Extractionless Analyses By Direct Analysis In Real Time (Dart) Mass Spectrometry
by Teris A. van Beek, Yao Shen and Radostina Manova. Laboratory of Organic Chemistry, Surface-bound Analytical Chemistry Group, Wageningen University, Dreijenplein 8, 6703 HB Wageningen, The Netherlands
Chemical analysis typically is a tedious process encompassing multiple steps. A flow chart of an analysis involving chromatography is depicted in Fig. 1. Of all these steps, sample preparation is often the most difficult and time-consuming step taking on average 60% of the total analysis time. Due to all the manual actions, it is also the most error-prone step and analytes can be lost in this step. Thus, if possible, the best sample preparation step is no sample preparation. To a lesser extent the same holds true for extraction. If a sample surface could be probed directly, one could forego an extraction step and save on solvents and time.
In 2004 Cooks et al. introduced Desorption Electrospray Ionisation (DESI) . In this technique, a surface is bombarded at high speed by charged solvent droplets, which free analyte molecules. Like in electrospray ionisation (ESI), charge is transferred from the solvent to analytes and charged analytes molecules are transferred into a mass spectrometer for analysis. This enabled the analysis of many surfaces without extraction or sample preparation and started the era of “ambient mass spectrometry”. Shortly thereafter, the DART (Direct Analysis in Real Time) technology was introduced by Cody et al. . Recently Monge et al. proposed four basic characteristics that should be present to be called ambient MS. These are: (1) ionisation in the absence of enclosures (i.e. open air operation); (2) no sample preparation (i.e. direct analysis); (3) direct interfacing to most “atmospheric pressure” mass spectrometers; (4) soft ion generation, with amounts of internal energy deposited equal or lower than those in their atmospheric pressure counterparts .
Since 2004, based on the ionisation mechanism more than 40 types of ambient MS have been published  but only a few are commercially available (Table 1). Roughly they can be grouped into ESI-based (e.g. DESI and LESA) and APCI-based ionisation (e.g. DART and ASAP), with ESI ionisation taking place in a solvent and APCI (atmospheric pressure chemical ionisation) in the gas phase. LAESI is a two-step technique in which analytes are released from the surface by an IR laser pulse. The released molecules are taken up in an ESI spray where ionisation takes place, followed by mass spectral analysis . All ambient MS methods have in common that they are fast and relatively simple. Quantitative analyses have been reported but they are better suited for present/absent analyses and fingerprinting of which there are many in the fields of forensics, doping control, homeland security, food safety and metabolomics.
Figure 1: Different steps in a chromatographic analysis.
Table 1. Commercially available ambient MS interfaces
Fig. 2. Upper: Schematic presentation of DART ionisation source. Lower: DART sampling zone with leaf fragment being probed; on the left DART source, on the right entrance of MS.
Of course ambient MS has its own share of constraints. In principle any released and dissolved molecule, which is ionised is measured. Thus DESI, DART or any other ambient MS technique is a dirty sampling method requiring additional selectivity. This is most conveniently and generally achieved by using either a high resolution (HRMS) or an MS/MS (MSn) instrument. By their nature they are less suitable for the probing of the interior of samples and as the release and ionisation is somewhat instable, ambient MS is less suitable for quantitative analysis. In the following we will illustrate the power of ambient MS by two DART applications.
With DART (Fig. 2) the sample is held in an open air (ambient) sampling zone where it interacts with excited helium species or products formed thereof (e.g. protonated water clusters or reactive oxygen species). Suitable molecules, i.e., not too polar and MW < 1000, on the surface are ionised, enter the MS and are measured. Measurements in both (+) and (-)-mode are possible and the gas temperature can be set from room temperature up to 550 °C. Generally less volatile substances require higher temperatures although often there is an optimum above which decomposition can occur.
Chinese star anise fruits (Illicium verum ) are a frequently used spice in Asian cuisine. In other parts of the world teas are used. The morphologically similar Japanese star anise (Illicium anisatum ) is only used for decoration purposes because it contains the neurotoxin anisatin (Fig. 3). Ingestion leads to nausea, hallucinations, and epileptic seizures. Intoxications of babies as well a large outbreak of toxicity among adults in The Netherlands have been reported [5,6]. Thus it is eminent that the two species can be unambiguously distinguished. However currently there are no reliable simple methods and the only excellent and sensitive HPLC-triple quad MS method requires extensive sample preparation (total of 13 steps) and is thus slow . As essentially only a qualitative distinction is required (anisatin is yes/no present?), ambient MS in the form of DART seemed a possible option.
Fig. 3. Chemicalstructure of anisatin (C15H20O8, MW=328)
When Japanese anisatin was interrogated by DART in negative mode at 400 °C, within seconds and without any sample preparation, a strong signal for deprotonated anisatin [M-H]- was seen at m/z 327.107 (Fig. 4). When Chinese star anise was analysed, no signal or a very weak signal was observed. Both mass spectra contained in the region m/z 327.0-327.3 an additional peak at m/z 327.072 with as most likely elemental composition C14H16O9. This could be bergenin, a fairly common plant metabolite. It stresses that additional selectivity is absolutely necessary. The results in ()-mode could be confirmed in (+)-mode (Fig. 5). In this case not [M+H]+ but rather [M+NH4]+ at m/z 346.150 (C15H24NO8) was obtained for anisatin. Ammonium adducts are frequently seen in (+)-DART-MS. Again a similar “digital” distinction was observed. Thus it is possible to classify any intact star anise sample by DART-HRMS as toxic Japanese or beneficial Chinese based on the toxin content. Retail teas containing finely ground ingredients could also be distinguished. However in that case first a tea was prepared and subsequently 2 µL of tea was placed on a glass rod and held in the DART sampling zone. By internal normalisation, even semi-quantitative assays proved possible and 1% of adulteration could be clearly detected . A comparison of the existing LC-MS procedure and the DART-HRMS approach with several figures of merit can be found in Table 2.
Fig. 4. Negative mode DART-orbitrap mass spectra from m/z 327.00-327.35 of one carpel of Japanese star anise (A) and Chinese star anise (B). Vertical scale in (B) is expanded 4 times. Peak at m/z 327.107 is anisatin [M-H]- (C15H19O8) .
Fig. 5. Positive mode DART-orbitrap mass spectra from m/z 346.00-346.35 of one carpel of Japanese star anise (A) and Chinese star anise (B). Peak at m/z 346.148 is anisatin, [M+NH4]+ (C15H24NO8) .
Table 2. Comparison of HPLC-MS/MS  and DART-HRMS method  for star anise analysis.
Fig. 6. DART set-up for analysing monolayers. Left: DART source with the exit pointing at the moving monolayer specimen. Right: entrance of mass spectrometer (white ceramic tube).
Fig. 7. Right: chronograms (scans) of NHS (3 , N-hydroxysuccinimide, top left) monolayer on silicon nitride (Si3N4, bottom left). The sample was passed between the DART and MS inlet thrice at 4.1-4.8, 7.0-7.8 and 21.3-22.0 min. Top right: total ion current (TIC) in (-)-mode; middle right: extracted ion current (EIC) of m/z 114.0184 [M-H]-; bottom right: EIC of m/z 229.0460 [2M-H]- (M = NHS).
An entirely different application of DART-HRMS lies in the field of organic monolayer analysis. Monolayers (1-3 nm thickness) on surfaces are of prime importance for sensing and antifouling purposes. They can be analysed by methods such as water contact angle measurements, ellipsometry, infrared reflection absorption-spectroscopy (IRRAS), and X-ray photoelectron spectroscopy (XPS). However none of these methods provides detailed molecular information. Thus it was investigated whether DART-MS is capable of releasing characteristic fragments of molecules making up the monolayer, providing in this way a fingerprint of the originally deposited monolayer. To this purpose the outlet of the DART was directed at a 45° angle at a few mm of the surface placed on a motorised rail with the inlet of the MS also a few mm from the surface opposite of the DART source (Fig. 6)
Activated esters constitute an important group of monolayers as they can serve as a stepping stone for making a wide variety of other monolayers, e.g. mostly by an efficient reaction with amines. In Fig. 7 the DART-HRMS scan of an NHS (N-hydroxysuccinimide ester) monolayer on silicon nitride is depicted. The modified part is scanned thrice and is already recognisable in the total ion current (TIC) profile but more characteristic are the ions related to the released NHS annion and the [2M-H]- anion. These are formed after in situ hydrolysis of the ester in the DART sampling zone, followed by ionisation by loss of a proton from the alcohol group. Other esters, like pentafluorophenyl ester behave similarly in negative mode.
The amides, which are formed after reaction of the above esters with amines, can also be analysed. Again the mechanism is in situ hydrolysis but in this case an amine is released, which is subsequently protonated and can then be measured in positive mode. Aromatic amines (benzylamines) also yield a fragment by loss of ammonia (formation of stable benzyl carbocation). An example is shown in Fig. 8. Here a mixed monolayer of 3 different amides was probed. For each of the three amides, the corresponding protonated amine as well as the characteristic fragment formed by loss of ammonia could be observed. This result cannot be achieved by any other currently existing analytical technique. Further studies have shown that DART-MS is also capable of analysing non-hydrolysable monolayers such as oligoethylene or polyethylene glycol (PEG) monolayers, thioether monolayers and sugar-containing monolayers . DART-HRMS combines speed, sensitivity in the pmol range, sufficient reproducibility and simplicity. A drawback is that it is unsuitable for the analysis of alkane monolayers. All in all one can conclude that DART-HRMS is a new and useful complimentary tool for the analysis of organic monolayers.
Fig. 8. Right: chronograms (scans) of mixed amide monolayer showing the 3 different released amines as well as the corresponding fragments formed after loss of ammonia (TIC and 6 EIC traces). Top left: structures of released amines 7 , 8 and 9 . Bottom left: structure of mixed monolayer on silicon nitride (Si3N4) where R corresponds with fluoro, chloro and bromobenzyl (= 7 , 8 and 9 minus the NH2 group) respectively.
 Z. Takáts, J.M. Wiseman, B. Gologan, R.G. Cooks, Science 306 (2004) 471.
 R.B. Cody, J.A. Laramée, H.D. Durst, Anal. Chem. 77 (2005) 2297.
 M.E. Monge, G.A. Harris, P. Dwivedi, F.M. Fernandez, Chem. Rev. 113 (2013) 2269.
 P. Nemes, A. Vertes, Anal. Chem. 79 (2007) 8098.
 D. Ize-Ludlow, S. Ragone, I.S. Bruck, J.N. Bernstein, M. Duchowny, B.M. Garcia Penã, Pediatrics 114 (2004) e653.
 E.S.D. Johanns, L.E. van der Kolk, H.M.A. van Gemert, A.E.J. Sijben, P.W.J. Peters, I. de Vries, Ned. Tijdschr. Geneeskd. 146 (2002) 813.
 I. Lederer, G. Schulzki, J. Gross, J.-P. Steffen, J. Agric. Food Chem. 54 (2006) 1970.
 Y. Shen, T.A. van Beek, F.W. Claassen, H. Zuilhof, B. Chen, M.W.F. Nielen, J. Chromatogr. A 1259 (2012) 179.
 R.K. Manova, F.W. Claassen, M.W.F. Nielen, H. Zuilhof, T.A. van Beek, Chem. Comm. 49 (2013) 922.
 R.K. Manova, S. Joshi, A. Debrassi, N.S. Bhairamadgi, E. Roeven, J. Gagnon, M.N. Tahir, F.W. Claassen, L.M.W. Scheres, T. Wennekes, K. Schroën, T.A. van Beek, H. Zuilhof and M.W.F. Nielen, Ambient surface analysis of organic monolayers using direct analysis in real time orbitrap mass spectrometry, Anal. Chem., 86 (2014) 2403.