Ion Source Complementarity for Characterization of Complex Organic Mixtures Using Fourier Transform Mass Spectrometry: A Review

ABSTRACT

Complex organic mixtures are found in many areas of research, such as energy, environment, health, planetology, and cultural heritage, to name but a few. However, due to their complex chemical composition, which holds an extensive potential of information at the molecular level, their molecular characterization is challenging. In mass spectrometry, the ionization step is the key step, as it determines which species will be detected. This review presents an overview of the main ionization sources employed to characterize these kinds of samples in Fourier transform mass spectrometry (FT‐MS), namely electrospray (ESI), atmospheric pressure photoionization (APPI), atmospheric pressure chemical ionization (APCI), atmospheric pressure laser ionization (APLI), and (matrix‐assisted) laser desorption ionization ((MA)LDI), and their complementarity in the characterization of complex organic mixtures. First, the ionization techniques are examined in the common direct introduction (DI) usage. Second, these approaches are discussed in the context of coupling chromatographic techniques such as gas chromatography, liquid chromatography, and supercritical fluid chromatography.

1. Introduction

In mass spectrometry, ionization is key because only the molecules that are efficiently ionized with a particular ion source are detected. There are many ionization techniques (Peacock, Zhang, and Trimpin ²⁰¹⁷). The physicochemical properties of the sample, such as state of matter, volatility, polarity, molecular mass, chemical, and thermal stability, determine their choice. The choice is also determined by the desired analytical approach: direct introduction (DI) or coupling with chromatography. The ionization technique used determines the ionized chemical space and therefore dramatically influences the obtained mass spectra. Albeit universal ionization sources have been searched for, to this day, no technique universally ionizes all the compounds in a complex organic mixture. The selectivity of an ionization technique may be important when groups of compounds, such as peptides or lipids, are targeted. However, when a broad overview of the existing compounds is needed, such selectivity can be problematic.

Soils, petroleum, sediments, carbonaceous chondrites, soot, and tholins are examples of highly complex organic mixtures involving numerous chemical functional groups and wide coverage of different chemical spaces within a relatively low mass range (50−600 Da) (Wilkins 1983; Laskin et al. ²⁰¹²; Smith et al. ²⁰¹⁸; Cooper et al. ²⁰²²). Note that by thus exemplifying and defining complex organic mixtures, we exclude, in this review, natural and anthropogenic high‐molecular weight molecules, such as proteins or polymers, and complex mixtures of peptides, which can form highly challenging mixtures to analyze. Nevertheless, complex organic mixtures of low molecular weight are found in many research fields such as energy, environment, health, planetology, and cultural heritage (Figure 1).

Figure 1.

Select most prevalent research fields involving complex organic mixtures analysis. [Color figure can be viewed at wileyonlinelibrary.com]

Historically, fossil‐derived mixtures, such as crude oil, coal, and bitumen, have for decades challenged analytical instrumentation due to their high molecular complexity at the isobaric and isomeric levels (Fernandez‐Lima et al. ²⁰⁰⁹; Abou‐Dib et al. ²⁰²²; Ruiz et al. ²⁰²³). The molecular description of these highly complex mixtures, often referred to as petroleomics (Marshall and Rodgers 2004), aims at being as exhaustive as possible. In addition, untargeted approaches are being developed in the study of air pollution (organic aerosols [OA]) (Schneider et al. 2022; Sueur et al. ²⁰²³), dissolved organic matter (DOM) (Nebbioso and Piccolo 2013; Zark, Christoffers, and Dittmar ²⁰¹⁷), natural organic matter (NOM) (Cooper et al. 2022; VanderRoest et al. ²⁰²⁴), metabolomics (Brown, Kruppa, and Dasseux ²⁰⁰⁵; Stopka et al. ²⁰¹⁹; Maia et al. ²⁰²³), planetology (Somogyi et al. 2005; Maillard et al. ²⁰¹⁸; Hertzog, Naraoka, and Schmitt‐Kopplin ²⁰¹⁹; Danger et al. ²⁰²⁰), or cultural heritage (van den Brink et al. ²⁰⁰¹; Garnier et al. ²⁰⁰⁹; Hertzog et al. ²⁰²³).

In the last decades, ultra‐high‐resolution mass spectrometry (UHRMS), mainly achieved by Fourier transform mass spectrometry (FT‐MS), has been used as the main analytical tool for the untargeted characterization of complex organic mixtures. It includes Orbitrap and Fourier transform ion cyclotron resonance (FT‐ICR) analyzers. Their principles have been widely reported in the literature (Marshall, Hendrickson, and Jackson ¹⁹⁹⁸; Makarov ²⁰⁰⁰). Both offer high performance in terms of mass resolution, mass accuracy, and dynamic range compared to other MS analyzers such as time of flight (TOF) (Figure 2) (Smith et al. 2018). Only with such ultra‐high‐resolution mass analyzers can both isobaric and elemental complexity be apprehended (Figure 2).

Figure 2.

Comparison between (a) isobaric and (b) elemental complexity found in complex organic mixtures. (c) Comparison of performance between a time‐of‐flight (TOF), Orbitrap Exploris 120, and FT‐ICR 12T mass analyzer with the zoom of a nominal mass justifying the use of FT‐MS analyzer for the molecular characterization of complex organic mixtures. Note that results showed correspond to internal results obtained from a pyrolysis pine oil, and lubricant additive samples. [Color figure can be viewed at wileyonlinelibrary.com]

Understanding the complementarity of the various ionization techniques is important to be able to apply each ionization source appropriately to study complex and heterogeneous organic mixtures. In this review, we will briefly describe five ionization techniques, namely electrospray (ESI), atmospheric pressure chemical ionization (APCI), atmospheric pressure photoionization (APPI), atmospheric pressure laser ionization (APLI), and laser desorption ionization (LDI), which correspond to the main ionization sources found for the analysis of complex organic mixtures. In a second part, the complementarity of these ionization sources will be displayed, with examples from the fields of energy, environment, omics and pharmaceuticals, planetology, and cultural heritage analysis. Finally, article investigating the complementarity of ionization sources used in hyphenated systems for isomer separation will be reviewed.

2. Ionization Sources

2.1. Introduction

Ionization sources can be classified into several categories (Gross 2017). The most common classification is based on their mode of operation, specifically whether the ionization takes place in a vacuum or at atmospheric pressure (API). Historically, the first ionization methods developed were vacuum ionization sources such as electron ionization (EI) and chemical ionization (CI) (Dempster 1918; Munson and Field ¹⁹⁶⁶). However, these ion sources are limited to volatile molecules and EI is considered as a harsh ionization source that induces a lot of fragmentation. Such harsh ionization sources are suitable for the characterization of complex organic mixtures only when they are used after a chromatographic separation, such as GC or GC×GC to associate molecular ions and fragments based on retention time alignment. Atmospheric pressure ionization sources are practically all considered soft ionization methods, primarily due to the collisional cooling, and are, therefore, suitable for complex organic mixtures even in the absence of chromatographic separation. In fact, the main ionization sources used for the molecular characterization of complex organic mixtures are atmospheric pressure ionization sources such as ESI, APCI, APPI, and APLI, as they are commercially available from all instrument vendors. The other main ionization source used for the molecular characterization of complex organic mixtures is the LDI source (Fenner and Daly 1966), which is not classified as API, because it can be performed under reduced pressure conditions. A scheme of each of these ionization sources which differ in their mode of operation is presented in Figure 3.

Figure 3.

Schematic description of the five main atmospheric pressure ionization sources used for the molecular characterization of complex organic mixtures. (a) Electrospray ionization (ESI), (b) atmospheric pressure chemical ionization (APCI), (c) atmospheric pressure photoionization (APPI), (d) atmospheric pressure laser ionization (APLI), and (e) laser desorption ionization (LDI). [Color figure can be viewed at wileyonlinelibrary.com]

Sample introduction in all API ionization sources exists for solid samples, solutions, or gas. Solids can be analyzed by evolved gas approaches such as atmospheric solid analysis probe (ASAP), or direct introduction probe (DIP). Laser desorption is limited to solids or dried solutions.

A graph showing the selectivity of ionization sources as a function of the polarity and molecular weight of the analytes is widely used in the literature (Halket et al. 2005; Li et al. ²⁰¹⁵; Wang et al. ²⁰¹⁵) and is shown in Figure 4. After a brief description of ESI, APPI, APCI, APLI, and LDI, we wish to illustrate how choosing an ionization source amounts to fitting on different “analytical eye glasses,” that focus on different chemical spaces in terms of polarity and molecular weight.

Figure 4.

Selectivity of the five main ionization sources used for the chemical description of the complex organic mixture in the function of the polarity and the molecular weight of analytes. The dotted line corresponds to the boundary between the area covered by conventional APCI and the area covered by APCI when hydrocarbons are used as dopants (indicated by the striped area). APCI, atmospheric pressure chemical ionization; APLI, atmospheric pressure laser ionization; APPI, atmospheric pressure photoionization; ESI, electrospray ionization; (MA)LDI, (matrix‐assisted) laser desorption ionization. [Color figure can be viewed at wileyonlinelibrary.com]

2.2. Functional Description of Ionization Sources

2.2.1. ESI

ESI is an ionization source that is highly selective for polar species. The ionization step, which takes place mostly in solution or in nanodroplets, can involve ionization competition and interference between analytes (King et al. 2000). Ionization is governed by the solution phase properties of the molecules and in particular their acidity and basicity. The pH and the pKa are therefore important factors (Wilm 2011). Negative mode ESI will favor the ionization of acidic molecules while positive mode will favor the ionization of basic molecules. The use of a modifier such as formic acid or ammonium hydroxide can also shift the equilibrium and thus promote ionization (Wilm 2011). In addition, ESI has the advantage of limiting ion dissociation. ESI results in multi‐charged ions for high molecular weight compounds such as proteins, but the ESI source also allows the ionization of small polar molecules. Ions with an even number of electrons, such as protonated molecules [M+H]⁺ in the case of a positive mode ionization and deprotonated molecules [M−H]⁻ in the case of a negative mode ionization, are thus mainly produced. The addition of a cation or an anion can also be involved to produce adduct ions such as [M+Na]⁺ or [M+Cl]⁻. The formation of adducts can be promoted by the addition of salts such as sodium acetate or ammonium acetate. Multimeric ions such as [2M+H]⁺ may also be observed.

In ESI, the analytes are introduced into the solution through a nebulizing needle (glass needle or steel capillary tube) at a low flow rate and then transferred to the gas phase according to a desolvation‐ionization phenomenon (Figure 3a).

The ESI source, described as a soft ionization source, has a wide ionization range in terms of mass, allowing the ionization of intact small and large molecules. It remains nevertheless dedicated only to the analysis of soluble molecules and more particularly polar to ionic compounds, which makes it a very selective source (Kruve 2020). In the case of complex organic mixtures, the use of a selective source can be essential for the detection of polar compounds present in low proportions. The detection of specific compounds has proven to be essential for understanding reactivity. For example, the unique signatures of DOM in Antarctic Lakes were obtained using ESI in negative ion mode (Niu et al. 2018). ESI in both positive and negative ion modes have also allowed selective detection of refractory nitrogen‐containing compounds in a vacuum gas oil (VGO), in feed and effluent.

2.2.2. Plasma Ionization and Direct Introduction Probes

2.2.2.1. APCI

The APCI source is a less selective ionization source. The ionization takes place in the gas phase and is governed by the properties of the gas phase, such as ionization energy (IE) and proton affinity (Marotta et al. 2003; Andrade et al. ²⁰⁰⁸). It is described as a robust, reliable, and very sensitive source with an efficient ionization less sensitive to matrix effects including ion suppression. Therefore, it can be considered for a wide range of applications. It is commonly used for the analysis of small polar and nonpolar compounds with low ESI response. However, operation at high temperatures requires thermally stable compounds, or induces some thermal degradation of labile species.

Ionization is initiated by a corona discharge generating nitrogen plasma (Carroll et al. 1975). The analytes are introduced through a tube heated at high temperatures (400°C−550°C) for volatilization (Figure 4). The use of a nebulizing gas facilitates the formation of the aerosol at atmospheric pressure. A high potential applied to a needle located just after this tube creates a corona discharge that ionizes the nebulizing gas and then the solvent molecules and analytes in the aerosol via primary N₂ ^+• ions (Andrade et al. 2008; Schwemer et al. ²⁰¹⁵). Since the first step is the formation of a nitrogen plasma, the driving force for ionization is therefore the IE of nitrogen which is 15.6 eV. This explains why APCI can in principle ionized a very wide range of molecules in terms of polarity. In the positive ionization mode, protonated molecules [M+H]⁺ or radical ions M^+• can be formed. The predominance of one or the other process depends on the solvent used and the physicochemical properties of the analyte such as the proton affinity and IE. In the presence of an aprotic solvent, the charge exchange process is favored, whereas conversely, the process of proton transfer is favored in the presence of a protic solvent or traces of moisture (Vaikkinen, Kauppila, and Kostiainen ²⁰¹⁶). Furthermore, the APCI source has shown some promise in the analysis of saturated hydrocarbons when small hydrocarbons are used as solvents (Manheim et al. 2020; Moulian et al. ²⁰²⁰).

Kim and Kim (2010) studied the evolution in the abundance of protonated and molecular ions of hydrocarbons and sulfur‐containing molecules in a petroleum sample while changing the solvent composition from toluene and methanol or acetonitrile to pure toluene: molecular ions for sulfur‐containing compounds increased in abundance in a similar manner to hydrocarbons with similar double bond equivalent (DBE) values suggesting benzothiophene groups. Similarly, Sanguineti et al. (2015) used APCI with heptane as a solvent to characterize bio‐oils from microalgae. They highlighted the presence of diverse classes of bio‐oil components such as sulfur‐, phosphorus‐, and chlorine‐containing compounds, which are unfavorable compounds for refining processes, in addition to the expected nitrogen‐containing compounds and hydrocarbon compounds. Hourani et al. (2013) used APCI with isooctane to detect polycyclic aromatic sulfur heterocycles (PASH) in VGO. In their study, they demonstrated the usefulness of the APCI source for characterizing of these species without any derivatization and with limited fragmentation.

Furthermore, direct introduction probes can be used with APCI for the analysis of solids, such as the ASAP and the direct insertion probe (DIP). DIP‐APCI FT‐MS has been used in particular for the characterization of additives (Lattimer and Polce 2011) and copolymers (Whitson et al. 2008).

2.2.3. Photoionization

2.2.3.1. APPI

As APCI, APPI is a less selective ionization source compared to ESI. APPI allows the ionization of a wide range of compound classes including nonpolar compounds that are difficult to detect by ESI (Huba, Huba, and Gardinali ²⁰¹⁶). It has a low susceptibility to ion suppression and matrix effects, wide linear dynamic range, and very low chemical noise since most impurities and solvent molecules are not ionized (Zimmermann 2021). As with the APCI source, the ionization occurs in the gas phase. Radical cations M^+• and to a lesser extent protonated species are formed by the absorption of V‐UV photons and a subsequent complex reaction scheme (Raffaelli and Saba 2003; Thomas et al. ²⁰²²). V‐UV photons can be generated by using plasma discharge lamps filled with inert gas, such as krypton (emitting at 10.6/10.0 eV), argon (11.7 eV), or xenon (9.6/8.4 eV) (Figure 3c) (McEwen ²⁰⁰⁷; Short, Cai, and Syage ²⁰⁰⁷; Hieta et al. ²⁰²¹; Neumann, Tiemann, et al. ²⁰²³; Neumann, Hansen, et al. ²⁰²³). The mechanism of photoionization with the APPI source also implies that the molecules are in the gas phase through an initial step of sample nebulization and evaporation. In the most direct gas phase reaction, the molecule, presenting a UV chromophore, will absorb a photon and form a radical ion with an odd number of electrons. Furthermore, detailed work, which would exceed the scope of this review, is presented by Kauppila (2021).

In practice, however, this direct APPI ionization is often not very efficient. To improve the ionization efficiency in APPI, a dopant can be added to the sample (Marotta et al. 2003; Kauppila, Syage, and Benter ²⁰¹⁷). In fact, in the presence of protic solvents, protonated molecules [M+H]⁺ can be produced (Syage 2004). Kauppila, Kostiainen, and Bruins (2004) showed that the use of anisole as a dopant allowed for a 100‐fold increase in the analytical sensitivity of analytes with low proton affinities in acetonitrile. The same observation was reported in the study of seven polycyclic aromatic hydrocarbons (PAHs) where the ionization efficiency was 1−2 orders of magnitude higher with dopant than without (Kauppila et al. 2002). The same group also proved that the charge exchange is favored for solvents with low proton affinity and that proton transfer is enhanced with the addition of a protic solvent.

DI of solids can also be used with APPI. In fact, Castilla et al. (2020) used DIP‐APPI FT‐MS to characterize lignocellulosic biomass samples. It was compared to DIP‐APCI. Using the same direct introduction probe for ground wood samples, the authors proved that APPI was more specific of lignin degradation compounds, while APCI covered a greater variety of oxygenated compounds. Podgorski et al. (2012) also used a custom APPI source, called desorption APPI (DAPPI), with a similar principle to DIP allowing for the in‐source desorption of intact bio‐char after combustion (at 250°C) and after pyrolysis (at 400°C).

APPI FT‐MS used by Chiaberge et al. (2013) for the untargeted characterization of crude oil samples allowing for the separation of 14 samples according to their geographical origin by using statistical analysis of the obtained FT‐MS data. Also, Vetere and Schrader (2021) used this ion source for the qualitative evaluation of sulfur‐containing species, which are difficult to detect, in heavy crude oil samples after chromatographic separation. In the field of environment, Headley et al. (2013) used APPI FT‐MS to evaluate the adsorption of components of the naphthenic acid fraction of oil sands by carbohydrate‐based materials.

2.2.3.2. APLI

The mechanism of APLI differs from conventional lamp‐based APPI in the way that instead of a single‐step V‐UV photoionization, APLI is based on a stepwise two‐photon ionization scheme (Figure 3d). The APLI mechanism is based on the classical resonantly enhanced multi‐photon ionization (REMPI) (Zimmermann 2021) but performed under elevated (atmospheric) pressure conditions. Due to its mechanism and similarity to APPI, APLI is particularly efficient for nonpolar aromatic species such as PAHs. However, APLI targets and enhances species whose first excited states have long lifetimes (Droste et al. 2005). Another specificity of the APLI is the need for a high photon density for the second absorption. This requires the use of a laser, because a high photon density also allows for a very efficient ionization (Streibel and Zimmermann 2014). Furthermore, the UV wavelength range used in APLI allows for a better ionization efficiency and selectivity than in APPI; in fact, in this range, most solvents and mobile phase molecules are transparent (e.g., H₂O, CH₃CN, and MeOH) (Constapel et al. 2005). Unlike APPI, APLI is less influenced by the use of a dopant (Kauppila, Kersten, and Benter ²⁰¹⁴).

APLI can be used for the detection of compounds in low concentration in the gas phase as shown by Schmidt et al. (1999). APLI is often found in coupling with separation techniques or compared to other ion sources, as in the study by Panda et al. (2011). Here they use APLI for the characterization of sulfur‐containing compounds in crude oil fraction and investigate the influence of the nebulization temperature on the resulting mass spectra. They also evidence the formation of mainly radical ions when using APLI. APLI is shown to efficiently ionize non‐ to low‐polarity compounds such as hydrocarbons, sulfur species, and low‐polarity oxygen species. In addition to aromatic hydrocarbons, aromatic sulfur compounds were also ionized by atmospheric pressure single photon ionization (ASPLI) (Rüger et al. 2021).

2.2.4. LDI

LDI allows the production of ions from a solid or liquid sample. The sample, in solution or powder form, is deposited on a conductive metal plate. Once the solvent is evaporated (for liquid deposition), the plate is introduced into the source, generally under vacuum. By laser irradiation, the photons tear off the molecules from the plate surface in the form of aggregates which will lead, in the case of ionization in positive mode, to radical ions M^+•, or protonated molecules [M+H]⁺ (Figure 4). Adducts such as [M+Na]⁺ can also be observed. Ionization mechanisms have been described in detail by Knochenmuss (2016). Compounds that are ionized show very little decomposition. Therefore, like the ESI source, it is considered a soft ionization source and will work well with complex mixtures containing compounds with chromophores. LDI can therefore be successfully applied to the study of aromatic rich samples for a variety of complex organic mixtures (Blackburn et al. 2017; Dhahak et al. ²⁰²⁰; Qian et al. ²⁰²²).

In some cases, the use of a matrix is essential. In fact, there are situations where the laser is unable to directly desorb and ionize the intact compounds. This is called matrix‐assisted laser desorption ionization (MALDI). MALDI is a process that involves the photo volatilization of a co‐crystallized sample with a small molecule, called matrix, in excess proportion (Dreisewerd 2003). The matrix is, in most cases, a small organic molecule with a high absorption coefficient at the laser wavelength. The analyte is incorporated into a solid or liquid organic mixture. Compared to LDI, it is the matrix that absorbs the energy of the photons emitted by the laser. The matrix consequently reaches an excited state. The main process reported is the desorption of the ions formed by proton transfer between the excited state of the matrix and the analyte (Liao and Allison 1995). Radical species may also be formed by a charge transfer process.

MALDI allows the study of molecules over a wide range of mass ( > 100,000 Da) and polarity. In addition, and compared to other techniques, it offers ease of use, high robustness, and high sensitivity. Its softness makes it a source of choice for the analysis of fragile species found in the biology/health research field such as neuropeptides (Kutz, Schmidt, and Li ²⁰⁰⁴), oligonucleotides (Li et al. 1996), or amino acids (Jing, Hao‐Yang, and Yin‐Long ²⁰⁰⁵). Furthermore, specific matrices can also be used to selectively enhance the ionization of a specific compound family present at a trace level, for example, vanadyl and nickel petroporphyrins in crude oil and its derivatives, as shown by Giraldo‐Dávila et al. (2018) and Sueur et al. (2023). In their respective studies, they showed that the use of an electron transfer (ET) matrix, instead of the commonly used proton‐transfer based matrix, greatly increased the selectivity of species with lower IE than the ET matrix. Indeed, the ET matrix hinders the ionization of higher IE by being easily ionized by MALDI and transferring their charge to the low IE analyte by an exchange of electrons. On the other hand, Mase et al. (2023) compared 12 matrices, including proton‐transfer and electron‐transfer matrices for the characterization of wood pyrolysis oils. They showed that some matrices enabled the ionization of a large number of species such as graphene oxide or dithranol, while others were specific to certain classes such as ferulic acid (FA) or 1,5‐naphtalenediamine (1,5‐DAN).

3. Ion Source Complementarity

The range of desolvation/desorption and ionization mechanisms makes each ion source more or less selective, either in terms of mass range or polarity. This selectivity, which at first sight might be seen as a drawback, is in fact a strength, since the combination of data obtained with different ionization sources can be used to achieve more comprehensive characterization of complex organic mixtures. In this review, we have selected several studies and reviews in the research fields of energy, environment, health, planetology, and cultural heritage that show how different ionization sources are used to provide complementary information on complex organic mixtures. Note that this list of articles is non‐exhaustive and limited to research published mainly within the last two decades.

3.1. Field of Energy

One of the key issues in the energy sector is crude oil and its derivatives. These samples have been extensively characterized by FT‐MS using various ion sources: ESI (Mapolelo et al. 2011), APPI (Abdul Jameel et al. 2019), APCI (Tose et al. 2015), APLI (Gaspar, Zellermann, et al. ²⁰¹²), and LDI (Terra et al. 2022). However, the use of several sources is often required to achieve the desired level of molecular information for both comprehensive and targeted descriptions of petroleum‐like mixtures.

Gaspar, Lababidi, et al. (2012) investigated the complementarity of five AP ion sources: ESI, heat‐assisted ESI (H‐ESI), APPI, APCI, and APLI, for the untargeted characterization of the asphaltene fraction of crude oil using a 12 T LTQ‐FT‐ICR MS. The asphaltene fraction was obtained by performing a SARA fractionation (Vazquez and Mansoori 2000). They show that APCI, APPI, and APLI allow the ionization of more compounds than ESI and its heated variant (Figure 5). In fact, ESI does not detect the nonpolar compounds that make up a large part of the asphaltene fraction, while nitrogen‐containing species dominate because they are selectively ionized by ESI. Nevertheless, ESI and H‐ESI are shown to be particularly efficient for nitrogen containing species. On the other hand, APPI, APCI, and APLI show a similar distribution of compound classes, that is, nonpolar species without heteroatoms. These three sources allowed the detection of compounds with lower DBE values and were also able to ionize a wider range of analytes, such as sulfur‐containing species, which were completely absent in the ESI and H‐ESI results. It is also noteworthy that APPI allowed the identification of the highest number of unique species.

Figure 5.

On the left: mass spectra enlarged at m/z 470 obtained with ESI, H‐ESI, APCI, APPI, and APLI with assigned molecular formulas. Formulas, found in APLI, are highlighted with red color. On the right: DBE distribution of the individual compound classes using the same ionization conditions for the characterization of asphaltenes. The scale is the number of assigned molecules (Gaspar, Lababidi, et al. ²⁰¹²). [Color figure can be viewed at wileyonlinelibrary.com]

With the aim of characterizing “supercomplex” crude oil mixtures, Panda, Andersson, and Schrader (2009) performed the analysis of a VGO fraction by FT‐ICR MS by using ESI, MALDI, APPI, APLI, and APCI as ion sources. Their study focuses on the ionization of PASH after a derivatization step that adds a methyl group to the PASHs. It is shown that only with ESI and MALDI can the derivatized form of the PASHs be observed. Furthermore, the compound distributions obtained by ESI and MALDI are very similar. Conversely, APPI, APLI, and APCI only give the nonderivatized forms, because the high‐temperature nebulizer, leads to the thermal removal of the methyl group. The results for all ionization sources highlight the ability of APCI to ionize high molecular weight species not detected by the other four sources. However, both APPI and APLI were able to ionize the widest range of compounds.

In their study to characterize of aged bitumen samples, Lacroix‐Andrivet et al. (2021) used negative mode ESI to selectively ionize polar and acidic heteroatom‐containing species, which include markers of aging. The results of this selective ionization were used to predict the aging state of bitumen samples. On the other hand, APPI was used in positive mode as a nonselective ion source to screen for low‐polarity compounds that could be used to determine the origin of the sample. Kondyli and Schrader (2020) evaluated the complementarity and combination of different AP ion sources, ESI, APPI, and APCI, for the characterization of light crude oil. This study shows the complementarity of the three sources as demonstrated in previous examples. The authors also perform dual‐source combinations to try to broaden the range of ionized molecules. Although some species can only be ionized by combining of two sources, these dual sources are more susceptible to matrix effects than the standard ion sources.

Compared to petroleum‐based materials and differences in molecular composition, the main source used for the characterization of bio‐oils, whether from lignocellulosic biomass or algae, is ESI. In fact, ESI is particularly indicated for the analysis of polar to midpolar species, with a selective ionization of acidic and basic compounds in negative and positive ion modes, respectively. Since these types of species are common in bio‐oils, ESI is adapted and therefore more widely used. However, several authors have shown the complementarity of ESI with other ionization sources to obtain a complete molecular description.

Hertzog et al. (2017) have used three ionization sources coupled with FT‐ICR MS to characterize a bio‐oil obtained from fast pyrolysis of lignocellulosic biomass. ESI and LDI were used in both positive and negative ion modes and APPI was used in positive ion mode only. As expected, the same molecular family C_xH_yO_z was observed with each ionization source. However, major differences were observed in the DBE versus carbon number plots showing the complementarity between these ionization sources. ESI allowed the ionization of compounds with higher oxygen atom content and lower DBE values whereas APPI allowed the ionization of less polar compounds with high DBE values and low oxygen atom content. LDI was considered to be intermediate in terms of DBE and the number of oxygen atoms between both ESI and APPI sources. However, a significant part of the features detected by LDI was specifically detected by this ionization method, providing important additional molecular information.

Mase et al. (2022) have also investigated the contribution of ionization sources in combination with FT‐ICR MS for the characterization of bio‐oils. APCI was compared with ESI and APPI, in the positive ion mode. Different dopants in ESI and dilution solvents in APCI were also compared. The APCI source proved to be particularly attractive because it ionized aliphatic molecules in addition to the other bio‐oil components also ionized by APPI and ESI.

3.2. Environmental Organic Matter

Motivated by environmental and climatic changes, the in‐depth analysis of environmental samples has gained more and more interest. Environmental complex organic mixtures encompass a wide variety of samples, from seawater to soil and ambient air and thus require specific analytical methods (Kim et al. 2022).

Among the complex organic mixtures in the environment, DOM plays an important role in the global carbon cycle and the resulting physical or chemical processes (Niu et al. 2018). Kurek et al. (2020) attempted at deciphering the complexity of DOM using ESI (−) and APPI‐FT‐ICR MS. They analyzed water samples from three sites in Florida: Suwanee River, Kissimmee River, and Wakulla Springs. Here, the dopant‐dependent selectivity of APPI is demonstrated using two dopants, tetrahydrofuran (THF) and toluene. Indeed, THF is shown to be capable of ionizing not only N‐containing species, similar to toluene, but also S‐containing species. However, THF and toluene seem to be complementary as they both allow the ionization of different species. Multi‐doped APPI and ESI (−) thus allow the detection of a wide range of compounds.

In addition to the biogenic DOM, anthropogenic DOM is of great research interest and can be found in the wastewater of cities or factories. He et al. (2021) studied wastewater samples from a petroleum refinery using ESI (+/−) and APPI‐FT‐ICR MS. The use of three different ionization techniques enabled the identification of compounds specific to anthropogenic DOM when compared to biogenic DOM from the Suwannee River standard sample. When examining a specific molecular class (N₁, O₁ or HC) in Figure 6, each ionization source allowed the ionization of different compounds that provided a more in‐depth description of the sample, demonstrating the complementarity of the three ionization sources. In fact, as shown in Figure 6, ESI (−) favored the ionization of neutral nitrogen‐containing species, such as carbazoles or benzocarbazoles, which are easily deprotonated in solution, while ESI (+) favored the ionization of basic nitrogen‐containing species, such as those with quinoline or pyridine moieties. Regarding the oxygen‐containing species, ESI (−) proved to be very efficient for the ionization of phenolic compounds (DBE 4) and fatty acids. On the other hand, APPI allowed the ionization of species with higher aromaticity in each compound class.

Figure 6.

DBE versus carbon number plots of N₁, O₁, O₂, and CH class detected by ESI (+/−) and APPI (+) FT‐ICR MS for a refinery wastewater DOM showing the complementarity of the ionization sources. Bubble size indicative of the abundance. The supposed structures were proposed based on the molecular composition (note: the structures are just possible but not detected) (He et al. 2021). [Color figure can be viewed at wileyonlinelibrary.com]

NOM, or humus, is produced by entropy‐driven decay of plant and animal tissues. Its complexity has been shown by Vinci, Piccolo, and Bridoux (2022) by using both ESI in negative mode and APPI in positive mode with an LTQ‐Orbitrap. They analyzed six fractions of a soil sample, three organo‐soluble and three water‐soluble. They showed the selectivity of APPI toward the aromatic and/or nitrogen‐containing species, while ESI (−) favored compounds with high oxygen/carbon ratio, which were not detected by APPI. The use of two ion sources allowed for a more detailed molecular characterization of the six fractions. In addition, Luo and Schrader (2021) also used ESI (+/−) and APPI FT‐MS to characterize contaminated soil samples.

Another category of environmentally relevant complex mixtures is OA. OA can originate from either biogenic or anthropogenic sources and elevated concentrations of OA in inhabited areas can pose significant health risks (Mauderly and Chow 2008). In addition, their impact on the environment is not negligible and has been studied since the 1950s (Haagen‐Smit ²⁰⁰²). Traditionally, OA has been characterized by a combination of sophisticated extraction and separation techniques, and targeted analyses (Johnston and Kerecman 2019). However, in the last decades, untargeted analyses of OA have emerged to identify as many compounds as possible (Nizkorodov, Laskin, and Laskin ²⁰¹¹), an approach for which FT‐MS is well suited. Kuang et al. (2018) characterized anthropogenic fine particles from the ambient air in Hong Kong by ESI and APPI‐FT‐ICR MS in positive ion mode. APPI allowed the detection of highly aromatic PAH while ESI allowed the detection of more aliphatic species, especially in the case of nitro‐PAHs. This study identified many more nonfunctionalized PAHs than the usually targeted 16 high‐priority PAHs and the 74 PAHs described in the cancer risk assessment. OA can also be divided into two fractions: water‐soluble organic aerosol (WSOA) and water‐insoluble organic aerosol (WIOA). Both fractions can be studied as shown by Choi et al. (2017) who compared WSOA and WIOA using ESI and APPI‐FT‐ICR MS operated in positive mode. It is also possible to consider only one of the fractions, as done by Schneider et al. (2022) in their comparison of biogenic and anthropogenic OA using ESI in both negative and positive modes.

3.3. Lipidomics and Metabolomics and Pharmaceutical Analysis

Metabolomics studies all the metabolites in a cell, an organ, or an organism (Brown, Kruppa, and Dasseux ²⁰⁰⁵). The chemical complexity of the mixture requires the use of several ion sources. Gray and Heath (2005) used the complementarity of ESI and APCI sources coupled to FT‐ICR MS to determine the cold acclimation mechanism in Arabidopsis by metabolic fingerprinting. The data generated by both sources were aligned and processed to obtain a list of all unique masses. Then, statistical analysis using PCA was used to conclude the acclimation of the plant. Similarly, Ahoroni et al. (2002) used ESI and APCI in a complementarity manner for nontargeted metabolomic analyses, focusing on strawberry. By combining both sources using both polarities, they identified about 5500 different mass peaks. They nevertheless underlined the fact that mass spectrometry alone did not allow the identification of isomers and that therefore potentially a much higher number of species existed. Despite this drawback, the study has revealed the process of accumulation of compounds that serve as raw materials and building blocks for the production of ripening‐associated metabolites. Calabrese et al. (2023) used MALDI with different matrices in addition to the classically used ESI for the analysis of plant root exudate metabolites. Plant root exudates are complex mixtures with a high salt concentration which induces matrix effects in ESI, while MALDI is more salt tolerant. Here, the complementarity of the two sources was demonstrated, as ESI and MALDI showed selectivity toward small molecules with low aromaticity and highly aromatic molecules, respectively.

Lipidomic studies are interested in the specific role of lipid molecular species in health. Lipids consist of a large number of structural and functional molecular species covering a wide range of mass and polarity. The review by Hu et al. (2009) provides a good overview of the recent advances in mass spectrometry in lipidomics. The identification of all the lipids in a mixture is complicated in a single analysis, so the use of different ionization sources is recommended. The ESI source in positive and negative ion mode is the most used ion source. It is mainly used for identification with tandem mass spectrometry MS/MS or coupled with liquid chromatography for quantification. In ESI, signal suppression can be observed and is a risk. For this reason, the MALDI source is often used as a complement because of its lower sensitivity to matrix effects. However, MALDI cannot be used in on‐line coupling with liquid chromatography and has limitations for quantification. Nevertheless, MALDI has seen in great development because it allows an imaging approach (Goto‐Inoue et al. ²⁰¹¹).

In the field of pharmaceutical analysis, Deschamps et al. (2023) have reviewed the principles of the two mass spectrometers, Orbitrap and FT‐ICR, and highlighted their application, development, and future perspectives. They showed the main use of the ESI source in this field. However, due to the important matrix effect found in these samples, they insist on the quality of the sample preparation. Other ionization sources, such as MALDI, APPI, and APCI, are used in cases where sample preparation is not possible but are less frequently found. Still related to matrix effects, coupling with liquid chromatography is widely used for the characterization of these types of samples. In this case, it is also the ESI source that will be most used. Finally, there is an interest in mass imaging using LDI and MALDI sources, which allow a great understanding of the bio distribution, metabolism, and accumulation of drugs in the human body.

3.4. Planetology and Astrochemistry

Mass spectrometers are regularly carried on spacecraft exploring the solar system to understand the origin, distribution, and evolution of organic matter. Innovative space instruments are therefore needed. In their study, Selliez et al. (2020) compared two instruments, a home‐built space‐designed Orbitrap, and a laboratory Benchmark FT‐ICR mass spectrometer. The first one was composed of a laser ablation ionization (LAb) source and the second one with an LDI source. Except for the difference observed due to the dynamic range, resolution, and mass range of the analyzers, both ionization sources presented similarities and allowed the ionization of the same nitrogen‐containing compounds.

Few comparisons of ionization sources have been performed on planetology samples, mainly due to the scarcity of samples and the small quantities available. In addition, the main works on planetology samples used LDI as the ionization source (Maillard et al. 2018; Danger et al. ²⁰²⁰). Hertzog, Naraoka, and Schmitt‐Kopplin (2019) have compared APPI and ESI in positive and negative ion modes for the characterization of the Murchison meteorite. Among the 16,000 unique features detected, only 4% were common to all analyses, clearly demonstrating the complementarity of the APPI and ESI ionization sources. The main difference is in the classes of ionized molecules (Figure 7). ESI (−) allowed the ionization of oxygen and sulfur‐containing molecules whereas ESI (+) allowed the ionization of nitrogen‐containing molecules. APPI (+) allowed the ionization of less polar molecules such as hydrocarbons and heteroatomic compounds with lower O/C, N/C, and S/C ratios compared to ESI. The same observation was made by Naraoka et al. (2023) who also compared ESI and APPI in both ionization modes for the characterization of the carbonaceous asteroid Ryugu. ESI allowed the ionization of molecules with higher O/C ratios, while APPI showed a narrower distribution at lower O/C ratios when ionizing hydrocarbon compounds.

Figure 7.

Venn diagram achieved from data obtained in ESI and APPI‐FT‐ICR MS in both positive and negative ionization modes. The heteroatom class distributions and the corresponding weighted average values gathered in the tables are given for specific and common features (Hertzog, Naraoka, and Schmitt‐Kopplin ²⁰¹⁹). [Color figure can be viewed at wileyonlinelibrary.com]

3.5. Cultural Heritage

Among the characterization techniques used in cultural heritage studies, UHRMS has shown particular interest, although it is considered a destructive technique and requires a sample collection, which is not always easy to obtain. Therefore, ionization techniques that require little sample quantity and sample preparation are preferred, such as MALDI.

Some work had focused on the complementarity of the ionization sources to determine which would provide better information. Vahur et al. (2012) have compared the MALDI and the APCI in positive ion mode for the analysis of dammar resin, which is a widespread varnish material found in paintings. They used 10 different solvents, because, with the years, the properties of the resin such as solubility changes, and it is therefore important to find the appropriate dilution solvent. They conclude that both MALDI and APCI sources are needed to get a complete picture of the dammar. MALDI allowed the ionization of the more polar components of the resin and APCI allowed the ionization of nonpolar components such as hydrocarbon and slightly oxidized hydrocarbon molecules. Similarly, Peets et al. (2020) have compared MALDI and ESI in both positive and negative ion modes for the characterization of red pigments from various plants, fungi, and insects. ESI allowed the ionization of sufficient compounds to discriminate between dyes (such as chloro‐compounds, several anthraquinones or carminic acid) and could be, therefore, considered a definitive technique, MALDI proved to be interesting in the case of some red dyes by giving mass spectra highly characteristic of the dye source, directly from suspensions or the fiber sample.

Another interesting work compared the four main ionization sources found, APCI, APPI, ESI, and LDI in negative ion mode for the characterization of Scotch whisky (Kew et al. 2018). Approximately 700 common molecular formulas were observed but each ionization source allowed the ionization of unique species. Among the four ionization sources, LDI lead to the highest number of molecular formulas assigned. However, the van Krevelen diagrams (Figure 8), of the molecular assignments obtained with LDI presented a distribution higher mass cask extractives, likely lignin‐derivatives compounds whereas APCI and ESI allowed the ionization of lower mass highly oxygenated compounds such as carbohydrates (high H/C and high O/C) and more aliphatic molecules such as fatty acids or alcohols (high H/C ratio and low O/C). The authors have thus shown each ionization source should be chosen according to the desired information.

Figure 8.

van Krevelen diagrams obtained with APCI, APPI, ESI, and LDI‐FT‐ICR‐MS in negative ion mode showing the complementarity of the ionization sources for the characterization of a scotch whisky (Kew et al. 2018). Each assignment from individual ionization modes was plotted with the size of the glyph representing the median abundance across four whisky samples. The color represents the mass of the peak. The number in the bottom right corner shows the number of unique formulas identified for each ionization source across four samples. [Color figure can be viewed at wileyonlinelibrary.com]

4. Ionization Sources for Hyphenated Chromatographic Methods

4.1. Introduction

Hyphenating chromatography and FT‐MS means, potentially, combining the high mass accuracy required to determine accurate molecular formulas and resolve isobars with the separation of isomers and reduction of the ion suppression effects, extending the dynamic range. However, coupling chromatography to FT‐MS involves some compromise and is not as straightforward as hyphenation with other mass analyzers such as TOF. Indeed, FT‐MS requires time to reach its highest resolving power while chromatographic peaks are transient. The hyphenation of FT‐MS with separations methods was recently reviewed by Gosset‐Erard et al. (2023). In Table 1, we have listed an overview of articles that involve hyphenated chromatography (including liquid chromatography, supercritical fluid chromatography, and gas chromatography) and FT‐MS techniques for the characterization of complex organic mixtures. The ionization source, operating mode, and sample preparation used for each application are specified. In the literature and because of the gas phase or solution state of the sample, the APCI source is widely used for gas chromatography, while the ESI source is preferred for liquid chromatography. In the following paragraphs, we chose to focus on research involving different ionization methods giving complementarity results with the same chromatographic system.

Table 1.

Overview of some hyphenated chromatographic and FT MS techniques. The ionization source, mode of use, and sample preparation used were specified for each.

References	Type of matrix	Hyphenated method	Ionization source	Mode	Sample preparation
Liquid chromatography
Shen et al. (2017)	Microbial and mammalian proteomic samples	UHPLC‐Orbitrap MS	ESI	Positive	Cells preparation
Pollier et al. (2011)	Transgenic M. truncatula hairy roots	LC‐FT‐ICR MS	ESI	Negative	Extraction
Scherling et al. (2010)	Plant species	UPLC‐FT‐ICR MS	ESI	Negative	Extraction
Pereira, Bhattacharjee, and Martin (2013); Pereira, Islam, et al. (2013)	Oil sands process‐affected water from Syncrude Canada	LC‐Orbitrap MS	ESI	Positive and negative	Filtration with 0.45 µm filter Acidification to pH = 2 and extracted by dichloromethane The residue dissolved in acetone
Triebl et al. (2017)	Complex biological sample	HPLC‐Orbitrap MS	ESI	Positive and negative	MTBE‐based lipid extraction
Undri et al. (2015)	Pyrolysis bio‐oils	LC‐Orbitrap MS	ESI	Positive and negative	Dilution
Xia et al. (2021)	Diesel, vacuum gas oil, and heavy oil	LC‐Orbitrap MS	ESI	Positive and negative	Dilution
Pereira and Martin (2015)	Oil sands from Northern Alberta	LC‐Orbitrap MS	APCI	Positive and negative	Extraction with dichloromethane
Reymond, Dubuis, et al. (2020)	Water‐soluble fraction of biomass	LC‐FT‐ICR MS	ESI and APCI	Positive and negative	Liquid−liquid fractionation Filtration
Reymond et al. (2019)	Fast pyrolysis bio‐oil from softwood	LC‐FT‐ICR MS	ESI and APCI	Positive and negative	Dilution
Lababidi et al. (2013)	Nitrogen‐rich crude oil	LC‐FT‐ICR MS	Home‐built APLI	Positive	Dilution
Lababidi and Schrader (2014)	Crude oil sample	HPLC‐FT‐ICR MS	ESI, APPI, APCI, and APLI	Positive	Desaphalted
Han et al. (2021)	Natural organic matter	LC‐FT‐ICR MS	ESI	Negative	No sample preparation
Jennings et al. (2022)	Wastewater effluent	LC‐FT‐ICR MS	ESI	Positive and negative	Ozonation and extraction
Qi et al. (2021)	Surface water and rainwater	LC‐FT‐ICR MS	ESI	Negative	Dilution
Rowland et al. (2021)	Arabian heavy gas oil distillate and Mackay Canadian bitumen distillate	LC‐FT‐ICR MS	APPI	Positive	Dilution
Kim et al. (2019)	Suwannee River fluvic acid and Upper Mississippi River NOM	LC‐FT‐ICR MS	ESI	Negative	Dilution
Supercritical fluid chromatography
Reymond, Le Masle, et al. (2020); Reymond et al. (2021)	Vacuum gas oil	SFC‐FT‐ICR MS	APPI	Positive	Dilution in the mobile phase
Pereira and Martin (2015)	Oil sands from Northern Alberta	SFC‐Orbitrap MS	APCI	Positive and negative	Extraction with dichloromethane
Yamada et al. (2013)	Lipids in animal plasma	SFC‐Orbitrap MS	ESI	Positive and negative	Extraction using Bligh and Dyer's method
Gas chromatography
Szulejko and Solouki (2002)	Gasoline samples	GC‐FT‐ICR MS	EI	Positive	No sample preparation
Luo, Heffner, and Solouki (2009)	Gasoline samples	GC‐FT‐ICR MS	CI	Positive	Headspace
Ortiz et al. (2014)	Naphthenic acid fraction compounds of oil sands	GC‐FT‐ICR MS	EI and CI	Positive	Filtration under vacuum, acidification to pH = 4.5, and extraction with Strata X‐A solid phase extraction sorbent Methylation using BF3/methanol
Ayala‐Cabrera et al. (2021)	Marine sediments	GC‐Orbitrap MS	APPI	Positive and negative	Soxhlet extraction
Kondyli and Schrader (2019)	Gas condensed of Arabic origin	GC‐Orbitrap MS	EI and APPI	Positive	Dilution
Barrow, Peru, and Headley (2014)	Oil sands from the Athabasca River basin	GC‐FT‐ICR MS	APCI	Positive	Filtration under vacuum, acidification to pH = 4.5, and extraction with Strata X‐A solid phase extraction sorbent Methylation using BF3/methanol
Palacio Lozano et al. (2022)	Pyrolysis softwood bio‐oil and its esterified products	GC‐FT‐ICR MS	APCI	Positive	Esterification with n‐butanol in the presence of H2SO4
Thomas et al. (2019)	Sediment core	GC‐FT‐ICR MS	APCI	Positive	Extraction with dichloromethane
Schwemer et al. (2015)	Particulate matter from diesel fuel and heavy fuel oil combustion	GC‐FT‐ICR MS	APCI	Positive	Extraction with methanol and dichloromethane Filtration with PTFE membrane filter
Zuber et al. (2016)	German brown coal pyrolysis	GC‐FT‐ICR MS	APCI	Negative	Extraction with benzene
Rüger et al. (2015)	Petroleum + lignocellulosic biomass	GC‐FT‐ICR MS	APCI	Positive	Thermogravimetry
Zacs, Perkons, and Bartkevics (2019)	Food sample	GC‐FT‐ICR MS	APCI	Positive	GPC then dilution in toluene
Smit et al. (2015)	Diesel sample	GC‐FT‐ICR MS	APCI	Positive	Extraction with methanol
Rüger et al. (2021)	Crude oil sample	GC‐FT‐ICR MS	APSPLI	Positive	No sample preparation
Benigni et al. (2016)	Crude oil sample	GC‐FT‐ICR MS	APLI	Positive	No sample preparation
Muller et al. (2012)	Gas oil	GC×GC‐FT‐ICR MS	APLI	Positive	No sample preparation

4.2. Complementarity of Ionization Sources in Hyphenated Methods

4.2.1. Gas Chromatography

The use of API sources for GC‐MS was reviewed in 2023 by Ayal‐Cabrera et al. (2023).

Several studies have reported the complementarity of ionization sources with hyphenated gas chromatography and FT‐MS techniques. In 2014, Ortiz et al. (2014) used gas chromatography coupled with FT‐ICR MS (9 T) for the characterization of environmental samples. They used the EI and the CI with different gases. The authors showed that the observed distribution and composition were influenced by the ionization technique and the choice of reagent. EI led to a pronounced hydrocarbon contribution, CI with ammonia led to more‐prominent nitrogen‐containing species, and CI with methane produced results intermediate between the other two methods. Kondyli and Schrader (2019) compared the EI and the APPI sources for the characterization of volatile gas condensate by GC‐Orbitrap MS. Both sources allowed the ionization of different compound classes such as hydrocarbons compounds and heteroatom‐containing compounds (oxygen, nitrogen, and sulfur) but with different intensities (Figure 9). Especially, the heteroatom‐containing species were well ionized with the APPI. The lower intensity of the sulfur‐containing compounds in EI was explained by the fact these species were unstable and fragmented very easily. They showed that the high energy of EI generated a lot of fragmentation and made the complexity even higher and therefore the characterization more difficult. However, they highlighted the use of EI low energy, which allowed the reduction of the complexity and therefore the distinction between fragment and molecular ions. Finally, they demonstrated that APPI is a good complementary source of EI for the characterization of complex organic mixtures allowing the detection of a wide number of compounds covering a large polarity range.

Figure 9.

Comparison of class distribution based on the number of the main assigned compounds between EI high and low energy and APPI for the characterization of gas condensate by GC‐Orbitrap MS (Kondyli and Schrader 2019). [Color figure can be viewed at wileyonlinelibrary.com]

4.2.2. Liquid Chromatography

The coupling of FT‐MS with chromatography allows to decrease the ionization competition and signal suppression that can occur with direct infusion ESI. Reymond et al. (2019) compared LC‐FT‐ICR and DI FT‐ICR using an ESI source in negative ion mode for the characterization of lignocellulosic‐based bio‐oil. The relative abundance of oxygenated compound classes was bimodal in LC‐FT‐ICR with a first distribution including compounds with 2−10 oxygen atoms centered on O4 and a second distribution with compounds with 11−18 oxygen atoms centered on O15. The relative distribution obtained in DI‐FT‐ICR was centered in O10 with compounds ranging from O3 to O18. Liquid chromatography enhanced the detection of heavy and moderately polar molecules. To obtain these good results in LC‐FT‐ICR, they shown that optimization of the desolvation and ionization steps is necessary to obtain a well‐controlled analytical method. Comparing ESI and APCI, they detected more than 2000 additional feature with ESI in negative ion mode compared to around 300 with APCI. Similar results were obtained by the same team using ESI and APCI sources in both positive and negative ion modes in the study of liquid−liquid extraction fractions from lignocellulosic biomass (Reymond, Le Masle, et al. ²⁰²⁰). However, the combination of both ionization sources is needed to achieve a comprehensive characterization of the sample by detection of lipid, phenolic, and carbohydrate compounds (Figure 10).

Figure 10.

van Krevelen diagram obtained by ESI and APCI in both positive and negative ion modes in LC‐FT‐MS for the characterization of liquid−liquid fractions (AQ for aqueous and ORG for organic) of lignocellulosic‐based biomass (Zark, Christoffers, and Dittmar ²⁰¹⁷). [Color figure can be viewed at wileyonlinelibrary.com]

Lababidi and Schrader (2014) compare four ionization techniques (ESI, APPI, APCI, and APLI) for the characterization of deasphalted crude oil using high‐performance liquid chromatography (HPLC) FT‐ICR MS. Using normal‐phase liquid chromatography with n‐hexane and isopropyl alcohol mobile phases, the chromatograms showed two distinct peaks, which included different distributions of compound classes (Figure 11). The first peak included hydrocarbons and other nonpolar compounds, while the second peak included nitrogen‐containing species. As expected, ESI allowed the ionization of only nitrogen‐containing compounds. APPI offered the widest spectrum of classes allowing the identification of nonpolar oxygen‐containing compounds that were not observed, or in very low abundance with both APCI and APLI sources.

Figure 11.

Class distribution based on the number of assigned formulas using ESI, APPI, APCI, and APLI sources obtained for the two peaks separated in LC‐FT‐ICR MS on a deasphalted crude oil (Lababidi and Schrader 2014). [Color figure can be viewed at wileyonlinelibrary.com]

5. Conclusion

The interest of FT‐MS lies in its unsurpassed resolving power and mass accuracy. These features allow the in‐depth characterization of complex organic mixtures, helping us to better understand our environment, from biological processes at the cellular level to the origin of life by analyzing the organic content of meteorites. However, taking advantage of complementary ion sources is important to make FT‐MS analysis as comprehensive as possible. In particular, the use of a highly selective ion source such as ESI can be of great interest for the detection of low abundance polar species but can lead to the loss of the big picture when nonpolar molecules are the main components of the mixture.

Indeed, as shown in this review, the use of multiple ionization techniques significantly expands the chemical space accessible to the user in terms of polarity, aromaticity, and mass. This complementarity of sources can also be added to the separation power of techniques compatible with FT‐MS, such as LC, GC, or SFC, to achieve a higher level of complexity in characterization, that is, isomer separation. However, the more ion sources and hyphenations are set up, the more complex the resulting data. This, and the increase in measurement time, are the main factors limiting the number of ion sources and separation techniques used for the characterization of complex organic mixtures. Therefore, efforts are constantly being made to shorten and simplify the processing of such complex data, and in some cases, the analysis can be automated to reduce experimental time. In general, it is not always feasible to use multiple ionization methods, but the choice of selective or nonselective ionization method is critical to obtaining the expected information. Overall, FT‐MS users must make a compromise between time and depth of characterization.

Author Contributions

Charlotte Mase: conceptualization, visualization, writing–original draft preparation. Maxime Sueur: conceptualization, visualization, writing–original draft preparation. Hélène Lavanant: supervision, conceptualization, writing–review and editing. Christopher Paul Rüger: supervision, conceptualization, writing–review and editing. Pierre Giusti: supervision, conceptualization, writing–review and editing. Carlos Afonso: supervision, conceptualization, writing–review and editing.

Data Availability Statement

Data are available on request from the authors.