Introduction

Ion mobility spectrometry (IMS) has undergone a series of developments in recent years that have enabled the technique to be applied in new research and application areas. Indeed the utility of IMS has expanded well beyond its established role for detecting explosives, narcotics and chemical warfare agents (CWAs). Recent applications include the analysis of pharmaceutical actives and formulations, and structural and conformational studies in biological and biomedical research [1–4]. Many of these developments arise from the hyphenation of IMS with mass spectrometry (IM-MS) [5].

In its simplest form, the drift velocity (v d, m s−1) of an ion in a drift tube (DT-IMS) under the influence of a uniform electric field gradient is proportional to the strength of the electric field gradient (E, V cm−1) with the proportionality coefficient being ion mobility (K, cm2 s−1 V−1), defined by Eq. (1):

$$ K = \frac{{v_d }}{E} = \frac{{d/t_d }}{V/d} = \frac{{d^2 }}{{t_d V}} $$
(1)

Where t d is the drift time, the time taken for ions to traverse the distance d (drift length), between the ion shutter grid and the detector, v d is the ion velocity and V is the potential difference applied to the drift tube. This relationship holds under low field conditions (≤ 2 Td, where 1 Td = 10−17 V cm2). The relationship between ion size, charge and mobility is described by the Mason-Schamp equation, a simplified form of which is given in Eq. 2:

$$ K = \left( {\frac{3ez}{16N}} \right)\left( {\frac{{2\pi }}{{\mu kT}}} \right)^{1/2} \frac{1}{{\Omega_D }} $$
(2)

Where e is the electronic charge (1.602 × 10−19 C) and z is the charge on the ion, N is the number density of the drift gas, µ is the reduced mass of the ion and drift gas, k is the Boltzmann constant (1.381 × 10−23 J K−1), T is the temperature of the drift gas and Ω D is the collision cross section (i.e. size and shape) of the ion. The units of the Mason-Schamp equation are nominally C s kg−1, equivalent to m2 s−1 V−1. Provided the conditions within the drift cell, such as temperature and buffer gas pressure are kept constant, the mobility of an ion is dependent on reduced mass, charge state and collision cross section.

It is common practice for comparison purposes and to correct for instrumental variations, to express ion mobility as reduced mobility (K0), normalised to standard conditions of temperature (T in Kelvin) and pressure (P in Torr) of the drift gas.

$$ K_0 = K\left( {\frac{273}{T}} \right)\left( {\frac{P}{760}} \right) $$
(3)

One of the key challenges in IMS is to measure the mobility of an ion accurately. Precision and resolution are limited by uncertainties in measuring the exact distance between the ion grid and detector (drift tube length), the temperature, pressure and composition of the drift gas, and ambiguity in timings such as ion injection and ion drift times. These uncertainties arise as a result of variations in instrument design, effects of temperature (causing thermal expansion or contraction of the drift tube) and pressure within the drift tube. Maintaining an accurately defined homogeneous electrical field is also non-trivial, and inhomogeneity in the electric field strength results in variations in ion drift times [6]. Thus, determining absolute ion mobilities using Eq. 1 may be problematic. This practical problem may be addressed by the adoption of mobility standards, where a compound of known mobility K 0ref (Eq. 4) is used to calibrate the drift tube and to determine the reduced mobilities of other analytes.

$$ K_{{0_{ref} }} = A.\frac{{d^2 }}{{t_d V}}\left( {\frac{273}{T}} \right)\left( {\frac{P}{760}} \right) $$
(4)

The correction factor A is usually incorporated with the drift length, d, and field defining voltage, V to give the cell constant, C, which is determined experimentally using the observed drift time of the mobility standard, Eq. 5.

$$ C = K_{{0_{ref} }} t_d \left( {\frac{273}{T}} \right)\left( {\frac{P}{760}} \right) $$
(5)

The reduced mobility of an unknown may be calculated using the cell constant or from the ratio of the drift times of the reference standard and the unknown compounds, Eq. 6.

$$ K_{0u} = K_{{0_{ref} }} \times \frac{{t_{{d_{ref} }} }}{{t_{du} }} $$
(6)

Where K 0u is the reduced mobility of the unknown, t dref is the drift time of the reference standard and t du is the drift time of an unknown compound.

In addition to low-field IMS, differential mobility spectrometry (DMS), also known as field asymmetric ion mobility spectrometry (FAIMS), presents further challenges for the determination of the relative low field and high field mobilities involved with this ion mobility variant. These are also sensitive to changes to the parameters discussed above, particularly with respect to the drift gas composition.

Several IMS reference standards have been proposed for determining cell constants, reduced mobilities and collision cross section for specific applications. For example, nicotinamide [7] for narcotics analysis, and dipropylene glycol monomethyl ether (DPGME) [8] and methylsalicylate [9] for field testing and verification of chemical agent monitors, as simulants for nerve and blister agents respectively. Nevertheless, the adoption of internationally agreed standards and standardisation procedures for a growing range of IMS applications and techniques is at a low level, particularly within proteomic and pharmaceutical analysis. Indeed there are no internationally accepted reference standards for the calibration and evaluation of mobility spectrometers (IMS or DMS/FAIMS systems). A set of IMS reference standards would allow traceable instrument calibration, and ensure that the validations of ion mobility spectrometry experiments could be undertaken in a practicable manner.

The widespread adoption of standards would also facilitate and support inter-laboratory exchange of mobility data obtained from different IMS systems generated by different operators and laboratories. The routine use of reference standards in ion mobility spectrometers would aid the validation of IMS methods, reduce ‘false positive’ results, increase the certainty of analyte assignments and lead to improvements in repeatability, reproducibility and robustness of IMS measurements, enabling cross examination/inter-comparison of results obtained using a diverse range of IMS systems and between different users and laboratories. Furthermore, IMS standards incorporated into quantitative mobility measurements would enable reliable spectral alignment; most important when combining very large data sets such as those generated within clinical and environmental studies.

Current status of standardisation in IMS

Several standards, listed in Table 1, have been reported for ion mobility-based applications and to test mobility spectrometers. The majority of the standards listed in Table 1 are readily available from most chemical catalogues, with the exception of TNT and DMMP, which may limit their application as reference standards. It is perhaps helpful to categorise mobility standards in terms of drift cell pressure and ion mode. Ambient pressure systems in the positive mode are generally used for the detection of drugs of abuse (narcotics), nerve agents, and in some pharmaceutical applications. Ambient pressure applications in the negative ion mobility mode include the detection of explosives, acidic gases and blister agents. Lower pressure drift tubes (50–100 Pa) are mostly associated with the hyphenation of ion mobility spectrometry to mass spectrometry in proteomic, metabolomic and pharmaceutical studies [4, 5, 10].

Table 1 Examples of IMS reference standards/calibrants currently being used for positive and negative ion mobility mode (*Denotes a reagent gas internal calibrant, K0 values obtained in air (A) or nitrogen (N))
Full size table

Ambient pressure, positive mode

The reactant ion peak (RIP) has been used as an internal standard, to which other IMS peaks are referenced [11, 12]. An RIP is commonly observed with ionisation sources such as 63Ni and corona discharge, and some researchers have argued that using the RIP is the most convenient and efficient way of calibrating their IMS systems. Such an approach was exemplified by Tabrizchi [11] who proposed using [(H2O)2H]+, as an internal reference standard. However, investigations by Eiceman and co-workers [6] concluded that an (H2O)2H+ RIP was an unsuitable chemical standard, since the reduced mobility of this ion was affected significantly by temperature (as shown in Fig. 1). The presence of ammonia or other drift gas dopants also affects the RIP response and drift time. Further, the landmark studies of Kebarle and co-workers described the equilibrium distributions of hydration states with changes in water concentration as well, making the isolation of (H2O)2H+ difficult [13].

Fig. 1
figure 1

The effect of temperature on the AP-IMS reduced mobilities of the reactant ion peak (RIP, [(H2O)2H]+), 2,4-lutidine ([2,4-DMP.H]+ and proton-bound dimer [(2,4-DMP)2H]+), 2,6-di-tert-butyl pyridine ([2,6-DtBP.H]+ and [(2,6-DtBP)2H]+) and dimethylmethyl phosphonate ([DMMP.H]+ and [(DMMP)2H]+) (reproduced from reference [6])

Full size image

The use of 2,4-dimethyl pyridine (2,4-DMP), commonly referred to as 2,4-lutidine or lutidine, was first proposed as a reference standard in the 1980s [14, 15] and it has since become a widely adopted positive mode standard [16]. Lutidine ions produce a single, sharp IMS peak with a reduced mobility (K0) value of 1.95 cm2 V−1 s−1 (in air [15] and nitrogen [17]). The lutidine dimer peak can also be observed with K0 of 1.43 cm2 V−1 s−1 [6, 18]. The mobility of lutidine has been examined using a range of ionisation techniques, including; 63Ni, electrospray ionisation [19], photoionisation [17] and matrix-assisted laser desorption/ ionisation (MALDI) sources [20], all of which produced comparable IMS responses. In the early investigations of Lubman [17] and Karpas [14], it was believed that the mobility of lutidine was independent of drift cell temperature. However, recent observations indicate that the K0 of lutidine increases with temperature [6], see Fig. 1. Many groups have used lutidine to calibrate their mobility spectrometers and have compiled reference libraries of reduced mobility values for a range of compounds normalised relative to the standard 2,4-DMP (lutidine) [17, 21, 22]. Thomas et al. [23] introduced a novel method of measuring drift gas temperature through the use of 2,4-lutidine to determine the cell constant, and demonstrated reliable calibration of an IMS system. Kanu and Hill [24] recently reported the reduced mobility of lutidine in drift gases of differing polarisability. Consistent with other studies, the reduced mobility of 1.95 ± 0.01 cm2 V−1 s−1 was obtained for the standard 2,4-lutidine in air and nitrogen. However, the K 0 value of lutidine in carbon dioxide and nitrous oxide was determined to be 1.20 ± 0.02 and 1.18 ± 0.01 cm2 V−1 s−1, respectively. This decrease in mobility was observed as a consequence of the effect on the ion collision cross section on increasing the polarisability and size of the drift gas. These findings demonstrate that the mobility of a reference compound cannot be transferred between different chemical regimes.

Eiceman et al. [6] examined the potential of 2,4-lutidine, dimethyl methylphosphonate (DMMP) and 2,6-di-tert-butyl pyridine (2,6-DtBP) as chemical standards for IMS using 63Ni ionisation. Results from this investigation indicated that, of the three compound tested, 2,6-DtBP would be a highly suitable candidate for standardising reduced mobilities, given that its mobility is largely independent of the effects of drift gas temperature, moisture and electric field strength. DtBP has been used successfully as a chemical standard for the calibration of ion mobility spectrometers using both electrospray ionisation [25] and 63Ni sources [6]. DtBP has been reported to generate a well defined IMS peak with a reduced mobility of 1.42 cm2 V−1 s−1, using air [6] and nitrogen [26] as the drift gas. The effect of temperature on the calculated reduced mobilities of DMMP, lutidine and DtBP are shown in Fig. 1 (taken from Eiceman et al. [6]). These data suggest that the proton-bound dimers [(2,4-DMP)2H]+ and [(DMMP)2H]+ may also have potential as IMS reference standards at temperatures below 78°C for lutidine and 180°C for DMMP, the temperature ranges at which these proton-bound dimers are stable. In contrast, an increase in reduced mobility was observed at high temperatures for the protonated [2.4-DMP.H]+ from a K0 of 1.82 to 2.05 cm2 V−1 s −1, within the temperature range of 78 to 250°C, and for [DMMP.H]+, K0 increased from 1.80 to 2.08 cm2 V−1 s−1 at 50 to 250°C (values estimated from Fig. 1).

In 2003, the US Environmental Protection Agency contracted the National Homeland Security Research Centre (NHSRC), as part of the Environmental Technology Verification (ETV) Program [8], to assess the performance of commercial portable ion mobility spectrometers for the detection of toxic industrial chemicals, CWAs and simulants. This verification process involved using dipropylene glycol monomethyl ether (DPGME) [27] and DMMP, to perform operational check tests, under a range of conditions and practices that mimic the in-field and on-site use of these portable IMS instruments. Such verification procedures were carried out by the United Nations Chemical Weapons Convention (CWC), to monitor mobility spectrometers for the detection of nerve agents, in particular alkyl methylphosphonofluoridates and related chemicals. In this case, the proton bound DMMP dimer (a simulant for chemical warfare agents) was used as the reference standard, with a reported K0 value of 1.40 cm2 V−1 s −1. The DMMP dimer was found to be invariant of temperature from 20 to 220°C and water concentrations of up to 2,500 ppm, making this compound ideal to test these ion mobility spectrometers. The effects of electric field and moisture on the reduced mobilities of lutidine, DMMP and DtBP were also explored and found to be negligible [6]. The same three compounds were used as IMS chemical standards to determine the effect of the reagent gas nitric oxide on reduced mobility [28].

The radical cation [M]+· of DtBP was observed when compatibility of DtBP calibrant compound with an atmospheric pressure photoionisation (APPI) source was investigated [29]. The DtBP radical cation, though of lower mass, drifted more slowly in ion mobility drift tube compared to the protonated species, possibly due to a change in spatial conformation. Formation of both types of ions leads to peak broadening and splitting and, therefore, may limit the compound’s compatibility with APPI.

The potential of employing DtBP as a reference compound was further investigated by Viitanen et al. [30, 31]. In this study, the effective drift tube lengths of ion mobility spectrometers were determined using 2,6-DtBP, to enable inter-comparison between IMS devices by adjustment of mobility scales. This method of mobility alignment not only allowed comparison of different instrumentation, with varying drift tube lengths, but also suggested that DtBP would be highly suitable to be used as a reference compound in conjunction with in-field mobility spectrometers.

Another group of promising candidate standardisation compounds for IMS are the tetraalkylammonium halides (TAAH), which have been examined as chemical standards for electrospray ionisation with ion mobility spectrometry by two groups [32, 33]. Viidanoja demonstrated that the reduced mobilities of the C2–C8, C10 and C12 (m/z range 130.0–690.8) TAAH were independent of temperature, drift field and chemical conditions within the drift tube, making them promising candidates as IMS reference standards. An additional feature of TAAH is the low tendency to form cluster ions with ESI solvent molecules and water molecules in the drift gas, due to the steric hindrance effects of the alkyl groups around the quaternary nitrogen. Cluster ions can often complicate IMS spectra, as several peaks may be observed for each compound, whereas with the TAAH a single, sharp IMS peak is generated for each member of the homologous series, covering a broad mass range, see Fig. 2. Other advantages of these compounds include low toxicity, ready availability and low cost.

Fig. 2
figure 2

ESI-IM-MS mass-selected ion mobility responses of aliphatic C2–C8, C10 and C12 tetraalkylammonium salts at m/z 130.0, 186.1, 242.1, 298.4, 354.5, 410.4, 466.5, 578.7 and 690.8 with reduced mobilities of 1.88, 1.56, 1.33, 1.15, 1.02, 0.92, 0.84, 0.73, and 0.67 respectively (reproduced from reference [32])

Full size image

The use of nicotinamide has been reported widely as a positive ion mode internal calibrant in commercially available IMS instruments [7, 34–37]. The selection of the nicotinamide was based primarily on its availability and ease of ionization. The IMS behaviour of nicotinamide is known to be non-ideal, as K 0 value has slight dependence on T, P and drift gas composition. However, the large amount of IMS data accumulated for these compounds enables compensation for K 0 deviations via software algorithms. In this approach, the K 0 values obtained for calibrants are always set at their tabulated values and the measurement scale for the instrument is adjusted accordingly. The obvious drawback of this approach is non-transferability of the calibration data in between instruments from different manufacturers, as the IMS behaviour can be instrument-dependant. Nicotinamide has been employed successfully not only as a reference standard, but also as a drift gas dopant for enhanced detection of analyte ions via proton transfer reactions. Bota and Harrington reported the use of nicotinamide with a reduced mobility of 1.85 cm2 V−1 s−1 [14] for the direct detection and quantification of trimethylamine in meat food products using IMS [34]. Nicotinamide has been used as an internal calibrant for the analysis of drugs of abuse including cocaine, heroin and LSD [7]. Vinopal et al. [38] also employed nicotinamide as an internal reference standard in a study to fingerprint and differentiate bacterial strains.

Wang et al. [39] constructed a solid-phase microextraction/surface enhanced laser desorption/ionisation-ion mobility spectrometer (SPME/SELDI-IMS), which was programmed in the positive mode. Nicotinamide was preloaded into the IMS drift gas as the internal standard for the analysis of drugs and other non-volatile compounds from urine samples. Studies by Keller, Miki et al. [40–42] demonstrated rapid screening and semi-quantitation of illicit drugs (including methamphetamine and ecstasy) directly from human hair and clothing for forensic applications using IMS. In this work, nicotinamide was also introduced into the drift cell as a calibrant and reactant gas for IMS. In addition, dibenzylamine (DBA) (K0 1.407 cm2 V−1 s−1) and trihexylamine (THA) (K0 1.060 cm2 V−1 s−1) were employed as internal standards for these specific applications. A proposed approach for environmental screening for organic contamination described hexaphenylbenzene (HPB) as a mobility reference for the positive ion mobility mode [43].

Ambient pressure, negative mode

A list of calibrants used for negative ion mobility mode is also presented in Table 1. IMS is most commonly employed in negative ion mode for the detection of explosives and CWAs. In contrast to the range of reference standards used for positive IMS mode, few standards exist for negative ion mobility mode.

Methyl salicylate, a CWA simulant is the most frequently used internal calibrant for the detection of negative ions including explosives [9, 36, 44, 45]. In a study by Patchett et al. [36], methyl salicylate was used as an internal calibrant for the negative ion mode detection of the drug-of-abuse, γ-hydroxybutyrate (GHB), with a hexachloroethane dopant. The authors reported the reduced mobility of methyl salicylate to be 1.62 cm2 V−1 s−1 and used this to determine the relative reduced mobilities of GHB (using Eq. 6). Buxton and Harrington [44] determined the reduced mobilities of pentaerythritol tetranitrate (PETN) and cyclotetramethylene tetranitrate (HMX) relative to methyl salicylate in negative ion mode. The K 0 of the methyl salicylate ion was calculated (using Eq. 6) by running a standard sample of trinitrotoluene (TNT), and comparing the experimental reduced mobility to the literature K 0 value for TNT (1.54 cm2 V−1 s−1).

Vinopal et al. [38] examined the potential of IMS in negative ion mode to differentiate between bacterial stains by the direct analysis of whole cells using 4-nitrobenzonitrile as the internal mobility standard. 4-nitrobenzonitrile has also been used with hexachloroethane as the dopant, for the detection of explosives [46], gunpowder stabilisers [47], and to correct for the operating pressure and temperature conditions of an ion mobility spectrometer [35]. In addition, a calibration procedure was also carried out using 4-nitrobenzonitrile as an internal calibrant and TNT as the external calibrant in the negative mode. This two step calibration process enabled K0 for the internal calibrant to be set so that the observed reduced mobility of the external calibrant TNT was adjusted to the defined/known K0 literature value of 1.45 cm2 V−1 s−1. [38].

An alternative reference standard for negative ion mobility mode is iodine, introduced as an internal standard in the sample mixture. Iodine has been used successfully as an internal calibrant for the identification and detection of trace amounts of toxic compounds such as chemical warfare agents, narcotics and explosives [48]. Dioctylphthalate and di(ethyl-hexyl)phthalate) have also been proposed as negative ion mode standards for use in environmental screening applications; alongside hexaphenylbenzene (HPB) in the positive ion mode (see above) [43].

Testing and validating commercial, security and military systems

One of the most valuable attributes of stand-alone IMS devices is their portable and robust nature, a strong reason for their widespread adoption for in-field and on-site operations. The use of reference standards and their associated procedures under such conditions must be compatible with the demanding requirements of the security and military user [1, 18, 24, 49]. The detection of chemical warfare agents, explosives and illicit drugs typically involves demanding and variable operating environments. Simple-to-use and fast procedures with straightforward data interpretation are priority specifications for applications such as these. It is essential that IMS devices are accurate and precise over a range of operating environments and that false positive results are minimised without imparting the risk of delivering false negatives.

Other key aspects that must be implemented in this sector are traceable protocols for verifying claims of enhanced detection capability; increased sensitivity, faster clear-down of contamination, or faster sampling. The adoption of “best-practice guides” for IMS users, with the implementation of protocols on optimising methods would allow cross referencing of results and assure high quality data.

The result of these operational imperatives is that commercial, security and military instrumental applications have led the way in implementing standardisation. For example, nicotinamide and 4-nitrobenzonitrile are embedded internal calibrants in some commercial IMS systems, used in combination with test and verification kits based on “lipsticks” impregnated with analytes, enabling the sampling, thermal desorption, and ion mobility operations to be tested. Chemical agent detectors use hand held “confidence-testers” containing vapour sources of DPGME and methyl salicylate for in field validation. Instrument performance in microfabricated DMS/FAIMS devices is characterised with dimethyl methylphosphonate (DMMP), diisopropylmethylphosphonate (DIMP) and methyl salicylate [9] or by monitoring the RIP to test instrument operation.

Quantitative evaluation

The National Institute of Standards and Technology (NIST), part of the federal agency within the U.S. Department of Commerce [50], are currently developing methods and referencing material standards to calibrate ion mobility spectrometers for the reliable detection of trace-levels of explosives and chemical warfare agents. NIST’s approach uses piezoelectric inkjet technology to generate fieldable standards for evaluating the performance of IMS systems in terms of quantitative accuracy and to assess false positive and false negative detection rates. Piezoelectric inkjet technology is a reproducible method that allows the preparation and printing of a range of explosive calibration standards. These include TNT, cyclotrimethylene trinitramine (also known as RDX) and PETN, as well as plastic explosives C4, semtex and detasheet [51]. These explosives were used to construct calibration curves from IMS responses at varying concentration levels and sampling times.

Concentration relationships and the quantitative evaluation of response curves at ambient pressure have a wider significance than the straightforward characterisation of an analytical response in ambient pressure ion mobility spectrometry. The ion mobility spectrum is the result of interacting kinetic and thermodynamic processes within the reaction region, and ultimately within the drift tube itself. The relationships between concentration, residence time and temperature control the distributions of the cluster ions seen in the mobility spectrum; the exact nature of which depends upon the nature of the reactant ion chemistry. Equation 7 shows the formation of an ion cluster whereby one or more water molecules are displaced from the reactant ion cluster by a molecule of the neutral standard (M).

$$ \left( {{\text{N}}_2 } \right)_{\text{n}} \left( {{\text{H}}_2 {\text{O}}} \right)_{\text{m}} {\text{H}}^{+} + {\text{M}} \rightleftharpoons \left( {{\text{N}}_2 } \right)_{{\left( {\text{n - i}} \right)}} \left( {{\text{H}}_2 {\text{O}}} \right)_{{\left( {\text{m - j}} \right)}} {\text{MH}}^{+} + {\text{i}}\,{\text{N}}_2 + {\text{j}}\,{\text{H}}_2 {\text{O}} $$
(7)

Increasing the concentration of the neutral standard and, or, its residence time in the reaction region will increase the tendency for the protonated monomer ion to go on to form a protonated dimer Eq. 8. Indeed the possibility exists to continue to react to form trimers and potentially larger cluster ions.

$$ \left( {{\text{N}}_2 } \right)_{{\left( {\text{n - i}} \right)}} \left( {{\text{H}}_2 {\text{O}}} \right)_{{\left( {\text{m - j}} \right)}} {\text{MH}}^{+} + {\text{M}} \rightleftharpoons \left( {{\text{N}}_2 } \right)_{{\left( {{\text{n - }}\left( {{\text{i}} + {\text{k}}} \right)} \right)}} \left( {{\text{H}}_2 {\text{O}}} \right)_{{\left( {{\text{m - }}\left( {{\text{j}} + {\text{i}}} \right)} \right)}} {\text{M}}_2 {\text{H}}^{+} + {\text{k}}\,{\text{N2}} + {\text{i}}\,{\text{H}}_2 {\text{O}} $$
(8)

The thermodynamic and kinetic elements underlying homogeneous and heterogeneous proton-bound ion clusters have been described [52, 53] and what emerges from these studies is the sensitivity of the ion mobility spectrum to changes in the critical factors of concentration, temperature and residence time in the reaction region, making the use of concentration relationships in the formation of ion clusters an attractive approach in specifying ion mobility standards.

The generation of calibration/standard curves is employed not only within military and security applications, but is often used for quantitative studies of other analytes. For example standard solutions of caffeine analysed using LC/AP-IMS have been used for performance evaluation [54], the determination of atrazine concentration in natural water [55] and quantification of morphine in mouse plasma. The analytical value of an IMS device in terms of true positive and false positive probabilities have been evaluated using a receiver operator characteristic (ROC) curve approach for determining the detection limit of an IMS chemical sensor [56].

Mobility standards for ion mobility-mass spectrometry

IMS coupled with mass spectrometry (IM-MS) is an area of major growth and has opened up many new applications of IMS, as a result of the complementary separation characteristics of the two techniques. IM-MS enables the identification of unknowns through the determination of the mass-to-charge ratio as well as the collision cross section (size and shape) of an ion. This offers a number of benefits such as simplification of mass spectral data, separation of isomers and conformers, reduction in spectral noise and chemical interferences, charge-state separation and structure elucidation [57–63]. The evolution of IM-MS and applications of the technique have recently been reviewed by Hill et al. [5].

The commercial availability of IM-MS is relatively new and, consequently, there are lack of protocols and standards for the independent validation of these instruments and in-house constructed spectrometers. When using such new technology, it is important that data are validated independently to allow cross examination of results and to demonstrate instrument performance, so that results will stand up to scientific and legal scrutiny. The Waters Synapt™ high definition (travelling-wave) ion mobility-quadrupole-time-of-flight mass spectrometer [64] and the ThermoFinnigan FAIMS interface for the ion trap and triple quadrupole mass spectrometers, based on the concentric cylinder design by Guevremont et al. [65–67], are examples of IM-MS that have recently been commercialised. The FAIMS interface is used to provide an extra degree of specificity, by allowing transmission of ions of selected differential mobility only, prior to MS and tandem mass spectrometry (MS/MS). The ion selection is based on differences in the mobility of an ion under high and low field conditions. A constant DC voltage, called the compensation voltage or CV, of the correct magnitude and polarity needs to be applied to compensate for the ion displacement resulting from the difference of K h (K during the high-field cycle) and K l (K during the low-field cycle), in order to enable the ions to travel between the DMS/FAIMS electrodes and be transmitted to the detection device. Each ion will have its characteristic CV under a certain set of conditions. At present, performance evaluation of the FAIMS-MS instrumentation is carried out using a range of compounds (for example; 1, 3, 6 polytyrosine mix, polyethylene glycols (PEG) and reserpine for positive ion mode, and taurocholic acid in the negative ion mobility mode), selected mainly to establish whether the instrument’s performance/compensation voltages (CV) are acceptable. The usual testing procedure involves spiking the calibrant into the device and measuring its CV value; if no response or a reduced or shifted response is observed (calibrants are normally expected to exhibit CV values within ± 1 V of their target range), this is indicative of a problem in instrument performance. The FAIMS-MS instrument may also be tested using the mass spectrometry calibration mixture (tuning solution) consisting of ultramark, caffeine and the MRFA peptide for day-to-day checks and to test the performance of the spectrometer.

The nature of the Waters Synapt™ Tri-wave IM-MS instrument necessitates its calibration, as there is no ab initio theoretical description which correlates drift time in the travelling wave field to the ion’s size or other parameters. The instrument is calibrated either by introduction of caesium sodium iodide [68] to ensure mass accuracy, or via correlation with libraries of ion mobility data and literature standards. The latter approach seems to be the preferred method for calibration, most commonly using species of proteomic origin [69–71]. The collision cross sections and ion mobilities data of a wide range of proteins and peptides are well documented, and therefore can be used to assess the instrument performance by comparison of known values with experimental measurements [70, 72, 73]. So far, mobility data obtained using the Synapt IM-MS instrument has shown good correlation with data obtained from other mobility separators for a mixture of peptides standards (containing bradykinin, substance P, bombesin, LHRH, and angiotensin I and II), tryptic peptides (from haemoglobin, horse albumin etc.) and other large protein and biomolecular complexes including cytochrome c, lysozyme, α-lactalbumin and myoglobin [64]. Calibration plots of collision cross section measurements and drift times of tryptic peptides and proteins have also been constructed to estimate collision cross sections of compounds of interest, for example the collision cross section of PEG was estimated using calibration plots of published cross sections of tryptic peptide ions obtained from human haemoglobin [73].

The proteins cytochrome c, ubiquitin and lysozyme enzyme have been proposed as mobility and collision cross section standards [64, 72, 74]. Lysozyme enzyme is a promising candidate as the presence of di-sulfide bonds enhances the stability of the structure, making lysozyme less susceptible to conformational changes. Several conformations of ubiquitin exist: compact, partially unfolded and unfolded, all of which have been well studied [75, 76]. The conformation of ubiquitin can be controlled under certain solvent compositions, and it is known that ubiquitin in the +13 charge state is the most stable conformation, generating a single, resolved IMS peak.

Other examples of candidate compounds that have been identified as suitable for use as collisional cross section standards are caesium iodide and sodiated PEG. The sodiated PEG complexes have stable collision cross sections, particularly in the nonamer conformation [73, 77]; which adopts a ball-like structure with the metal ion encapsulated in the centre. Such compounds experience minimal interactions with the drift gas, as the carbon and oxygen atoms are present in the inner structure of the molecules.

For peptide and protein analysis, C60 and C70 fullerenes have also been used as internal standards in conjunction with MALDI-ion mobility-time-of-flight mass spectrometry (MALDI-IM-TOF-MS). C60/C70 were co-deposited with the peptide mixture onto the MALDI target plate [4, 78, 79] for ion mobility separation prior to mass analysis. The fullerenes displayed different mass-mobility distributions to that observed for peptide ions, and the drift times of the compact fullerenes were lower than peptide ions of the same mass, as indicated by Fig. 3. Peptide ions could therefore be easily differentiated from the calibrant responses. In addition, due to the rigid and stable structural characteristics of fullerenes, these molecules possess constant collision cross sections which tend to be independent of factors such as temperature [77].

Fig. 3
figure 3

Plot of drift time vs. m/z for six peptides and the fullerene molecules C60 (m/z = 720.6) and C70 (m/z = 840.7), which were used as internal calibrants for MALDI-IM-TOF MS analysis (modified from reference [78])

Full size image

Sysoev et al. [80] used a standard test mixture consisting of lutidine, 2,6-DtBP and tetrabutylammonium iodide, to test the potential of electrospray ionisation coupled to IM-MS incorporating an atmospheric pressure drift tube. The instrument was used for the rapid screening of complex samples, primarily for drug discovery and environmental screening.

IMS spectral libraries

Spectral libraries of ion mobility data have often been used as references to calibrate ion mobility spectrometers. A number of groups have established libraries of drift times, collision cross sections, and reduced mobilities for selected analytes under specified conditions, all of which are accessible to IMS users. Hill et al. [15] presented tables of reduced ion mobilities for compounds reported during the period 1970–1985 using ambient pressure IMS. Additional information, such as K0 of product ions, the carrier and drift gases used in each case and the temperature of the drift tube were also defined in this review. Karpas [14] compiled the reduced mobilities (relative to the reference standard lutidine) of aliphatic and aromatic amines (molecular weight range 32–521) in air, at 150°C, 200°C and 250°C. More recently, Eiceman and Karpas [22] provided an extensive spectral library (supplied on a CD-ROM with reference [22]) of drift times and reduced mobilities for a range of compounds, including alcohols, amines, organo-phosphates and pyridines. In some cases, the reduced mobilities of these compounds were calculated relative to 2,4-lutidine. In a study by Matz et al. [81], reduced mobilities were tabulated for cocaine and its metabolites, amphetamines, benzodiazepines, and small peptides in four different drift gases (helium, argon, nitrogen and carbon dioxide). The reduced mobilities of a range of compounds including lutidine, crown ethers, amino acids and peptides were reported by Bramwell et al. using nano-ESI, and reduced mobilities compared with values obtained by APCI (63Ni) [19]. In order to facilitate the creation of libraries of IMS spectral data, the IUPAC JCAMP-DX electronic data exchange format was adapted [82, 83]. This was developed to allow exchange of IMS spectral data, inter-comparison between IMS users, laboratories and systems, as well as validation of new ion mobility spectrometers. The data exchange software can be downloaded from the IUPAC JCAMP-DX website [84] or via the Institute for Analytical Sciences (ISAS) [85] website.

In the field of biomolecule analysis, a database of collision cross section measurements (in helium) of proteins and peptides, including ubiquitin [86], cytochrome c [71], polyaminoacids, and tryptic peptides from common proteins such as haemoglobin and bovine serum albumin are available on the Indiana University (Clemmer group) website [69, 70]. A database of cross sections and reduced mobilities for a range of singly-charged proteolytic peptide ions is also available [87, 88]. It is common practice to express mobility data of proteins and peptides as collision cross section measurements, as proteomic analysis using IMS often involves structure elucidation and conformation studies. Purves et al. compared collision cross section measurements of bovine ubiquitin obtained in nitrogen using the FAIMS instrumentation [76], with literature values in helium reported by Clemmer [89] using drift tube-IMS as a means of instrument calibration. These results indicated a good correlation between collision cross section measurements, as similar conformation and charge states were observed on both IMS systems. This work also demonstrated the potential of using ubiquitin (particularly in the higher charge state) as a collision cross section standard for protein measurements.

Bowers and Jarrold have developed computer programs such as the Sigma program [77, 90, 91], which employs the projection approximation algorithm, and Mobcal [92, 93], based on trajectory and exact hard sphere scattering calculations. These programs are available to determine mobilities and theoretical collision cross sections of ions, thus allowing comparison between experimental cross section measurements and modelling for structure elucidation.

Selection of ion mobility standards

UK national measurement system initiative in ion mobility reference materials

The Chemical and Biological Metrology programme, part of the UK National Measurement System (NMS), aims to provide the necessary metrology infrastructure for the establishment of validated protocols and standards for current and emerging uses of IMS [94]. The first objective of this initiative is the investigation of the feasibility of producing a set of IMS reference materials and protocols for an extended application base to allow harmonisation and traceability of IMS data. Currently, IMS users and manufacturers are being consulted about their requirements for standards. Further studies are also investigating possible causes of interferences in the use of IMS standards, with the intent of improving confidence in analyte identification and the reliability of IMS results. Candidate compounds (of m/z range 102–105) for use in test and calibration mixes to enable evaluation of instrumental performance and mobility and cross-section determination, as well as quantitation accuracy, are being developed.

Specifications for ion mobility reference materials

Ion mobility measurements are affected by experimental conditions (for example ionic charge, drift gas composition, temperature, pressure and humidity), and instrumentation performance (factors include electric field stability and homogeneity, timing accuracy and digital signal processing). Therefore, standards are required to enable nominally fixed instrumentation factors to be isolated from variable experimental factors for instrumental performance evaluation and the determination of reduced mobility/collision cross section. Mobility reference materials for assessing instrumentation should be insensitive to changes in experimental parameters, while complementary standards that are highly sensitive to changes in experimental conditions enable the operational parameters and state of a system to be assessed. The selection criteria for candidate ion mobility reference standards encompass five criteria (some of these specifications are in opposition to each other); therefore the standards should possess the following characteristics:

Stability

The mobility of ions should ideally be resilient to conditions within the drift cell such as, temperature, pressure, electric field strength, and should be reproducible in a range of drift gas compositions, including increasing polarisability (He, N2, CO2, SF6). Special attention is needed for hyphenated systems where the effect of injection energy of ions into a low-pressure drift cell needs to be characterised, as the formation of excessively energetic, metastable ions would lead to fragmentation of clusters, resulting in changes in the collision cross section and observed mobility.

Concentration

The concentrations of reference standards need to be specified carefully to ensure that monomer dimer, and potentially other cluster ion relationships, are defined and exploited. The information on the mobility of the monomer and dimer ions are augmented with data on their relative intensities which provides useful information on temperature and gas flows within the instrument.

Sensitivity to change

Responsive to changes in temperature, pressure, electrical field strength and drift gas in a defined and reproducible way. The levels of water in the drift gas may also influence drift times and differential mobility behaviour. Increase in humidity levels of the drift gas can result in changes in the ion chemistry (via cluster reactions between waters molecules and sample ions, and hydrolysis), which can often lead to the formation of additional ion species that may complicate the spectra. IMS peak broadening effects and shifts in reduced ion mobilities are also frequently observed [95, 96]. This is an especially important factor for portable IMS systems that are used in-field.

Practicality

Inexpensive, readily available in sufficient purity, with low toxicity, and stable under normal storage conditions for extended periods of time.

Relevance to the target analyte

Reference standards should reflect the chemical nature of the target analyte. The standard’s characteristics should be comparable to the analyte’s volatility, polarity, mass and concentration.

Ion formation

Candidate reference standards should form well defined ion species with known reduced mobility and m/z (using IM-MS) values, and should display consistent and stable IMS responses with little variation between users and instruments.

A number of ionisation techniques have been employed for sample introduction to mobility spectrometers depending on the nature of the analyte of interest. Compounds can be divided into categories according to whether they are sufficiently volatile (generally small molecules, including explosives and CWAs) to be ionised using 63Ni, corona discharge and photoionisation, or those that are amenable to electrospray and matrix-assisted laser desorption/ionisation (proteins and peptides for example). Therefore, source compatibility of potential candidates and parameters affecting ionisation (such as cone voltage, gas flow rate, temperature etc.) also need to be addressed. In addition, reference standards need to be developed for use in both positive and negative ion mode.

Preliminary findings

There is no single compound that meets these criteria for all application areas, IMS systems and ionisation techniques. Consequently groups of candidate reference standards need to be assessed over a range of IMS systems (DMS, FAIMS, linear-IMS and low-pressure IMS), and in the presence of various reagent gases and modifiers with the intent of selecting a calibration mixture, or a set of reference compounds suitable for use as universal IMS standards.

Candidate standards may be categorised into two main groups: instrumental/performance evaluation reference materials and standards for the determination of reduced mobility/collision cross section. The K0 and Ω of the latter type of standards should be well defined, invariant of experimental parameters and insensitive to gas impurities and moisture content within the drift cell, the former should behave in a predictable way to changes in experimental/instrumental conditions.

As indicated earlier, lutidine, DtBP and the tetraalkylammonuim halides have already been proposed as IMS calibrants and appear to be highly suitable candidates as instrumental and K0/Ω standards for low molecular weight analytes. The growing interest in the use of IM-MS for proteomic analysis means that standards of high molecular weight and biological origin are also required. Lysozyme enzyme and ubiquitin in the +13 charge are attractive candidates, whilst Group I metal PEG complexes merit further scrutiny. Similarly, the fullerenes are very promising candidates due to their rigid (stable) structural properties. As discussed earlier, the fullerenes have already been employed as an internal calibrant (C60 and C70) for peptide/protein analysis.

An important additional consideration for mobility standards for proteomic analysis is the interaction between drift gases and the conformation of the ionised biopolymer. Such relationships have yet to be completely described and require further study. Furthermore, the distribution of charge and conformational/folding states of protein standards may be influenced by the conditions in the ionisation source, and the effects of pH and solvent compositions have yet to be comprehensively characterised. Higher charge states can be observed for some proteins in acidic conditions, as the proteins become denatured and more basic protonation sites become available [71, 97]. Structures can be stabilised by bound metal ions, so that the collision cross section remains constant and resists changes in conformation as a result of factors such as temperature, pressure and electric field.

In addition to providing a set of IMS reference standards, the production of best practice guides on the experimental design and validation of protocols for ion mobility systems would greatly benefit the community. The use of IMS data mining software for peak clean up, measurement of peak shape (apex determination) and comparison of drift time to a reference standard would enhance the quality of ion mobility data. Such data processing is particularly important when combining very large data sets (for example in environmental and clinical analysis). The development of current databases and spectral libraries of mobility, m/z, and collision cross section measurements (under a given set of instrumental and environmental conditions) for a range of compounds relevant to different application areas would also facilitate calibration and evaluation of mobility spectrometers.

Conclusions

The range of IMS applications is increasing beyond established military and security in-field monitoring roles. For example, the development of new methods and applications of IMS in bioanalytical and pharmaceutical science, the technique seems certain to continue to expand outside its established user base. Although the use of several IMS reference standards have been reported for a range of IMS and IM-MS systems, including 2,4-lutidine, DtBP, nicotinamide, and TAAHs for positive ion mode and 4-nitrobenzonitrile and methyl salicylate in the negative ion mobility mode, the current status of standardisation for the wider application base is at a low level and indeed there are no internationally accepted reference standards for the calibration and evaluation of mobility spectrometers. The UK NMS Chemical and Biological Metrology Programme is anticipating and addressing the need for the user community to adopt the use of traceable standards to expedite the research and development activity in ion mobility spectrometry.

This review outlines the current status and benefits of standardisation in IMS and identifies issues/challenges that would need to be addressed in the future. The key requirements/characteristics, in terms of instrumental parameters and environmental conditions (such as temperature, electric field, humidity and drift gas composition) that an ‘ideal’ reference standard must possess to be used as a universal standard, have been discussed. There is also a need to adopt standard testing protocols to facilitate data comparison between different IMS systems and users. It is important to continue to compile databases and spectral libraries of mass, mobility and collision cross sections for a range of compounds (under various instrumental and environmental conditions), and should be accessible by IMS users and manufacturers. The ultimate goal of the NMS programme is to foster the adoption of widely recognised mobility standards across the IMS research and user community enabling more comparable measurements on a global scale.