Analysis of model Pd- and Pt-containing contaminants in aqueous media using ESI-MS and the fragment partitioning approach

Abstract

Ubiquitous usage of Pd- and Pt-containing nanoparticles in automotive catalytic converters is an important potential threat to the environment. The unavoidable release of transition metal species to the environment and their contact with water give rise to the poisoning of ecosystems by heavy metal compounds. Electrospray ionization mass spectrometry and the newly-developed fragment partitioning approach show that a variety of metal species may be formed upon contact of metal salts with water. A series of monometallic complexes, homonuclear clusters and heteronuclear clusters of palladium and platinum were detected and characterized. The study has revealed a critical danger of metal contamination due to easy formation of transition metal clusters, which may be much more toxic than corresponding monometallic complexes.

---

1. Introduction

Due to their unique chemical and physical properties, platinum group metals (PGM) have broad applications in various areas of industry, advanced technologies, and research.^1,2 The three platinum group metals – Pd, Pt and Rh – are of paramount importance in the automobile industry as components of catalytic converters. It is important to realize the scale of precious metal consumption in the manufacturing of autocatalysts. As a representative example, >70% of the approximately 200 tons of palladium produced annually world-wide are used in the manufacturing of catalytic converters (Fig. 1a).

Fig. 1 Palladium demand by application (a)³ and a general scheme of automotive catalyst operation (b).⁴

The goal of the catalytic converter is to transform toxic exhaust gases CO, C_xH_y and NO_x into harmless CO₂, H₂O, and N₂ (Fig. 1b). Conversion occurs on the surface of nanosized precious metal catalyst (Fig. 1b) and is indeed very efficient for environmental protection.⁵ However, this fascinating technology has a dangerous drawback. Small amounts of the metals may leave the surface of the catalyst during catalytic converter operation. This process is known as leaching and is readily facilitated by the presence of coordinating molecules that are available in the surrounding gas phase (H₂O, CO, NO, CO₂ etc.).⁶ The outstanding growth of the automotive industry has resulted in increased environmental pollution by palladium (together with platinum and rhodium), which is emitted with exhaust gases from catalytic converters. The emitted palladium is deposited along roadways, on soil and plants, and in the water.⁷ Ecotoxicological measurements have shown that the palladium content in road dust reaches 250–300 μg kg⁻¹.⁸ The negative effects of platinum group metals on the biosphere cannot be excluded due to their increasing accumulation in the environment, high biological availability of transition metals, and unknown toxicological and ecological influence.⁹ It has been shown that solid waste released from autocatalysts to the environment retains high chemical reactivity.^5,10

Therefore, the investigation of PGM behavior under environmental conditions is an important problem. Electrospray ionization mass spectrometry (ESI-MS) is the method of choice for studying environmental behavior of transition metal complexes, enabling the direct analysis of water contamination. In the present work, we studied model palladium and platinum compounds in aqueous media using the highly sensitive and soft ESI-MS as an analytical technique^11–13 to examine the forms in which these metals can be present in surface water. In addition to monometallic species, we also addressed the possibility of the formation of polynuclear transition metal clusters in solutions, which are known to exhibit high chemical reactivity¹⁴ and have significant biological influence.¹⁵

2. Results and discussions

2.1. ESI-MS data analysis based on fragment partitioning

ESI-MS is a highly sensitive and informative analytical tool with numerous advantages for practical application. However, for transition metal compounds, even a simple analyzed sample can produce rather complicated spectra with hundreds of signals.¹¹ A routine ESI-MS analysis of the model solution of Na₂PdCl₄ in water carried out in the present work showed that 1536 spectral lines are present. The presence of numerous signals in the spectra is caused by the easy coordination of various ligands and the association of solvent molecules with the metal center, as well as the facile degradation and transformation of the transition metal species. In addition, each individual compound gives rise to multiple spectral lines in the high-resolution ESI-MS spectrum due to isotopic distribution.

A conventional mass spectrum analysis for the studied sample is based on the calculation of the brutto formula C_n1H_n2O_n3Cl_n4N_n5Pd_n6 and variations of coefficients n1–n6. In spite of significant time and effort, it was not possible to carry out a complete analysis of the registered mass spectrum using this standard approach. Several brutto-formula alternatives for each signal complicate the spectra assignment, and it is often difficult to unambiguously connect the brutto-formula with a reliable structure of the complex.

To solve this problem, we utilized another approach based on the partitioning of the molecular structures into a possible set of possible fragments and making variations based on fragment analysis (Fig. 2) rather than per-element analysis. To illustrate the idea, the composition of the detected species was analyzed in the form of (Metal)_n1(Solvent)_n2(Na⁺)_n3(H⁺)_n4(Ligand_1)_n5(Ligand_2)_n6…(Ligand_X)_nx, which significantly simplifies and speeds up the analysis compared with the fitting of the general brutto-formula Pd_n1C_n2H_n3O_n4Cl_n5N_n6. Resolving the data in terms of coordinated ligands and solvents directly connects the detected MS signal with the chemical nature of the species. The fragment partitioning approach is not limited to particular structures, and the data analysis can be carried out in a general form of (Metal_1)_n1(Metal_2)_n2…(Fragment_1)_n3(Fragment_2)_n4.

Fig. 2 Algorithm of fragment partitioning analysis of ESI-MS spectra.

The developed algorithm involves the following steps (Fig. 2): (1) the selection of peaks and fragments for the analysis, (2) the calculation of coefficients n1 – nx, (3) the evaluation of possible structural variations in the fragments, (4) the sorting and scoring of the results, and (5) the confirmation of the resolved peaks by simulations of isotopic patterns. A dedicated software program was created to implement the developed algorithm (see ESI†).

2.2. ESI-MS study of an aqueous solution of Na₂PdCl₄

Sodium tetrachloropalladate(II) was chosen as the model palladium compound for the study. The ESI-MS analysis of its aqueous solution showed the existence of a variety of species in solution, both in positive (Fig. 3) and negative (Fig. 4) ion modes. In spite of the simple structure of the starting metal salt, Na₂PdCl₄, rather complicated spectra containing 1291 and 245 signals were recorded for the positive and negative ion modes, respectively. A detailed analysis of the ESI-MS spectra using the developed fragment partitioning approach enabled the identification of a variety of metal species in solution (Table 1).

Fig. 3 ESI-MS spectrum of aqueous solution of Na₂PdCl₄ in positive ion mode. Signals of mononuclear palladium species are shown in red, binuclear, in blue, and trinuclear, in green; the signal shown in magenta corresponds to the [erucamide·Na]⁺ adduct. The insert shows the fine structure of the signal at m/z 460.72.

Fig. 4 ESI-MS spectrum of aqueous solution of Na₂PdCl₄ in negative ion mode. Sodium-free ions are shown in red, monosodium adducts, in green, and disodium adducts, in blue. The insert shows the fine structure of the signal at m/z 624.45.

Table 1 Species detected in the ESI-MS spectra of Na₂PdCl₄

Positive ion mode			Negative ion mode
Composition	m/z	Rel. int., %	Composition	m/z	Rel. int., %
NaPdCl2(CH3CN)	241.86	1.08	PdCl3	214.81	0.76
Na2PdCl3	258.79	1.64	NaPdCl4	270.77	0.40
NaPdCl2(CH3CN)2	282.88	2.96	Na2PdCl5	328.73	3.66
Na2PdCl3(CH3CN)	301.81	1.53	Pd2Cl5	390.65	100.00
NaPdCl3(CH3CN)2	316.75	2.26	NaPd2Cl6	448.61	70.11
Pd2Cl3(CH3CN)2	402.78	3.13	Na2Pd2Cl7	508.57	10.83
NaPd2Cl4(CH3CN)	419.70	12.08	Pd3Cl7	568.50	21.05
NaPd2Cl4(CH3CN)2	460.72	49.74	NaPd3Cl8	624.45	61.78
Na2Pd2Cl5(CH3CN)	477.66	14.62	Na2Pd3Cl9	684.41	34.08
Na3Pd2Cl6	502.19	70.36	Pd4Cl9	742.38	7.00
NaPd3Cl6(CH3CN)2	633.47	3.40	NaPd4Cl10	800.31	6.44
Na2Pd3Cl7(CH3CN)2	653.50	3.22	Na2Pd4Cl11	861.26	5.16
Na3Pd3Cl8	672.43	4.82	Pd5Cl11	918.22	8.08
Na4Pd3Cl9	732.39	3.47	NaPd5Cl12	977.17	8.23
Na5Pd3Cl10	790.35	2.21
Na6Pd3Cl11	855.74	3.43

---

In positive ion mode, the ESI-MS spectrum (Fig. 3, Table 1) is dominated by binuclear palladium species of various compositions. Because positive ions of transition metal complexes formed during electrospray ionization tend to be electrophilic and coordinatively unsaturated, several signals corresponding to solvent adducts can be detected in the spectrum. The ESI-MS spectrum in negative ion mode (Fig. 4, Table 1) was more informative because the Na₂PdCl₄ starting material can easily generate anionic species following the dissociation of the sodium cations. Additionally, compared with the spectrum in positive ion mode, a significantly lower signal overlap was observed. The following trends can be distinguished: signals of sodium-free anions (red colored) show the highest intensity in the case of binuclear species; for monosodium adducts (green colored), signals of dinuclear and trinuclear species are nearly the same intensity; and for disodium adducts, the most intensive signal corresponds to trinuclear clusters.

Thus, it may be proposed that sodium ions stabilize trinuclear chloropalladate anions during electrospray ionization because in the case of larger cluster anions (Pd₄, Pd₅), there is no significant difference in signal intensity between the sodium-free chloropalladates and the sodium containing adducts; therefore, the stability of polynuclear clusters does not significantly depend on the presence of sodium. Comparisons between the spectra in both ESI-MS ion modes show that binuclear chloropalladates are prevalent in solution because the corresponding signals show the highest intensity in the spectra regardless of ion mode and particle composition.

2.3. ESI-MS study of an aqueous solution of Na₂PtCl₆

Sodium tetrachloroplatinate(II) was chosen as the model compound for the spectral study of platinum complexes. As in the previous sample, in spite of the simple structure of the metal salt Na₂PtCl₆, complicated spectra containing numerous signals were detected (Fig. 5 and 6). The nature of the signals in the spectra was elucidated using the approach described above (Table 2).

Fig. 5 ESI-MS spectrum of aqueous solution of Na₂PtCl₆ in positive ion mode. Signals of mononuclear platinum species are shown in red, binuclear, in blue, trinuclear, in green, tetranuclear, in cyan, pentanuclear, in orange, and hexanuclear, in violet; the signal shown in magenta corresponds to the [erucamide·Na]⁺ adduct. The insert shows the fine structures of the signals corresponding to binuclear Pt species.

Fig. 6 ESI-MS spectrum of aqueous solution of Na₂PtCl₆ in negative ion mode. Sodium- and OH-free ions are shown in red, and sodium adducts, in green, OH-containing sodium adducts are shown in blue, and binuclear Pt species, in cyan. The insert shows the fine structure of the signal at m/z 430.76.

Table 2 Species detected in the ESI-MS spectra of Na₂PtCl₆

Positive ion mode			Negative ion mode
Composition	m/z	Rel. int., %	Composition	m/z	Rel. int., %
Na3(OH)PtCl3	386.84	16.64	PtCl3	300.87	1.87
Na3PtCl4	404.81	14.38	PtCl4	335.84	34.11
Na3(OH)PtCl5	456.78	80.20	Pt(OH)Cl4	352.84	1.89
Na3PtCl6	476.74	37.40	NaPtCl4	358.83	21.04
Na4(OH)PtCl6	516.74	9.43	PtCl5	372.81	38.80
Na4PtCl7	534.70	4.92	NaPtCl5	398.82	2.33
Na5Pt2(OH)2Cl10	892.57	6.76	NaPt(OH)Cl5	410.80	27.52
Na5Pt2(OH)Cl11	912.54	7.65	NaPtCl6	430.76	100.00
Na5Pt2Cl12	930.50	4.91	Na2Pt(OH)Cl6	470.75	27.12
Na6Pt2(OH)2Cl11	952.53	4.20	Na2PtCl7	488.72	6.61
Na6Pt2(OH)Cl12	970.49	4.73	Na3Pt(OH)Cl7	528.71	3.60
Na6Pt2Cl13	988.46	2.57	Na3PtCl8	546.68	2.05
Na7Pt2(OH)2Cl12	1008.48	1.33	Na4Pt(OH)Cl8	586.67	0.44
Na7Pt2(OH)Cl13	1026.45	1.47	Na5Pt2Cl8	788.62	1.64
Na7Pt3(OH)3Cl15	1328.37	1.21	Na2Pt2Cl11	824.55	1.99
Na7Pt3(OH)2Cl16	1347.33	2.24	Na6Pt2Cl9	848.58	3.49
Na7Pt3(OH)Cl17	1366.29	2.80	Na3Pt2Cl12	884.51	9.31
Na7Pt3Cl18	1382.26	1.85
Na8Pt3(OH)3Cl16	1405.28	1.78
Na8Pt3(OH)2Cl17	1423.25	2.05
Na8Pt3(OH)Cl18	1442.20	0.99
Na8Pt3Cl19	1463.25	0.83

---

In positive ion mode, mononuclear platinum species dominate in the ESI-MS spectrum (Fig. 5, Table 2). Another significant feature of the spectrum is the presence of a large number of signals corresponding to OH-containing ions; this leads us to suggest that chloroplatinates are more susceptible to hydrolysis and reactions with water than chloropalladates. This fact is in accordance with the reported potentiometric, spectrophotometric and NMR data on the hydrolysis of [PtCl₆]²⁻.¹⁶ The intensity of the signals corresponding to species containing 4–6 Pt atoms is very low; thus, their analysis results are unreliable, and the composition data for these ions are not shown in Table 2.

The ESI-MS spectrum of Na₂PtCl₆ in negative ion mode (Fig. 6, Table 2) also differs significantly from the analogous Na₂PdCl₄ spectrum. Again, the signals of the mononuclear species dominate the spectra, and the presence of OH-containing ions is clearly observed. In the case of the Na₂PtCl₆ solution, the signals correspond to Na⁺-free species (red color; Fig. 6) are lower in intensity than those for the sodium-adducts (green and blue color; Fig. 6). Thus, by comparing the spectra of Na₂PtCl₆ in both ESI-MS ion modes, we can propose that a solution of chloroplatinate in water gives rise to predominately mononuclear Pt species with appreciable amounts of partially hydrolyzed ions.

2.4. ESI-MS study of the mixture of Na₂PdCl₄ and Na₂PtCl₆ in aqueous solution

The release of both metals to environment may initiate the formation of heterometallic species in water solutions. To examine the possibility of the formation of heteronuclear Pd–Pt clusters in the solution containing both transition metals, an aqueous mixture of Na₂PdCl₄ and Na₂PtCl₆ was studied using ESI-MS.

In positive ion mode (Table 3), the signals of the heteronuclear species showed very low intensity, whereas in negative ion mode, the intensity of the heteronuclear species signals was rather high (Table 3). It is important to note that all detected heteronuclear clusters contained a single Pt atom, whereas the quantity of Pd atoms in the clusters varied from 1 to 3. This finding is in accordance with the observed difference between Pd and Pt in the tendency to form polynuclear species (cf. Fig. 3 and 5). Thus, these experiments have shown that polynuclear heterometallic clusters can be easily formed if both metal ions are presented in solution.

Table 3 Heteronuclear species detected in the ESI-MS spectra of a mixture of Na₂PdCl₄ and Na₂PtCl₆

Positive ion mode			Negative ion mode
Composition	m/z	Rel. int., %	Composition	m/z	Rel. int., %
Na3PdPtCl8	654.59	1.34	NaPdPtCl8	606.59	16.16
Na4PdPtCl9	712.54	1.86	Na2PdPtCl9	666.55	26.25
Na5PdPtCl10	768.50	0.86	NaPd2PtCl10	784.43	1.28
Na6PdPtCl11	828.45	0.34	Na2Pd2PtCl11	844.38	3.28
Na7PdPtCl12	886.40	0.30	Na3Pd2PtCl12	902.34	6.07
Na5Pd2PtCl12	946.33	0.66	NaPd3PtCl12	962.29	0.39
Na6Pd2PtCl13	1006.30	0.24

---

As a representative example, comparisons of the simulated isotope pattern distribution and the experimentally measured spectrum for the heterometallic cluster Na₂PdPtCl₉⁻ are shown in Fig. 7. Excellent agreement between the measured and calculated spectra confirms the identified composition of the ions. Verification of the peak assignment was carried out for all resolved signals in the present study (see ESI†). It should be noted that the assessment of the signal fine structure is a well-established and reliable tool in mass-spectrometric analysis.

Fig. 7 Experimental and calculated isotope patterns for the heteronuclear ion Na₂PdPtCl₉⁻.

Overall, ESI-MS is a very efficient analytic tool for studying solutions of transition metal complexes. Nevertheless, it is important to mention some known drawbacks. The outcome of the analysis may be concentration dependent¹² and may be very sensitive to the presence of organic contaminants.¹⁷ The impact of the initial presence of the ions in solution and the possibility of ion generation during the electrospray process are also important considerations.^12–14,18

The mechanism of formation of polynuclear clusters in solution is of much importance to understand reactivity of metal particles and to reveal possible pathways under catalytic conditions. It was proposed that leaching of metal species and oriented attachment of metal species are the key processes for tuning the performance of catalytic reactions and they have a large impact on metal contamination.^6,19 More detailed studies on this fascinating topic will be a subject of out future research work.

3. Conclusions

Although studied palladium and platinum salts have simple composition, the corresponding solutions possess rather complicated ESI-MS spectra. Detailed spectral analysis is an important task for studying potential risk of environmental pollution, as well as for the fields of catalysis and organometallic chemistry. These Pd and Pt salts are ubiquitously used as catalyst precursors and staring materials for preparation of more complex compounds. Signal assignment in the spectra of starting materials is required for mechanistic studies and reaction monitoring.

The ESI-MS study of aqueous solutions of Na₂PdCl₄ and Na₂PtCl₆ has revealed that the platinum salt predominantly consists of mononuclear species in solution, whereas binuclear species predominate in palladium salt solutions. An analysis of the aqueous mixture of two salts has shown the possibility of the formation of heteronuclear chlorinated clusters with the general formula of the metal core PtPd_n (n = 1…3), which can be detected in positive and negative ESI-MS ion modes. These findings are highly important from an ecotoxicological point of view because the chemical properties and biological activity of monometallic complexes, homonuclear clusters and heteronuclear clusters may differ substantially. The implementation of the fragment partitioning approach and development of corresponding software enabled rapid analysis of complicated ESI-MS spectra of metal-containing compounds. The present study has revealed a potential danger of metal contamination due to easy formation of transition metal clusters, which can be much more toxic, as compared to the corresponding monometallic complexes.

The overall amount of metals can be determined by known methods (ICP-MS, spectroscopy, etc.), while revealing the structure and understanding reactivity of metal contaminants is a challenge. As far as environmental issues are concerned, it should be noted that the power of ESI-MS is to reveal the nature of metal complexes and to study transformations of transition metal complexes in solution. We anticipate the rapid development of the ecotoxicological analysis of Pd- and Pt-contamination of the environment, and high performance analytic methods will provide valuable insights into the molecular mechanisms of these processes.

4. Experimental part

4.1. Preparation of the solutions

A total of 0.1 mmol of a corresponding salt (Na₂PdCl₄ or Na₂PtCl₆·6H₂O) was dissolved in 10 mL of HPLC-grade water (Sigma) and stirred overnight at room temperature. In the mixed metal cluster formation experiment, equal volumes of two solutions were stirred overnight at room temperature. Prior to the ESI-MS measurements, each sample was diluted 100-fold by HPLC-grade acetonitrile (Merck).

4.2. ESI-MS measurements

High resolution mass spectra were recorded using a Bruker maXis instrument (Bruker Daltonik GmbH, Bremen, Germany) equipped with an electrospray ionization (ESI) ion source. The measurements were performed in positive (+MS) and negative (−MS) ion modes (HV capillary: 4500 V (for positive) and 3000 V (for negative); HV end plate offset: −500 V) with the following scan range in m/z: 50–3000. External calibration of the mass spectrometer was performed with Electrospray Calibrant Solution (Fluka). Direct syringe injection was used for the analyzed solutions in MeCN (flow rate: 3 μL min⁻¹). Nitrogen was used as the nebulizer gas (0.4 bar) and dry gas (4.0 L min⁻¹); the drying temperature was set at 180 °C. The recorded spectra were processed using the Bruker DataAnalysis 4.0 software package.

4.3. Software development

The program used to perform the fragment partitioning approach was developed in Qt Creator 5.5 using C++ language. The target mass was decomposed as a linear combination of fragment exact masses with natural coefficients. The maximum coefficient of the fragment was limited by the integer part of the quotient obtained by dividing the target mass by the mass of the fragment. Because the coefficients are natural and bounded from above, it was possible to check all variants using restriction criteria (target mass and tolerance).

Acknowledgements

ESI-MS and chemical studies were supported by Russian Science Foundation (RSF Grant 14-13-01030). Development of mathematical procedures was partially supported RFBR (Grant 14-03-31465). The authors gratefully acknowledge Dr Julia V. Burykina for the assistance and helpful discussions.

References

1. (a) H. Renner, G. Schlamp, I. Kleinwächter, E. Drost, H. M. Lüschow, P. Tews, P. Panster, M. Diehl, J. Lang, T. Kreuzer, A. Knödler, K. A. Starz, K. Dermann, J. Rothaut, R. Drieselmann, C. Peter and R. Schiele, in Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2000; (b) L. B. Hunt and F. M. Lever, Platinum Met. Rev., 1969, 13, 126; (c) Chemistry of the Platinum Group Metals. Recent Developments, ed. F. R. Hartley, Elsevier Science, 1991; (d) Precious Metals for Biomedical Applications, ed. N. Baltzer and T. Copponnex, Woodhead Publishing, 2014.
2. Catalytic applications: (a) A. R. Muci and S. L. Buchwald, in Practical palladium catalysts for C–N and C–O bond formation, ed. N. Miyaura, Springer, Berlin Heidelberg, 2002, pp. 131–209; (b) J. Hartwig, Organotransition Metal Chemistry: from Bonding to Catalysis, University Science Books, Sausalito, 2009; (c) Metal-Catalyzed Cross-Coupling Reactions and More, ed. A. de Meijere, S. Bräse and M. Oestreich, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2013; (d) Olefin Metathesis: Theory and Practice, ed. K. Grela, John Wiley & Sons Inc., Hoboken, New Jersey, 2014; (e) Homogeneous Catalysis for Unreactive Bond Activation, ed. Z.-J. Shi, John Wiley & Sons Inc., Hoboken, New Jersey, 2015; (f) J. M. Thomas and W. J. Thomas, Principles and Practice of Heterogeneous Catalysis, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2015; (g) Hydrofunctionalization, ed. V. P. Ananikov and M. Tanaka, Springer, Berlin, 2013.
3. A. Cowley, PLATINUM 2013 Interim Review, Johnson Matthey PLC, Royston, 2013.
4. Adapted from http://www.preciousmetals.umicore.com/recyclables/SAC/catalyticconverter/.
5. (a) J. Kašpar, P. Fornasiero and N. Hickey, Catal. Today, 2003, 77, 419; (b) P. J. Millington and A. P. E. York, Platinum Met. Rev., 2012, 56, 58; (c) C. Morgan, Johnson Matthey Technol. Rev., 2014, 58, 217.
6. A. S. Kashin and V. P. Ananikov, J. Org. Chem., 2013, 78, 11117.
7. (a) G. C. Lough, J. J. Schauer, J.-S. Park, M. M. Shafer, J. T. DeMinter and J. P. Weinstein, Environ. Sci. Technol., 2005, 39, 826; (b) Palladium Emissions in the Environment: Analytical Methods, Environmental Assessment and Health Effects, ed. F. Zereini and F. Alt, Springer, 2006.
8. K. Boch, M. Schuster, G. Risse and M. Schwarzer, Anal. Chim. Acta, 2002, 459, 257.
9. (a) K. Ravindra, L. Bencs and R. Van Grieken, Sci. Total Environ., 2004, 318, 1; (b) Immunotoxicology of Environmental and Occupational Metals, ed. J. T. Zelicoff and P. Thomas, CRC Press, 1998.
10. E. Mieczynska, A. Gniewek and A. M. Trzeciak, Appl. Catal., A, 2012, 421–422, 148.
11. (a) R. B. Cole, Electrospray Ionization Mass Spectrometry; Fundamentals, Instrumentation, and Applications, John Wiley, New York, 1997; (b) J. H. Gross, Mass Spectrometry, Springer, Berlin, 2004.
12. (a) L. P. E. Yunker, R. L. Stoddard and J. S. McIndoe, J. Mass Spectrom., 2014, 49, 1; (b) K. L. Vikse, Z. Ahmadi and J. S. McIndoe, Coord. Chem. Rev., 2014, 279, 96.
13. D. Schröder, Acc. Chem. Res., 2012, 45(9), 1521.
14. L. V. Romashov, L. L. Khemchyan, E. G. Gordeev, I. O. Koshevoy, S. P. Tunik and V. P. Ananikov, Organometallics, 2014, 33, 6003.
15. M. Tsoli, H. Kuhn, W. Brandau, H. Esche and G. Schmid, Small, 2005, 8–9, 841.
16. (a) C. M. Davidson and R. F. Jameson, Trans. Faraday Soc., 1965, 61, 2462; (b) B. Shelimov, J. F. Lambert, M. Che and B. Didillon, J. Am. Chem. Soc., 1999, 121, 545.
17. D. B. Eremin and V. P. Ananikov, Organometallics, 2014, 33, 6352.
18. V. B. Di Marco and G. G. Bombi, Mass Spectrom. Rev., 2006, 25, 347.
19. (a) R. H. Crabtree, Chem. Rev., 2012, 112, 1536; (b) N. Y. S. Phan, M. van Der Sluys and C. W. Jones, Adv. Synth. Catal., 2006, 348, 609; (c) C. Deraedt and D. Astruc, Acc. Chem. Res., 2014, 47, 494; (d) A. Trzeciak and J. J. Ziolkowski, Coord. Chem. Rev., 2007, 251, 1281–1293; (e) E. O. Pentsak, A. S. Kashin, M. V. Polynski, K. O. Kvashnina, P. Glatzel and V. P. Ananikov, Chem. Sci., 2015, 6, 3302; (f) V. P. Ananikov, ACS Catal., 2015, 5, 1964.

---

Footnote

† Electronic supplementary information (ESI) available: Description of the developed program and ESI-MS spectra are available in the electronic supplementary information. See DOI: 10.1039/c5ra22257e

---