Anal. Chem. 2007, 79, 5364-5372
Advanced Liquid Chromatography-Mass Spectrometry Interface Based on Electron Ionization A. Cappiello,* G. Famiglini, E. Pierini, P. Palma, and H. Trufelli
Istituto di Scienze Chimiche “F. Bruner”, Universita` di Urbino “Carlo Bo”, 61029 Urbino, Italy
Major progress in interfacing liquid chromatography and electron ionization mass spectrometry is presented. The minimalism of the first prototype, called the Direct-EI interface, has been widely refined, improved, and applied to modern instrumentation. The simple interfacing principle is based on the straight connection between a nanoHPLC system and a mass spectrometer equipped with an EI source forming a solid and reliable unicum resembling the immediacy and straightforwardness of GC/ MS. The interface shows a superior performance in the analysis of small-medium molecular weight compounds, especially when compared to its predecessors, and a unique trait that excels particularly in the following aspects: (1) It delivers high-quality, fully library matchable mass spectra of most sub-1 kDa molecules amenable by HPLC. (2) It is a chemical ionization free interface (unless operated intentionally) with accurate reproduction of the expected isotope ion abundances. (3) Response is never influenced by matrix components in the sample or in the mobile phase (nonvolatile salts are also well accepted). A deep evaluation of these aspects is presented and discussed in detail. Other characteristics of the interface performance such as limits of detections, range of linear response, and intra- and interday signal stability were also considered. The usefulness of the interface has been tested in a few real-world applications where matrix components played a detrimental role with other LC/MS techniques. When liquid chromatography-mass spectrometry (LC/MS) is considered, electrospray ionization (ESI) automatically comes into mind. It accounts for almost 90% of all modern analytical methods based on LC/MS. A few other options, such as atmospheric pressure chemical ionization or much less frequently atmospheric pressure photoionization complete the picture, bringing the total to a perfect 100%. The revolution brought by the development of “soft” ionization techniques moved liquid chromatography closer to mass spectrometry, allowing the development of powerful instrumentations for new challenging analytical applications. However, although LC/MS opened the door to large and thermally labile molecules in complex matrixes, its practical application is still far from the immediacy * Corresponding author. Tel. +390722303344. E-mail:
[email protected].
5364 Analytical Chemistry, Vol. 79, No. 14, July 15, 2007
Figure 1. Scheme of the interface and mechanism of ionization.
and simplicity of GC/MS. While the evolution of coupling GC and MS has reached its full maturity bringing a well established, relatively simple, and inexpensive instrumentation, the LC/MS interfacing progress, far more challenging than GC/MS, has far to complete its growth. The reason for the complexity of LC/MS is in the fact that the domains of LC and MS are often seen as antagonists. Components that improve chromatographic separations can interfere with the analyte ionization or contrast with other vital function of the interfacing mechanism. In the effort of overcoming a series of evident difficulties, LC/MS interfaces use different strategies and because of that they are usually expensive and complicated compared to the intrinsic simplicity of GC/MS. Difficulty in LC/MS can be summarized as follows: (1) Solvent restriction. An HPLC mobile phase is a liquid of variable composition, and sample components are led to the detector together with a given solvent combination. Separation and ionization of analytes often depend on different and, sometimes, antagonist mobile-phase components. (2) Sample restriction. Liquid chromatography analyzes all samples that can be dissolved in solvents. The analytes may vary dramatically in weight, polarity and stability and this poses a huge variability in terms of response and system requirements on both interface and mass spectrometer. In spite of this challenging premise, the demand for new LC/ MS methods is increasingly strong in many fields of research. For its original approach and its worldwide success, electrospray ionization is considered a revolutionary progress in LC/MS. However, it is not a flawless interface. Its strict dependence on the mobile-phase composition sometimes limits chromatographic flexibility and introduces some important analytical shortcomings in quantitative analysis such as signal suppression, nonlinear 10.1021/ac070468l CCC: $37.00
© 2007 American Chemical Society Published on Web 06/15/2007
Figure 2. (a) Experimental LC/MS spectrum of 100 pg of atrazine recorded in FIA at a flow rate of 300 nl/min, mobile phase water/acetonitrile 1:1 v/v. NIST library search is reported on the bottom of the figure. (b) Experimental GC/MS spectrum of 100 pg of atrazine. NIST library search is reported on the bottom of the figure.
response, or adduct ion formation that are often found in realworld applications. In addition, the typical low fragmentation limits
compound characterization and the use of reliable libraries so that MSMS is routinely operated for most applications. Analytical Chemistry, Vol. 79, No. 14, July 15, 2007
5365
Table 1. Isotopic Cluster Ion Abundances (%) at Different Mobile Phase Flow Rates theoretical
flow 0.1 µL/min
flow 0.2 µL/min
flow 0.3 µL/min
compound
M
M+1
M+2
M+3
M
M+1
M+2
M+3
M
M+1
M+2
M+3
M
M+1
M+2
M+3
caffeine dimethyl phthalate cyanazine atrazine terbutryn
100 100
10 11
0.9 1.4
0.06 0.1
100 100
10 13
0.9 2
0.1 0.1
100 100
11 11
1 2
0.2 0.1
100 100
12 11
1 0.9
0.3 0.1
100 100 100
12 11 14
33% 33 5.4
0. 3.4 0.6
100 100 100
15 11 15
33 33 5.3
nd 3.6 0.2
100 100 100
16 11 15
30 3% 5.9
nd 2.9 0.4
100 100 100
17 17 12
33 35 6.2
nd 3.0 0.5
theoretical
flow 0.1 µL/min
compound
M
M+2
M+4
M+6
M+8
M
M+2
M+4
M+6
M+8
lindane
52
10
80
34
8
51
100
79
33
8
Table 2. Performance Evaluation during Continuous Buffer Introduction sample
1
2
3
4
5
6
7
8
9
mean
SIM (peak area) scan (NIST probability value)
410 028 94.4
460 755 94.7
422 274 94.6
427 099 96.0
477 094 94.2
439 711 95.3
468 662 94.2
495 140 95.3
470 016 94.7
455 278 94.8
Electron ionization, which pioneered the LC/MS development, is still the rule in GC/MS with no rivals on the scene. Surprisingly, the initial success of EI in coupling liquid chromatography and mass spectrometry, mainly offered by the particle beam interface, was too rapidly disregarded and confined to a very limited market niche.1 As a matter of fact, the development of new, soft, liquidbased ionization techniques polarized most researchers’ attention and caused the rapid disappearance of almost any EI-based interface from the market, although a considerable number of small-medium molecular weight substances can provide very good EI spectra, whether they are in a liquid phase or not. The simple operating principle of EI shows several advantages over other ionization techniques: in EI, the absence of ion-molecule or ion-ion reactions minimizes the influence of the mobile phase in the ionization process and reduces all those weaknesses that can be related to the interface such as signal suppression, influence of the matrix, need for desalting, or postcolumn solvent modifications. Moreover, the typical EI spectrum is very informative, and its high reproducibility allows an easy comparison with thousands of spectra from commercially available electronic resources (such as NIST or Wiley). EI detection can also benefit from recently developed sophisticated algorithms such as the one developed by the National Institute of Standards and Technology (NIST) and AMDIS (Automated MS Deconvolution and Identification System), which extracts the analytes’ mass spectra in complex chromatographic mixtures in the presence of several overlapping peaks. The fact that EI and LC operating conditions remain distant cannot be ignored, and it might explain some of the frustration observed in the past: poor sensitivity, signal instability, reduced linearity, and disappointing optimization. However, under specific circumstances, coupling EI MS with LC can be highly rewarding. Amirav and Granot exploited the potential of supersonic molecular beam in producing intense molecular ion in EI spectra from LC separations.2-4 Enhanced molecular ion is provided together with the library-searchable fragments for improved confidence level 5366
Analytical Chemistry, Vol. 79, No. 14, July 15, 2007
RSD, % 6.6 0.63
in the identification. Our research group has devoted a considerable effort to update and improve the design and the performance of the particle beam interface.5 A significant improved sensitivity for any mobile-phase composition was achieved by drastically reducing the mobile-phase intake. The microparticle beam interface and its evolution, called capillary EI (CapEI, commercialized by Waters, Milford, MA, in 1999), use microscale flow rates (1-5 µL/min) and were successfully employed in the analysis of several classes of compounds and with different modifiers added to the mobile phase.6-15 It is noteworthy to point out that this approach showed, for the first time, a significantly improved tolerance for nonvolatile buffers in mass spectrometry, as demonstrated in two applications.16,17 (1) Chapman, J. M.; Knoy, C.; Kindscher, K.; Brown, R. C. D; Niemann, S. Poster presented to Kansas City Life Sciences Day, 2006. (2) Granot, O.; Amirav, A. Int. J. Mass Spectrom. 2005, 244 (1), 15-28. (3) Amirav, A.; Granot, O. J. Am. Soc. Mass Spectrom. 2000, 11 (6), 587-591. (4) Granot, O.; Amirav, A. In Advances in LC/MS Instrumentation; Cappiello A., Ed.; J. Chromatography Library 72; Elsevier: Dordrecht, The Netherlands, 2007; pp 45-63. (5) Cappiello, A.; Famiglini, G.; Palma, P. Anal. Chem. 2003, 75, 497A-503A. (6) Cappiello, A. Mass Spectrom. Rev. 1996, 15, 283-296. (7) Cappiello, A.; Famiglini, G.; Bruner, F. Anal. Chem. 1994, 66 (9), 14161423. (8) Cappiello, A., Famiglini, G. Anal. Chem. 1995, 67 (2), 412-419. (9) Cappiello, A.; Famiglini, G.; Mangani, F.; Tirillini, B. J. Am. Soc. Mass Spectrom. 1995, 6, 132-139. (10) Cappiello, A.; Famiglini, G.; Tirillini, B. Chromatographia 1995, 40 (7-8), 411-416. (11) Cappiello, A.; Famiglini, G.; Palma, P.; Berloni, A.; Bruner, F. Environ. Sci. Technol. 1995, 29 (9), 2295-2300. (12) Cappiello, A.; Famiglini, G.; Lombardozzi, A.; Massari, A.; Vadala`, G. G. J. Am. Soc. Mass Spectrom. 1996, 7, 753-758. (13) Cappiello, A.; Famiglini, G. J. Am. Soc. Mass Spectrom. 1998, 9, 993-1001. (14) Cappiello, A.; Famiglini, G.; Mangani, F.; Careri, M.; Lombardi, P.; Mucchino, C. J. Chromatogr. 1999, 855, 515-527. (15) Cappiello, A.; Balogh, M.; Famiglini, G.; Mangani, F.; Palma, P. Anal. Chem. 2000, 72, 3841-3846. (16) Cappiello, A.; Famiglini, G.; Rossi, L.; Magnani, M. Anal. Chem. 1997, 69 (24), 5136-5141. (17) Cappiello, A.; Famiglini, G.; Mangani, F.; Angelino, S.; Gennaro, M. C. Environ. Sci. Technol. 1999, 33, 3905-3910.
Table 3. Performance Evaluation of the Interface for Four Selected Compounds
linear regression equation R2 LOD (SIM), pg LOD (SCAN), ng RSD, %
dimethyl phthalate
atrazine
lindane
mestranol
Y ) 2772x - 166 0.9999 2 0.2 1.1
Y ) 1274x + 1606 1 10 0,5 1.5
Y ) 1592x - 7595 0.9984 10 0.5 3.7
Y ) 2114x + 2522 0.9988 20 1 4.3
To further reduce the adverse effects of mobile-phase vapors and system complexity, we designed a new interface, in which the eluate is directly introduced into the EI source18-23 at nanoscale flow rates (
