Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Gas phase ion-molecule chemistry of ironcyclopentadienyl and ironcyclopentadiene ions with C1-C4 alcohols… Korenkova, Eva 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1998-0081.pdf [ 4.9MB ]
Metadata
JSON: 831-1.0059552.json
JSON-LD: 831-1.0059552-ld.json
RDF/XML (Pretty): 831-1.0059552-rdf.xml
RDF/JSON: 831-1.0059552-rdf.json
Turtle: 831-1.0059552-turtle.txt
N-Triples: 831-1.0059552-rdf-ntriples.txt
Original Record: 831-1.0059552-source.json
Full Text
831-1.0059552-fulltext.txt
Citation
831-1.0059552.ris

Full Text

Gas Phase Ion-Molecule Chemistry of Ironcyclopentadienyl and Ironcyclopentadiene Ions with C1-C4 Alcohols by Fourier Transform Ion Cyclotron Resonance Mass Spectrometry by Eva Korenkova B.Sc., Komenskeho University, 1992 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR T H E D E G R E E OF MASTER OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES Department of Chemistry We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 1997 © Eva Korenkova, 1997 ln presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of o W i s t ^ The University of British Columbia Vancouver, Canada Date ^ A V M " - ^ Ik^ V^^S. DE-6 (2/88) Abstract The gas phase ion-molecule chemistry of organometallic ions CpFe + (1) and C5H.6Fe+ (2), shown below, has been examined by Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometry. 1 2 These ions, generated by electron impact ionization from organometallic neutral precursors ferrocene and cyclohexadienone iron tricarbonyl, were allowed to react with neutral methanol, ethanol, n-, i-propanol, and n-, i-, f-butanol. Selected isotopologues of employed alcohols were used to infer the reaction mechanism. Abundances of ionic reactants and products were monitored as a function of reaction time to obtain kinetic data. A general trend was observed that the reactivity of ions 1 and 2 with alcohols increases with the increasing length of the alkyl chain. The most important difference in the gas phase chemistry of these two ions was noted for the reactions with methanol. CpFe + forms only an ion-neutral complex without any bond activation. CsHgFe"1" was found to dehydrogenate methanol by a competitive insertion into C-H and H -0 bonds. Taking the ratio of ADO collision rates and the measured reaction rates it was found that ion 2 has a twofold increase in reaction efficiency as compared to ion 1. For the reactions of CpFe + with higher alcohols the major processes observed were dehydration and loss of alkene. Dehydrogenation of alcohols was a minor reaction. In all CsHyFe+Zalcohol systems except for methanol, reductive elimination of water and alkene was always accompanied by loss of a hydrogen molecule. Experiments with labeled alcohols have shown that one of the - i l l -hydrogen atoms in the eliminated neutrals originates on a cyclopentadiene ring. This was rationalized by proposing that the reacting form of C5H.6Fe+ ion has a CpFe + -H structure, formed by an intramolecular 6-hydrogen transfer that results in a metal-hydrogen bond. -iv-Table of Contents Abstract ii Table of Contents iv List of Tables vi List of Figures vii List of Abbreviations and Symbols xii Acknowledgments xiii Chapter 1 Introduction 1 1.1 Research Interest 1 1.2 Gas Phase Ion-Molecule Reactions of Organometallic Ions with Alcohols 5 1.2.1 Methanol 5 1.2.2 Ethanol 8 1.2.3 n-Propanol 9 1.2.4 /-Propanol 12 1.2.5 C4 and Higher Alcohols 14 1.3 Experimental Section 18 1.3.1 Operating Principles of FT-ICR and Experimental Techniques 18 1.3.2 Experimental Parameters 24 1.3.3 Basic Kinetics of Gas Phase Ion-Molecule Reactions 25 References 28 Chapter 2 Gas Phase Ion-Molecule Reactions of Cyclopentadienyliron Ion ( r | 5 - C 5 H 5 ) F e + with Alcohols 32 2.1 Reactions with Methanol 34 -V-2.2 Reactions with Ethanol 40 2.3 Reactions with n-Propanol 46 2.4 Reactions with j-Propanol 54 2.5 Reactions with n-, i- and r-Butanol 60 2.6 Summary and Conclusions 67 References 75 Chapter 3 Gas Phase Ion-Molecule Reactions of Cyclopentadieneiron Ion C 5 H 6 Fe+ with Alcohols 76 3.1 Reactions with Methanol 79 3.2 Reactions with Ethanol 86 3.3 Reactions with /-Propanol 100 3.4 Reactions with n-Propanol I l l 3.5 Reactions with n-,i- and f-Butanol 116 3.6 Summary and Conclusions 126 References 133 -vi-List of Tables Table 2.1 Primary product distributions for the ion-molecule reactions of CpFe+ with alcohols 33 Table 2.2 Primary product distributions for the ion-molecule reactions of CpFe + with alcohols 68 Table 2.3 Experimental and theoretical rate constants for the reaction of CpFe + with alcohols 72 Table 3.1 Primary product distributions for the ion-molecule reactions of C 5 H 6 Fe+ with alcohols 78 Table 3.2 Ionic products observed for the reaction of C5H^Fe+ with methanol and its isotopologues 83 Table 3.3 Ionic products observed for the reaction of C 5H.6Fe+ with ethanol and its isotopologues 91 Table 3.4 Ionic products observed for the reaction of CpFe(OH)+ with ethanol and its isotopologues 94 Table 3.5 Formation of CpFe^H^)"1" observed for the reaction of C5H6Fe+ with /-propanol and its isotopologues 104 Table 3.6 Formation of CpFe(OC3Hs)+ observed for the reaction of C 5 HgFe + with /-propanol and its isotopologues 109 Table 3.7 Primary product distributions for the ion-molecule reactions of C 5 H 6 Fe+ with alcohols 128 Table 3.8 Experimental and theoretical rate constants for the reaction of C 5 H 6 F e + with alcohols 130 -Vl l -List of Figures Figure 1.1 Cubic ICR cell 19 Figure 1.2 FT-ICR ion excitation and detection 20 Figure 1.3 FT-ICR experimental pulse sequence of events for single resonance, double resonance (DBL), and collision-induced dissociation (CID) experiments 22 Figure 2.1 Temporal variation of ion abundancies in the CpFe + /CH 3 OH/Cp 2 Fe system at a total pressure of 1.2 x 10"6 Torr 35 Figure 2.2a-b Partial triple resonance mass spectra illustrating the reactions of CpFe+ with a) C H 3 O H at delay time 200 ms, b) C H 3 O D at delay time 250 ms 37 Figure 2.2c-d Partial triple resonance mass spectra taken at delay time 200 ms illustrating reactions of CpFe+ with c) C D 3 O H , d) C D 3 O D 38 Figure 2.3a Triple resonance mass spectrum illustrating the reactions of CpFe + with ethanol at delay time 200 ms. The total pressure is 1.2 x 10~6 Torr 40 Figure 2.3b-c Triple resonance mass spectra illustrating the reactions of CpFe + with b) C H 3 C H 2 O D at delay time 250 ms, c) C H 3 C D 2 O H , delay time 200 ms. Total pressure of neutrals is 1.2 x 10 - 6 Torr 41 Figure 2.3d-e Partial triple resonance mass spectra illustrating the reactions of CpFe+ with d) C D 3 C H 2 O H , delay time 150 ms, e) C 2 D 5 O D , delay time 210 ms. Total pressure of neutrals is 1.2 x 10 - 6 Torr 42 Figure 2.4 Temporal variation of normalized ion intensities in the CpFe+/EtOH/Cp2Fe system 43 -vm-Figure 2.5 Partial triple resonance mass spectra taken at delay time 200 ms illustrating the reactions of CpFe + with a) n - C 3 H 7 O H and b) n - C 3 H 7 O D 48 Figure 2.6 Temporal variation of ion intensities in the CpFe+/n-PrOH/Cp2Fe system. Total pressure of neutrals is 1.2 x 10"6 Torr 49 Figure 2.7 Partial triple resonance mass spectra illustrating reactions of CpFe(H20)+ with n-propanol. a) Cp(H 2 0) + formed by reaction 2.11 and isolated by multiple resonance after a 100 ms delay time, b) Products of the reaction of Cp(H 2 0) + with n-propanol after additional 100 ms reaction time 52 Figure 2.8 Partial triple resonance mass spectra illustrating the reactions of Cp(C3Hg)+ with n-propanol. a) CpFe(C 3H5) + formed by reaction 2.12 and isolated by multiple resonance after a 100 ms delay time, b) Products of ion-molecule reactions of CpFe(C3Hg)+ with n-propanol after additional 150 ms delay 53 Figure 2.9a FT-ICR triple resonance mass spectrum of CpFe + reacting with i-propanol taken at a reaction delay time 180 ms 55 Figure 2.9b-c FT-ICR triple resonance mass spectra of CpFe + reacting with /-PrOD and *'-PrOH-d6, respectively. Reaction delay times are 100 ms (spectrum b) and 200 ms (spectrum c) 56 Figure 2.9d FT-ICR triple resonance mass spectrum of CpFe + reacting with /-PrOH-dg taken at a reaction delay time 250 ms 57 Figure 2.10 Temporal variation of ion intensities in the CpFe+//-PrOH/Cp2Fe system 57 Figure 2.1 la-b Triple resonance mass spectra illustrating reactions of CpFe + with a) n-BuOH, reaction time 140 ms and b) n-BuOD, reaction time 100 ms 61 -ix-Figure 2.11c Triple resonance mass spectrum illustrating reactions of CpFe + with rc-BuOH-fJ.9 at a delay time 100 ms 62 Figure 2.12 Triple resonance mass spectrum taken at delay time 150 ms illustrating reactions of CpFe + with /-BuOH 62 Figure 2.13a-b Triple resonance mass spectra illustrating reactions of CpFe + with a) r-BuOH, reaction delay time 150 ms and b) r-BuOH-din> reaction delay time 200 ms 63 Figure 2.14 Temporal variation of ion intensities in the CpFe +/n-BuOH/Cp 2Fe system 65 Figure 2.15 Temporal variation of ion intensities in the CpFe +//-BuOH/Cp 2Fe system 66 Figure 2.16 Temporal variation of ion intensities in the CpFe+/r-BuOH/Cp 2Fe system 66 Figure 2.17 The trend in reactivity of CI to C4 alcohols with CpFe + ion 72 Figure 3.1. Temporal variation of ion intensities in the C 5H 6Fe+/MeOH/(C 6H 60)Fe(CO) system 79 Figure 3.2a-b Partial triple resonance mass spectra illustrating the reactions of C 5 H 6 Fe+ with a) C H 3 O H at delay time 120 ms, b) C H 3 O D at delay time 200 ms 81 Figure 3.2c-d Partial triple resonance mass spectra illustrating the reactions of C 5 H 6 Fe+ with c) C D 3 O H at delay time 180 ms, d) C D 3 O D at delay time 150 ms 82 Figure 3.3a Partial triple resonance mass spectrum illustrating the reactions of C5HgFe + with ethanol at delay time 120 ms. Total pressure of neutrals is 1.2 x 10"6 Torr 87 -X-Figure 3.3b-c Partial triple resonance mass spectra illustrating the reactions of C 5 H 6 F e + with b) C H 3 C H 2 O D at delay time 100 ms and c) C H 3 C D 2 O H at delay time 100 ms 88 Figure 3.3d-e Partial triple resonance mass spectra illustrating the reaction of C 5 H 6 F e + with d) C D 3 C H 2 O H at delay time 100 ms and e) C 2 D 5 O D at delay time 150 ms. Total pressure of neutrals is 1.2 x 10"6Torr 89 Figure 3.4 Temporal variation of ion intensities in the C 5 H 6 Fe + /EtOH/(C 6 H 6 0)Fe(CO) 3 system 90 Figure 3.5a-b Partial triple resonance mass spectra illustrating the reactions of C 5 H 6 Fe+ with a) / - C 3 H 7 O H at delay time 150 ms, b) / - C 3 H 7 O D at delay time 130 ms 101 Figure 3.5c-d Partial triple resonance mass spectra illustrating the reactions of C 5 H 6 F e + with c) i - ( C D 3 ) 2 C H O H at delay time 100 ms, d) ; - C 3 D 7 O D at delay time 150 ms 102 Figure 3.6 Temporal variation of ion intensities in the C 5 H 6 Fe+//-C 3 H 7 OH/(C 6 H 6 0)Fe(CO) 3 system 103 Figure 3.7 Triple resonance mass spectra illustrating the reaction of C5HgFe+ with a) n - C 3 H 7 O H at delay time 150 ms b) n - C 3 H 7 O D at delay time 90 ms. The total pressure of neutrals is 1.2 x 10 - 6 Torr 113 Figure 3.8 Temporal variation of ion intensities in the C 5 H 6 F e + / n - P r O H / ( C 6 H 6 0 ) F e ( C O ) 3 system 114 Figure 3.9a-b Partial triple resonance mass spectra taken at a delay time 150 ms illustrating the reactions of C 5 H 6 F e + with a)n-BuOH and b) n-BuOD 118 -xi-Figure 3.9c Partial triple resonance mass spectrum taken at a reaction delay time 150 ms illustrating the reactions of C5H 6 Fe+ with n-BuOH-d 9 119 Figure 3.10 Partial triple resonance mass spectra taken at a delay time 120 ms illustrating reactions of C 5 H 6 F e + with /-BuOH 119 Figure 3.11 Partial triple resonance mass spectra illustrating reactions of C 5 H 6 F e + with a) t-BuOU at delay time 120 ms and b) r-BuOH-d 1 0 at delay time 100 ms 120 Figure 3.12 Temporal variation of ion intensities in the C 5H 6Fe+/n-BuOH/(C 6H 60)Fe(CO) 3 system. 121 Figure 3.13 Temporal variation of ion intensities in the C 5 H 6 F e + / / - B u O H / ( C 6 H 6 0 ) F e ( C O ) 3 system 121 Figure 3.14 Temporal variation of ion intensities in the C 5 H 6 F e + / r - B u O H / ( C 6 H 6 0 ) F e ( C O ) 3 system 122 Figure 3.15 The trend in reactivity of C1 to C4 alcohols with C 5 H 6 F e + ion 130 - X l l -List of Abbreviations and Symbols B magnetic field induction BDE Bond Dissociation Energy CTX) Collision-Induced Dissociation Cp Cyclopentadienyl ( C 5 H 5 ) DBL DouBLe Resonance ED Electron Deficiency EI Electron Impact ionization Eqn. Equation eV electron Volt FT-ICR Fourier Transform Ion Cyclotron Resonance mass spectrometry kHz kiloHerz MHz MegaHerz m/z mass-to-charge ratio of ions ms millisecond OMN OrganoMetallic Neutral rf radiofrequency SGL SinGLe resonance TPL TriPLe resonance V Volt co cyclotron frequency - X l l l -Acknowledgments I would like to express my sincere gratitude to my research supervisor, Dr. Melvin B. Comisarow, for his guidance, support and patience during the course of this work, to my colleague Dr. Jinru Wang for enlightening discussions and suggestions, to Dr. G. Eigendorf and Lina for the opportunity and the assistance in learning other mass spectrometric techniques, and to the staff of electronic shop and mechanical shop, especially to Martin, Milan and Ron, for their invaluable work on repairing the instrument. I extend special thanks to my parents and to my friends Rob, Steve and Kevin for their continuing moral support. -1-Chapter 1 Introduction In the past two decades there has been a keen interest in the gas phase transition metal ion chemistry. Since the first observations of metal ion activation of saturated organic compounds by Allison, Freas, and Ridge1, naked metal ions, ligated metal ions and small metal cluster ions have been observed to be highly reactive and interesting species. The gas phase reactions of bare metal ions provide an insight into intrinsic properties of metals in the absence of complicating factors such as ligand/solvent effects and ion pairing. Periodic trends can be evaluated as a function of the electronic structure of the metal ion. The metal-ligand bond energies and other thermochemical properties can be applied to predict reaction mechanisms. When these data are applied to the condensed phase systems the effect of solvent can be evaluated. The results from gas phase studies, because of their relative simplicity, are ideal for comparing with theoretical calculations. 1.1 Research Interest Jacobson and Freiser2 studied the reactions of the group 8 transition-metal ions (Fe +, C o + and Ni + ) with cyclic hydrocarbons in the gas phase by FT-ICR MS. They observed that reactions of M + with cyclopentane and cyclopentene yield M-(c-C5H.6)+ ion. The nature of the M-(c-C5H6)+ ion was probed by using collision-induced dissociation and H/D exchanges. They observed that the Fe-c-C5H6+ and Co-c-C5H6 + undergo six H/D exchanges in the presence of excess deuterium (reaction 1.1, where M = Fe, Co). (C 5 H 6 )M+ + 6D 2 -> (C 5D 6)M+ + 6HD (1.1) -2-To explain this observation, they proposed a mechanism that invokes a rapid equilibrium between the cyclopentadiene complex 1 and the hydrido-7U-cyclopentadienyl complex 2 (reaction 1.2, where M = Fe, Co). (1.2) H 1 2 Byrd and Freiser examined the gas phase reactions of R h + with propane and 2-methylpropane using FT-ICR M S 3 . The secondary reactions of the product ion Rh(C3H5R)+ (R=H for propane, R=CH3 for 2-methylpropane) suggest that this ion does not have a simple metal-propene structure 3, but may rearrange to a metal-hydride complex 4. This is supported by H/D experiments with deuterium. The 5 H/D exchanges with deuterium that the RhC3H5R + undergoes can be rationalized by a rapid equilibrium between the propene and hydrido-7t-allyl species (equation 1.3). R R (1.3) R h + R h + \ H 3 4 Hydrido-7t-allyl complexes formed by reversible insertion into an allylic carbon-hydrogen bond (3—»4) have been proposed as intermediates for the catalytic isomerization of olefins4. Complexes containing a metal-hydrogen bond formed by B-hydrogen transfer have been proposed as intermediates for C-C and C-H bond activation of hydrocarbons by gas phase bare or ligated transition-metal ions2 (Scheme 1.1). R R H I 6-H X / - R H M + M + - M + >• | c-c/ I M + + R C H 2 C H 2 R ' C H 2 C H 2 R ~ ^ r \ R C - H \ H H H I B - H \ / - H 2 M + M + ~~ M + * | R R R R Scheme 1.1 R C H C H 2 R / \ Kan 5 studied gas phase chemistry of organometallic compounds of Fe and Cr that contained cyclopentadiene ( C 5 H 6 ) , cycloheptatriene ( C 7 H 8 ) and cyclohexadienone (C6H6O) ligands. He observed that some ions that resulted from gas phase ion-molecule reactions had unusually low reactivity despite their high electron deficiency. Previous studies on ion-molecule clustering reactions of 18-electron organometallic complexes has shown that the reactivity of an ion is typically governed by the electron deficiency from 18 electrons of the central metal ion: the greater the electron deficiency, the greater the reactivity6. Kan 5 observed that the reactivity of (C5H6)Fe+ is smaller than that of (r|5-C5H5)Fe+ (ED=6). This means that the C 5 H 6 ligand does not have a cyclopentadiene structure with r | 4 hapticity that would result in the electron deficiency ED=7 on the metal centre. Kan proposed that the C 5 H 6 ligand rearranges by an intramolecular 6-hydrogen transfer from cyclopentadiene structure to form hydrido-7t-cyclopentadienyl complex (reaction 1.2, where M=Fe). The 15-hydrogen transfer results in the increase of the coordination hapticity from T ) 4 to r | 5 + r\l which lowers the electron deficiency on the iron centre by two (i.e. to ED =5). Lower electron deficiency explains the observed lower reactivity of the C s ^ F e * ion. A similar rearrangement was proposed to explain a low reactivity of C 5 H 6 C r + , C 7 H 8 C r + , C 7 H 8 C r 2 ( C O ) + , C 6 H 6 OFe+ and C 6 H 6 O F e 2 + ions5. -4-The driving force for the 8-hydrogen transfer is high degree of electron unsaturation on the metal centre and resonance stabilization upon formation of aromatic or highly delocalized ligand. The C-H bond dissociation energy and the spatial position of hydrogen atom may also play part in the 8-hydrogen transfer process. As the above examples demonstrate, complexes with a metal-hydrogen bond are either products or intermediates in many gas phase organometallic reactions. It is therefore desirable to have at hand a reliable method that could be used to probe such a bond. So far, isotope exchange with deuterium 7 - 1 1 and ethene-d412, CID experiments11 and specific reactions with NH 3 1 3 > 1 4 have been used for this purpose but often they work only with partial success. In this research we will examine ion-molecule reactions of two organometallic ions with C1-C4 alcohols. The first ion is the (r|5-C5H5)Fe+ complex that does not have a metal-hydrogen bond. The second ion of interest is the (C5H6)Fe+ complex where a metal hydrogen bond is possible and was proposed to explain the unusually low reactivity of this complex with its precursor neutral5. We are interested if and how the reactions of these two ions with alcohols differ from each other. Further, Kan 5 proposed that the specific ion-molecule reactions with methanol be used to probe the metal-hydrogen character of organometallic complexes in the gas phase. We will investigate the potential of the ion-molecule reactions with ethanol and higher alcohols for probing the metal-hydrogen bond. -5-1.2 Gas Phase Ion-Molecule Reactions of Organometallic Ions with Alcohols Reactions of bare metal ions with alcohols and other monosubstituted alkanes have been reviewed by Eller and Schwarz 1 5> 1 6. The number of reports on the study of ligated metal cations 1 7- 1 8 and anions 1 9 - 2 1 is limited. 1.2.1 Methanol With the exception of A u + , late transition metal ions C r + 2 2 , M n + 2 3 , F e + 22> 2 4 ; C o + 2 5 , N i + 2 5 , C u + 2 6 > 2 7 and Ag+ 2 6 > 2 7 do not activate any bonds in methanol. The only reaction observed is condensation, reaction 1.4. M+ + alcohol -> M(alcohol)+ (1.4) M = Cr, Mn, Fe, Co, Ni, Cu, Au, Ag Because of a strong Au-H bond (D(Au-H) = 77.5 kcal/mol) and large ionization potential of gold, (IP(Au) = 212.7 kcal/mol) the dominant reaction of A u + 2 7 with methanol is hydride abstraction, reaction 1.5. In the earlier study Allison and Ridge 2 5- 2 8 observed F e O H + as the only ionic product of the reaction of Fe + with methanol. They proposed that F e O H + was formed by the insertion of the metal into a C-0 bond of alcohol. Evidence for this assumption was obtained using reactions 1.6 and 1.7. Au+ + C n H 2 n+iOH - » C n H 2 n O H + + AuH (1.5) FeCO+ + C H 3 O H - » FeCH 3OH+ + CO (1.6) FeCH 3OH+ + C D 3 O H -> FeOHCD 3OH+ + C H 3 (1.7) -6-Exclusive loss of CH3 in reaction 1.7 excludes a symmetrical collision complex of the type [(CH 3OH)Fe(CD 3OH)+]*. Thus, CH 3-Fe+-OH, rather than CH 3 OH-Fe+, is formed in reaction 1.6. In a later study Huang et al. 2 2 found Fe+ to be unreactive with methanol. It was concluded that the F e O H + ion observed by Allison and Ridge was due to a reaction of electronically excited F e + formed during the electron impact ionization of Fe(CO)5. Schwarz et al. studied the [Fe,C,H4,0] + isomers by using collisional activation mass spectrometry24. The dominant reactions in the collisional activation mass spectrum of [Fe,C,H4,0]+ isomer generated from Fe(CO)5/CH 3OH mixture by a 100-eV electron beam are due to the loss of C H 3 O H and the elimination of C H 3 . A weak, but structure-indicative signal at m/z 32 ( C H 3 O H + ) was observed as well. These findings indicate that the [Fe,C,H4,0] + isomer is an ion/molecule complex C H 3 0 ( H ) F e + , in which the intact methanol molecule serves as a ligand. The ions indicative of the insertion of the Fe + into a C-O bond of methanol were of very low intensity (<1% of base peak) rendering the CH 3-Fe+-OH structure less likely. It has been observed previously that the reactivity of metal ions with hydrocarbons changes upon ligation. This has also been demonstrated for the reactions of metal ions with alcohols. While bare Fe+ is unreactive with methanol, it was shown that the F e C H 3 + liberates methane via oxidative addition of the O-H bond to the metal centre17. The reactivity of metal ions also changes upon electronic, vibrational or translational excitation. The translationally excited C o + ions generated in the guided ion beam apparatus react with C H 3 O D to yield seven ionic products, formed in reactions 1.8-1.1429. Co+ + CH3OD -> C0OD+ + C H 3 (1.8) -> C0CH3+ + OD (1.9) -> CH 2OD+ + CoH (1.10) -> CoH+ + C H 2 O D (1.11) -> CoD+ + CH3O (1.12) Co+ + CH3OD -» C0OCH3+ + D (1.13) -> CoCH 2+ + HOD (1.14) At low energies, the products from the activation of the C-0 bond, reactions 1.8 and 1.9, dominate the spectrum. At energies above ~4 eV the most likely processes are reactions 1.10-1.11 and 1.12-1.14 that correspond to activation of the C - H and O-D bond, respectively. Unlike Cr+ that is unreactive with methanol, M o + dehydrogenates methanol to yield C H 2 0 M o + 2 2 . A correlation has been found between the first-row metal-ligand bond dissociation energy and a promotion energy from the 3d n configuration to the 3dn_74s^ configuration30. This correlation suggests that bonding between first-row transition metal ions and organic substrates involves s orbital electrons. C r + has the highest promotion energy thus it is relatively unreactive. Unlike first-row transition metals, bonding between the second-row transition metals and organic substrates is thought to occur through the interactions of d orbitals31. Second-row transition metals have larger d orbitals relative to their s orbitals than do first-row transition metals. Higher reactivity of M o + is therefore owing to its expanded-size d orbitals. Reactions of early transition metal ions are dominated by the metal insertion into the C-O bond of methanol, probably due strong metal-oxygen bonds. T i + is also able to break C-H and O-H bonds of methanol, although T i O H + , formed by the metal insertion into a C-0 bond, is the dominant product (over 50%)32. Similarly, the reactions of T a + with methanol afforded T a O H + and TaO + , resulting from the cleavage of the C-0 bond 3 3 ' 3 4 . Alkali metal ions Li+, Na+ and K+ 2 5 , Mg+ 35> 3 6 as well as A1+ 36> 3 7 , Ga+ and In+ 3 7 are unreactive towards methanol. A l + and M g + afford only a product of condensation, reaction 1.4. -8-1.2.2 Ethanol p e+ 22, 255 C o + 25> 3 8 and N i + 2 5 dehydrate ethanol by activating the C-O bond. Competitive loss of water or alkene generates products in reactions 1.15 and 1.16, respectively. M+ + C n H 2 n + i O H - » M ( C n H 2 n ) + + H 2 0 (1.15) M+ + C n H 2 n + i O H -> M ( H 2 0 ) + + C n H 2 n (1.16) M = Fe, Co, Ni, Rh, Au, A l For reaction of N i + with ethanol, also a small amount of dehydrogenation (-5%) was observed, reaction 1.17. M+ + C n H 2 n + i O H -> M(C n H 2 n O)+ + H 2 (1.17) M = Ni, Pd, Mo, Cu Second-row transition metal ions Pd + 3 9 and M o + 2 2 only dehydrogenate ethanol. M o + generates even a double dehydrogenation product Mo(C 2 H 2 0) + . Unlike bare Fe + that cleaves C-0 bond, the Fe(C4Ff6)+ ion, where C4H6 is butadiene, activates the O-H bond of alcohol1 8, reaction 1.18. This demonstrates again that not only the properties of M + but also the nature of the ligand determines the reactivity of the organometallic ion as well as the site of bond activation. Fe(C 4H 6)+ + C H 3 C H 2 O H - » Fe(C 4H 8)+ + CH3CO (1.18) High resolution FT-ICR experiments showed that R h + generates the isobaric RhC2H4+ and R h C O + ions, reactions 1.15 and 1.19, respectively40. Rh+ + C 2 H 5 O H -> RhCO+ + C H 4 + H 2 (1.19) -9-C u + 27> 4 1 mainly dehydrogenates ethanol, reaction 1.17, in addition to condensation, reaction 1.4. For reasons that are presented in the discussion of reaction 1.5, the dominant reaction of A u + 2 7 with ethanol is hydride abstraction. It was proposed that the hydride from the a-C is abstracted by metal ion. A u + also dehydrates ethanol, however, only the retention of alkene is observed, reaction 1.15. With use of C D 3 C H 2 O H and C2H5OD it was shown that dehydration involves 6-hydrogen transfer. Ag+ 2 7 > 4 2 , G a + 3 7 and In+ 3 7 and Mg+ 35> 3 6 undergo only condensation reactions with ethanol, reaction 1.4. A1+ 36> 3 7 dehydrates ethanol with the exclusive retention of the water ligand, reaction 1.16. Cr+ 2 2 , M n + 2 3 as well as alkali metal ions 2 5 are unreactive with ethanol. 1.2.3 n-Propanol Most reactions of transition metal ions with n-propanol are similar to those with ethanol. In the FT-ICR study2 2 Fe + was found mostly to dehydrate n-propanol (67%), reactions 1.15 and 1.16. The other product ions Fe(CH30H) + and Fe(C2H.50)+ were due to metal insertion into a C-C bond of alkyl chain2 2. The metastable ion (MI) decomposition spectra showed only products of dehydration with exclusive retention of alkene43. For C o + 38> 4 3 and N i + 4 3 loss of C H 3 O H and C 2 H4 was observed, besides dehydration. Labeling experiments showed that in M + mediated dehydration of n-propanol (M=Fe, Co, Ni) all C H bonds serve as a hydrogen source in the lost water molecule. The corresponding loss of C3H6 from C3H70H/M+ observed for C o + and N i + is accompanied by extensive hydrogen exchange processes. For C o + induced dehydration the data in MI spectra of n-propanol isotopomers show near-equivalent contribution of hydrogen from a- and B-methylene units. Based on these results it was proposed that the dehydration proceeds through the C o + insertion into a CB-Oy bond rather than through oxidative addition of C-0 bond (Scheme 1.2, 7->8). Subsequent insertion of ethylene unit in the -10-M+-CH3 bond generates 9 , rendering the original a- and 13-methylene groups indistinguishable. Ni +: 80% OH H (Y) / H,C Ni + \ / \(o) CH 2 CH2OH (6) Ni +: 20% Co+: 100% H,C \ M + — OH I I H,C CH, (7) H CH,OH \ / Ni + (7) I (6) H2C = CH 2 H3C OH \ / M + (fi) I (a) H2C = CH 2 8 w CH2(H)OH (6) (7) H2C = CH 2 (7) CH3OH (a) (B) H2C = CH 2 OH 10 ^ l l — M + M + \ OH H20 + C 3H 6 OH B-H 6(a) M+ a(B) Scheme 1.2 Proposed mechanism for generation of CH3OH, C2H4 and for dehydration of n-propanol. From ref. 43. In the n-C3H70H/Co + system lost ethylene molecule consists exclusively of a- and 6-methylene units and CH3OH is built up from the intact OH and CH3 groups. This is in -11-line with the proposed mechanism for the C o + mediated dehydration of C 3 H 7 O H . Complex 8 (Scheme 1.2) can be viewed as the immediate precursor for the detachment of C2H4 and C H 3 O H . The major fraction of ethylene and methanol in the n-C 3H70H/Ni+ system is generated by a cleavage of the Cot/CB bond of CH3CH2-CH2OH, followed by 8-hydrogen transfer from the ethyl group (Scheme 1.2, 5—>6). From a common intermediate 6 either C H 3 O H or C2H4 can be eliminated. Under FT-ICR conditions C r + and M o + dehydrogenate n-propanol, for M o + only the products of double and triple dehydrogenation Mo(C3H40) + and Mo(C3H20) + , respectively were observed22. In an ion-beam study a translationally excited C r + yielded besides the dehydrogenation product C r ( C 3 H 6 0 ) + also a Cr(C2H40)+ due to metal insertion into a C-C bond 4 4 . C u + mainly dehydrogenates n-propanol but Cu(H20) + and C u ( C 3 H 6 ) + as well as Cu(C 3 H 7 OH)+ were observed27. Ag+ 26> 27> 4 2 and Mg+ 35> 3 6 undergo only a condensation reaction, reaction 1.4, with n-propanol. Similarly, the predominant pathway of silver cluster ions A g x (x=l-5) with n-propanol, as well as j-propanol and ?-butanol, is adduct formation45. Initial addition of alcohol to Ag2+ and Ag4 + is accompanied by a displacement of silver atom to form Ag(n-PrOH) + and Ag3(n-PrOH)+, respectively. A l + reacts with C3 and C4 alcohols by hydroxide transfer, reaction 1.20, and by dehydration, reaction 1.15. H2O is always retained by A1+ in the dehydration36. M+ + C n H 2 n + i O H C n H 2 n + i + + MOH (1.20) M = Al, Au For Au+, AuH formed by hydride abstraction is the main product in addition to the loss of H2O from dehydration and AuOH from hydroxide abstraction27. -12-1.2.4 /-Propanol The dominant reaction of Fe+ 22> 2 5 , Co+ 2 5 , Ni+ 2 5 , Cr+ 2 2 , A1+ 36> 3 7 ' , Mg+ 3 6 , C u + 2 7 with /-propanol is dehydration, reactions 1.15 and 1.16. The M ( C 3 H 6 ) + and M ( H 2 0 ) + ions are also observed for A u + along with the products of hydride and methanide abstraction AuH and A U C H 3 , respectively27. Fe + 2 5 , C o + 2 5 and A1+ 3 6 also induce hydroxide abstraction generating M O H neutral and C3H7+. Ag+ 2 6 , Cr+ and M o + 2 2 mainly dehydrogenate i-propanol, M o + even doubly and triply. Some dehydrogenation is observed for C u + 2 7 , as well. G a + , In + 3 7 and L i + 2 5 yield only the adduct ion M(i-C3H70H)+, that is also one of the dominant products for C u + 2 7 . Beauchamp et al.46 and later Allison and Ridge 2 5 studied the reactions of L i + , Na+ and K+ and Fe+, C o + and N i + with alkyl halides and alcohols. For the reaction of alkali metal ions they observed that the reactivity of R-X is related to the enthalpy change, AHexch> for the halide (hydroxide) exchange process, reaction 1.21 M+ + R-X -> R+ + M X AHexch (1.21) This correlation between reactivity and A H e x c h suggests a mechanism that involves transfer of halide (hydroxide) to the metal cation and generation of charge at the carbon (Scheme 1.3). That is, initial association of an alkali ion and organic halide or alcohol leads to the formation of a chemically activated species 11 which may either dissociate to an M X and carbonium ion or may rearrange to 12 with subsequent competitive ligand loss to afford 13 and 14. The energy necessary to move X (X=halide, OH) away from the charged carbon centre is the source of the energetic barrier to reaction. This mechanism was later termed ion/dipole mechanism. -13-x M + + M 4 11 I H H -AA/^ » + MX X H M + 12 I M ( H X ) + 13 + >=< "1 M + I >=< 14 + H X Scheme 1.3 Ion/dipole mechanism for the reaction of alcohols with metal ions. From ref.16 However, for the reactions of transition metal ions Allison and Ridge25 proposed a mechanism that involves metal insertion into the R-X bond and subsequent 8-H transfer (Scheme 1.4). The essential feature of this mechanism is the absence of any significant energetic barrier between 15 and 16. x H M + + \ M + 15 X H \ / M + I >=< 16 M + I > = < + H X 17 M ( H X ) + + 18 Scheme 1.4 Metal insertion/8-hydrogen transfer mechanism for the reaction of alcohols with metal ions. From ref.25 In a recent review, it was suggested that the observed M(C3H6)+ and M(H20)+ ions for the dehydration of n-, i-propanol are in fact formed by the ion/dipole mechanism rather than by the insertion/8-hydrogen transfer mechanism16. This is based on the fact that the same product ions are observed for Fe+, Co+ and Ni+, the ions that have insertion -14-ability, as well as for C r + , M o + and group 11 ions, all of which have limited insertion ability because of d 5 or d 1 0 electronic configuration. 1.2.5 C4 and Higher Alcohols Under FT-ICR conditions Fe+-mediated dehydration of n-butanol prevails over the dehydrogenation22. With the increasing chain length, products from the insertion into a C-C bond begin to dominate the spectra. For example for n-hexanol, they constitute 90% of the product distribution, the remaining 10% is formed by dehydrogenation, reaction 1.17, and dehydration is absent all together22. In the MI spectra Fe +-mediated dehydrogenation of C4-C8 alcohols constitutes the dominant decomposition mode of the ROH/Fe+ complex47. Tsarbopoulos and Allison 3 8 studied the reactions of C o + with 1-alkanols with the alkyl chain length varying from three to eight carbon atoms. They observed that as the alkyl chain length increases, insertion into the C-OH bond and subsequent dehydration becomes a minor pathway and attack of C-C bonds dominates. To account for the observed products (i.e. loss of alkanes, alkenes and alcohols from C n H 2 n + i O H / C o + , where n>6) they proposed a formation of five- or six-membered ring intermediates upon the initial complexation of Co+ with the OH group. This intermediate brings specific skeletal carbon atoms into close proximity with the metal ion. The resulting C o + - C interactions weaken the adjacent C-C bonds making them susceptible to insertion. The formation of the five- and six-membered ring intermediates along with all of the possible insertion intermediates and the resulting products are shown in Scheme 1.5 for 1-hexanol. The hydrogen formation is explained by hydrogen transfers from both the OH group and the "exocyclic" part of the alkyl chain. -15-n-C6H1 3OH Co+ C 6 H 1 3 - C o + - O H CH 3 CH 2 ^ ~ C o + H2C J OH r 1 H2C CH 2 \ / CH 2 C 2 H 5 - Co+ - (CH2) 4 OH C3H7 ~r - C o + H2C J OH — \ / H2C — CH 2 C 3 H 7 - Co+ - C 3H 6OH C 6 H 1 2 + C0(H2O)+ C 2 H 6 + CoC 4H 7OH+ C 2 H 4 + CoC 4H 9OH+ H 2 + Co(C4H7OH)(C2H4)4 C 3 H 6 + CoC 3H 7OH+ C 3H 7OH + CoC 3 H 6 + "C 4 H 9 - Co+ - C 2H 4OH C 4 H 8 + CoC 2H 5OH+ Scheme 1.5 Metal insertion into a C-C bond of 1-hexanol and formation of cyclic intermediates. From ref. 38. The MI study of labeled 1-pentanols47 and 1-hexanol48 led to a revision of the previously proposed mechanisms for the transition-metal ion mediated formation of hydrogen, alkanes and alkenes in the gas phase. Huang et al.22 proposed that dehydrogenation involves an initial metal insertion into an O-H or unspecified C-H bond. However, no labeling experiments have been done to substantiate this mechanism. In a later study Priisse and Schwarz 4 7 showed that in the reaction of F e + and C o + with n-pentanol hydrogen is formed nearly exclusively from the 00- and (GO - Impositions of the pentyl chain. The OH group does not serve as a hydrogen donor. The reaction follows a 1,2-elimination and can be described in terms of a mechanism called "remote functionalization". In this mechanism the C-H bond of a terminal C H 3 group of a flexible alkyl chain can be oxidatively added to the transition-metal ion M + which is "anchored" at -16-the OH group (Scheme 1.6, 19—»20). The insertion is followed by a 8-hydrogen shift (20—>21) or B-cleavage of the C-C bond (20—>22) to generate intermediates from which eventually reductive elimination of H 2 or alkene detachment occurs. M + H \ H O C H 2 I C H 2 ( C H 2 ) „ 19 J I M + / \ H O C H 2 C H , ( C H 2 ) N -20 J H H \ / M + R H O | | C H , ^ ( C H 2 ) ^ 21 H 2 C H ^ \ / C H 2 M + / \ H O C H , 22 C , H , 2 "4 Scheme 1.6 "Remote functionalization" mechanism, n>2. From ref. 47 The MI spectra of the C o + complex of labeled n-hexanol 4 8 showed that the hydrogen is formed by a "remote functionalization", in sharp contrast to the previously proposed C-C insertion/8-hydrogen transfer mechanism38 (Scheme 1.5). It was also demonstrated that the loss of C3H6 moiety (as described in Scheme 1.5) corresponds in fact to the consecutive loss of C2H4/H2. The dominant reaction of C u + 2 7 with n-, i- and r-butanol is dehydration, reactions 1.15 and 1.16 and condensation, reaction 1.4. C u + dehydrogenates only n-butanol. Small amounts of CuOH generated by hydroxide abstraction were observed for i- and f-butanol, and Q1CH3 formed by methide abstraction, reaction 1.22, for f-butanol. -17-M+ + C 4 H 9 O H - » C 3 H 6 O H + + M C H 3 (1.22) M = Cu, Au A u + 2 7 reacts with C4 alcohols mainly by hydride abstraction (except for f-butanol) and by hydroxide abstraction, in addition to dehydration and dehydrogenation. For the reaction of A u + with r-butanol, methide abstraction is important as well. A g + 2 7 gives only adduct ions Ag(C4H90H)+ with n-butanol. No products were observed in Ag+/i-butanol system, f-butanol yields mainly the dehydration product Ag(C4Hs)+. A l + 3 7 generates A1(H.20)+ and AlOH with n-butanol. With f-butanol only a product of hydroxide abstraction AlOH was observed. G a + and In + 3 7 form only adduct complexes with n- and r-butanol. n-, and j'-butanol cluster with L i + and N a + 2 5 at higher pressures. L i + dehydrates r-butanol25> 4 9 . M g + 3 5 ' 3 6 cationizes n-butanol and dehydrates /-, and f-butanol. Several mechanisms have been proposed in the literature to explain the reactions of alcohols with transition-metal ions. For smaller alcohols, n<4, competitive loss of H2O and alkene seem to be the most important process. Their formation was rationalized by invoking either a metal insertion into a C-0 bond/8-hydrogen transfer mechanism25- 2 8 (Scheme 1.4) or by ion/dipole mechanism16 (Scheme 1.3). The controversy over which mechanism is in fact operative, has not yet been fully resolved. For higher alcohols (n>4) the "remote functionalization" mechanism 4 7- 4 8 (Scheme 1.6) has been proposed to explain the generation of H2 and ethene (or higher alkenes). However, to account for the loss of smaller alcohol neutrals from the R O H / M + system a direct insertion of the metal ion into a C-C bond has to be invoked4 7 (Scheme 1.5). -18-1.3 Experimental Section 1.3.1 Operating Principles of FT-ICR and Experimental Techniques Since its introduction in 197450- 5 1 , Fourier Transform Ion Cyclotron Resonance Mass Spectrometry (FT-ICR MS) has found application in a wide variety of fields. These areas include the study of reaction pathways and ion/molecule reaction chemistry15, mass measurement of short-lived isotopes, sequence analysis of peptides and proteins, study of drugs, metabolites and organometallic compounds, analysis of semiconductor and polymer surfaces. The many advantages of FT-ICR MS include high resolution and mass accuracy, ability to analyze high molecular weight compounds, simultaneous ion detection, ion trapping and manipulation by double and triple resonance techniques, positive/negative-ion detection by simple software changes, adaptability to a wide range of ionization techniques, multiple MS experiments. The usefulness of FT-ICR MS as an analytical technique for solving real-world problems is evident from an exponential increase in the number of FT-ICR spectrometers used world-wide. Several literature reviews5 2"5 4 and a book 5 5 on FT-ICR theoretical aspects, techniques and applications were published recently. The FT-ICR approach is distinctive in that during the detection, the ions are not spatially separated and collected individually but rather their cyclotron motion is detected while they are trapped in the ICR cell (Figure 1.1). The ICR cell consists of three pairs of parallel electrodes. A direct voltage is applied to two trapping plates that are perpendicular to the magnetic field in order to confine the ions within the cell boundaries. Through the transmitter plates a rf electric field is applied to excite the ion cyclotron motion. This cyclotron motion is then detected by the receiver plates. The cell is contained within a high vacuum chamber which is centered in a homogeneous magnetic field. Various designs of ICR ion traps have been reviewed recently 5 6- 5 7. -19-Figure 1.1 Cubic ICR cell 5 8 . The trapped ion undergoes large-amplitude cyclotron motion in the xy plane and also oscillates along the z axis in the electrostatic potential well. An ion of charge q moving in a static magnetic field B = fik with velocity v is subjected to a Lorentz force F L = <?vB. The force restricts the ion's motion in the plane perpendicular to the magnetic field B to a circular cyclotron orbit. The Lorentz force is counterbalanced by the outward-directed centrifugal force F c = mv 2/r. A stable orbit results when the two forces are equal as shown in equation (1.23) qBx\ = m\2/r (1.23) Using \/r = ca we can write the cyclotron equation (1.24), where coc is the ion cyclotron frequency (Oc = q Blm (1.24) All ions of a given mass-to-charge ratio have the same cyclotron frequency, independent of their initial velocity. This eliminates the need for translational energy 'focusing'. Through the cyclotron equation the ionic mass-to-charge ratio, m/q, is effectively converted to a cyclotron frequency, that can be measured with high accuracy. The ICR frequencies for -20-ions in the usual mass range (15- 10 000 amu) typically lie between a few kHz and a few MHz and are in the radio frequency region. The measurement of cyclotron frequencies is accomplished in two steps: excitation of cyclotron motion and detection of excited cyclotron motion59 (Figure 1.2). a) b) I(t) resonant ion co=coc Figure 1.2 FT-ICR ion excitation and detection a) Excitation of ion cyclotron motion by an oscillating rf electric field occurs when the resonant condition, co=coc, is achieved b) Detection of the excited cyclotron motion During the ionization event ions of the same m/q and initial velocity are created with random phase. Their incoherent motion cannot produce a detectable signal. The purpose of the excitation of the cyclotron motion is to create coherent motion of the ions and to increase the radius of the resulting ion packet. The increase of the ion cyclotron radius improves the sensitivity of measurement. The radius of the ion orbit is given by equation 1.2559 r = Vpt/co d B (1.25) -21-where V p is the peak value of the excitation voltage, t is the duration of the pulse and d is the separation between the two transmitter plates. The equation shows that the radius of the ion orbit is independent of the ion mass. When an alternating rf electric field with frequency co is applied to the transmitter plates, the energy is transferred from the rf oscillator to the kinetic energy of the ions when the resonance condition co=C0c is reached. The radius of the ion orbit increases and the ions are excited spirally outward. This is the principle of ion cyclotron resonance. Ions excited by a resonant rf electric field to an orbit that is large (~1 cm) but still within the cell boundaries will be detected with optimal sensitivity. By increasing the duration or the amplitude of the excitation voltage the radius of the ion packet can be increased to such extent that these ions will collide with one of the transmitter or receiver plates and be absorbed by them. This is the basis of the ion ejection technique60. During the detection event the coherently moving packet of ions of the same mass-to-charge ratio induces a charge on the receiver plates. An alternating current results with a frequency equal to the cyclotron frequency of the ions. In practice, the current signal is converted to a voltage signal V m by the circuit displayed in Figure 1.2. The voltage is given by equation 1.2659 Vm(t) = N q r/d C cos(cot) (1.26) where N is the number of ions, C is the circuit capacitance, co is the cyclotron frequency of the ion and t is time. The signal voltage is proportional to the number of ions and the radius of the excited cyclotron motion. The presence of ions of different mass-to-charge ratio gives rise to a composite voltage signal V(t) V(t) = l V m ( t ) . (1.27) which is then amplified, digitized and Fourier transformed. -22-FT-ICR MS is a pulsed technique. A typical pulse sequence is given in Figure 1.3. Ion Formation Ion excitation Ion Ion (rf sweep) detection quench DL3 - n — Repeat Ion Formation Ion ejection 1 (rf sweep) D L l - n -Single Resonance Ion excitation Ion Ion (rf sweep) detection quench DL3 -n — Repeat Double Resonance Ion Formation Ion Ion ejection 1 ejection 2 (rf sweep) (rf sweep) Ion acceleration (fixed rf) Ion excitation Ion Ion (rf sweep) detection quench D L l - U -DL2 — 11— DL3 - 11 -Repeat Collision-Induced Dissociation Figure 1.3 FT-ICR experimental pulse sequence of events for single resonance, double resonance (DBL), and collision-induced dissociation (CID) experiments. The initial step of the experiment is the production of gas phase ions from the sample. Basically all ionization methods have been adapted for use with FT-ICR MS. These include electron ionization, chemical ionization, matrix assisted laser desorption/ionization MALDI, laser desorption/ionization L D , fast atom bombardment FAB, electrospray ionization ESI and plasma desorption PD. In some cases the conflicting pressure requirements for ion formation and ion detection have been reconciled by -23-employing external ion sources or differentially pumped dual trap ion cells6 1. The dual cell also permits efficient coupling of gas chromatography to FT-ICR MS by allowing the chromatographic effluent to be introduced directly and continuously . The variable delay time allows for ion-molecule reactions or insertion of other pulses. The delay is set to zero for normal mass spectra. As explained above, generation of a detectable signal requires that ions be excited coherently to a suitable ion cyclotron radius. The goal is to excite all ions in the mass spectral range of interest to the same cyclotron orbit radius to give a flat spectrum without mass discrimination. Several approaches to ion excitation have been used with varying degree of success. In the impulse excitation a single-frequency pulse of finite width and height is used. This waveform is unsuitable for broadband and/or highly selective excitation because of its 'sine' shape and a requirement for voltage amplitudes of several thousand kilovolts. The chirp or frequency sweep excitation51 is the most common excitation method. In this case the frequency is swept linearly over the range from the lowest to the highest frequencies desired in the spectrum. The advantages of chirp excitation are its rather simple implementation and the ease with which a wide mass range can be excited without the need for large rf amplitudes and expensive amplifiers. In special applications, when ions are to be excited or ejected over several different mass ranges, the Stored Waveform Inverse Fourier Transform (SWIFT) 6 2 excitation gives the most satisfactory control over the excitation characteristics. Its average mass selectivity is better than that of the frequency sweep because the SWIFT time-domain waveform is turned on and off only once, whereas the frequency sweep must be turned on and off N times for excitation over N distinct mass ranges. The detection event has been described above. At the end of the experiment the ions are removed from the cell by applying a differential voltage to the trapping electrodes. In this work we will use two extremely powerful features of FT-ICR that allow manipulation of the ions. These are ion ejection techniques of double (DBL) and triple -24-(TPL) resonance to infer the reaction pathways and collision-induced dissociation (CID) to study the structure of the product ions. The pulse sequence of a D B L experiment is schematically illustrated in Figure 1.3. Following the ionization event a high power radio frequency sweep is applied which excites the cyclotron motion of the ions in the selected mass range. The radius of their orbit increases until they collide with the walls of the cell and thus are lost. The ion of interest can then be excited by another rf sweep or pulse and detected. If two rf sweeps are used before the ion excitation the sequence is called triple resonance. In a CID experiment (Figure 1.3) the ion of interest is first isolated by double or triple resonance. It is then accelerated and allowed to pass through a collision gas. The energy imparted to the ion by collisions induces fragmentation. The ionic fragments are then mass analyzed. 1.3.2 Experimental Parameters All experiments were performed on a home-built instrument controlled by a Nicolet FTMS-1000 program from a computer console. A 2.54 cm cubic ICR cell, placed between the poles of electromagnet with magnetic field intensity 1.9 Tesla, was used as the ion trap and reaction cell. The trapping voltage was kept at +1V. Background pressure was maintained at about lxlO"9 Torr by an ion pump backed up by Edwards Diffstack diffusion pump. The samples were introduced through a leak valve. Liquid samples were submitted to several freeze-pump-thaw cycles before introduction to the vacuum system in order to remove non-condensable gases. The organometallic neutrals were kept at a pressure lx lO- 7 Torr and alcohols at a pressure about lxlO" 6 Torr. The pressure in the vacuum chamber was measured by a Bayard-Alpert ionization gauge (Varian model 971-1008). Because the Bayard-Alpert gauge has a selective response to different gases, the pressure of alcohols was corrected for this different sensitivity according to Bartmess and Georgiadis63. For the CUD experiments nitrogen, introduced through a pulsed valve, was -25-used as a collision gas. All neutral molecules were used as supplied. (CsHs^Fe and ( C 6 H 6 0 ) F e ( C O ) 3 , C H 3 O H , C H 3 O D , C D 3 O H , C D 3 O D , C 2 H 5 O H , C 2 H 5 O D , C D 3 C H 2 O H , C H 3 C D 2 O H , C 2 D 5 O D , n-, i - C 3 H 7 O H , i - C 3 H 7 O D , ( C D 3 ) 2 C H O H , / - C 3 D 7 O D , n-, i-, ? -C 4 H 9 OH, n -C 4 H 9 OD, r -C 4 H 9 OD and f -C 4 D 9 OD were purchased from Aldrich Chemical Co.. n - C 3 H 7 O D and n - C 4 D 9 O H were obtained from Cambridge Isotope Laboratories Inc.. The mass spectra were calibrated using the following procedure. First, the mass spectrum was calibrated using only the calculated molecular masses for major isotopic peaks of known reactant ions, i.e. (C5H 5 ) 5 4 Fe+, (C5H 5 ) 5 6 Fe+, (C5H6) 5 4Fe+ and (C5H6) 5 6 Fe + , and the product ions of self ion-molecule reactions of alcohol. In most cases, due to a high partial pressure of alcohol, the products of clustering reactions of alcohol ( C n H 2 n + i O H ) 2 H + , and ( C n H 2 n + i ) 2 O H + were observed in the spectra at longer delay times. The calculated masses of these ions were also used for the calibration. In this way the masses of nearest neighbouring product ions could be calibrated and their formulae determined. These new masses where then included into a calibration table and the mass spectrum was recalibrated. The whole procedure was repeated for ions with higher m/z ratio until all the peaks in the spectrum were calibrated. 1.3.3 Basic Kinetics of Gas Phase Ion-Molecule Reactions At a pressure around lx lO - 7 Torr and higher the concentration of neutral molecules is about thirty thousand times higher than that of the ions. The general ion-molecule reaction M+ . + A products (1.28) therefore follows a pseudo-first order rate law. The experimental rate constant k can be calculated from the slope of the plot of log(normalized intensity) vs. reaction time for the reactant ion M+. The reaction time equals the DL3 (see Figure 1.3), a delay between the -26-isolation of a reactant ion and the detection of product ions, that is gradually increased. In our experiments the situation is slightly complicated by the presence of two types of neutrals, the organometallic neutral, from which the reactant ion is formed and the alcohol. Therefore, the decay of the reactant ion M+ (M=CpFe, CsHgFe) is not only due to the reaction with alcohol but also due to a simultaneous self ion-molecule reaction with ferrocene or with (r)4-C6H60)Fe(CO)3. Although the extent of the interfering self ion-molecule reactions is minimized by introducing the alcohol at a pressure ten times higher than that of organometallic neutral, it still has to be taken into consideration. For the overall reactions M+ •+ ROH -> products (1.29) M+ + OMN -> products (1.30) where M=CpFe or C s ^ F e and organometallic neutral OMN=Cp2Fe or (r)4-C6rl60)Fe(CO)3, respectively, the rate for the decay of the reactant ion M+ can be expressed as _WLA = kROH[M+][ROH] + ksimr[M+][OMN] (1.31) at where k R Q H is the rate constant for the reaction 1.29 and k^mr is the rate constant for the self ion-molecule reaction 1.30. After integration this gives \nlMA = _k t (132) [M+]0 ohs where k o b s (1.33) -27-is the slope of the graph log[M +]/[M +]o vs. reaction time. The [M +]o and [M+] are normalized intensity of the reactant ion at a zero reaction time and at a particular time DL3, respectively. To minimize the errors caused by intensity measurement fluctuations, the relative intensities of ion peaks ( J 4 , 1 2 , ..Ii) in a spectrum were converted to normalized intensities given by nomalized intensity of ion i = (1-34) From the equation 1.33, kROH c a n D e expressed as k t m = - j ^ . k j ™ m (1.35) R0H [ROH] s,mr[ROH) As can be seen from equation 1.35, the determination of the reactivities of the metal ions with the various alcohols required several experiments. First, the rate constant ksimr for the disappearance of M + as a function of time was measured for each metal ion in reaction with its precursor neutral (no alcohol present). The values of ksimr were found to be 2.6 x 10"10 cm3molecule_1s"1 for self ion-molecule reactions of CpFe + with ferrocene and 6.7 x 10"10 cm 3molecule _ 1s" 1 for self ion-molecule reactions of CsF^Fe"1" with (r|4-C6H60)Fe(CO)3. In a separate experiment, the rate constant for disappearance of M+ was measured for each metal ion in reaction with a given alcohol and with its neutral precursor. The difference between the rate constants in the above two experiments for each metal ion was taken as the experimental rate constant kROH for the metal ion/alcohol reaction. Determination of the rate constant for reactions 1.29 and 1.30 requires a knowledge of partial pressure of neutral species. This pressure was measured by a Byard-Alpert gauge and corrected for different sensitivities of alcohols according to Bartmess and Georgiadis63. -28-References (1) Allison, J.; Freas, R. B.; Ridge, D. P. Journal of American Chemical Society 1979,101, 1332-1333. (2) Jacobson, D. B.; Freiser, B. S. Journal of American Chemical Society 1983,105, 7492-7500. (3) Byrd, G. D.; Freiser, B. S. Journal of American Chemical Society 1982, 104, 5944-5949. (4) Jacobson, D. B.; Freiser, B. S. Journal of American Chemical Society 1985,107, 72-80. (5) Kan, Z . , University of British Columbia, Vancouver, 1993. (6) Russell, D. H.; Fredeen,D. A.; Tecklenburg, R. E . In Gas Phase Inorganic Chemistry; Russell, D. H., Ed.; Plenum Press: New York, 1989, pp 117-134. (7) Byrd, G. D.; Freiser, B. S. Journal of American Chemical Society 1982, 104, 5944-5950. (8) Carlin, T. J.; Sallans, L . ; Cassady, C. J.; Jacobson, D. B.; Freiser, B. S. Journal of American Chemical Society 1983,105, 6320-6321. (9) Jacobson, D. B.; Freiser, B. S. Journal of American Chemical Society 1983,105, 7492-7500. (10) Jacobson, D. B.; Freiser, B. S. Journal of American Chemical Society 1985,107, 5876-5883. (11) Jacobson, D. B. Organometallics 1988, 7, 568-570. (12) Jacobson, D. B.; Freiser, B. S. Journal of American Chemical Society 1985,107, 72-80. (13) Buckner, S. W.; Freiser, B. S. Journal of American Chemical Society 1987,109. (14) Buckner, S. W.; Freiser, B. S. Polyhedron 1989, 8, 1401-1406. (15) Eller, K.; Schwarz, H. Chemistry Reviews 1991, 91, 1121-1177. -29-(16) Eller, K. Coordination Chemistry Reviews 1993,126, 93-147. (17) Blum, O.; Stockigt, D.; Schroder, D.; Schwarz, H. Angew. Chem. Int. Ed. Engl. 1992,37, 603-604. (18) KarraB, S.; Schroder, D.; Schwarz, H. Chem. Ber. 1992,125, 751-756. (19) Buckner, S. W.; VanOrden, S. L . Organometallics 1990, 9, 1093-1097. (20) McDonald, R. N.; Jones, M . T. Organometallics 1987, 6, 1991-1992. (21) McElvany, S. W.; Allison, J. Organometallics 1986, 5, 416-426. (22) Huang, S.; Holman, R. W.; Gross, M . L . Organometallics 1986, 5, 1857-1863. (23) Uppal, J. S.; Staley, R. H. J. Am. Chem. Soc. 1982,104, 1238-1243. (24) Schroder, D.; Fiedler, A.; Hrusak, J.; Schwarz, H. / . Am. Chem. Soc. 1992, 114, 1215-1222. (25) Allison, J.; Ridge, D. P. J. Am. Chem. Soc. 1979,101, 4998-5008. (26) Jones, R. W.; Staley, R. H. / . Phys. Chem. 1982, 86, 1669-1674. (27) Weil, D. A.; Wilkins, C. L . J. Am. Chem. Soc. 1985,107, 7316-7320. (28) Allison, J.; Ridge, D. P. J. Am. Chem. Soc. 1976, 98, 7445-7447. (29) Chen, Y . - M . ; Clemmer, D. S.; Armentrout, P. B. J. Am. Chem. Soc. 1994,116, 7815-7826. (30) Armentrout, P. B.; Halle, L . F.; Beauchamp, J. L . J. Am. Chem. Soc. 1981, 103, 6501-6502. (31) Mandich, M . L . ; Halle, L . F.; Beauchamp, J. L . J. Am. Chem. Soc. 1984,106, 4403-4411. (32) Guo, B. C ; Derns, K. P.; A. W. Castleman, J. J. Phys. Chem. 1992, 96, 4879-4883. (33) Wise, M . B.; Jacobson, D. B.; Freiser, B. S. J. Am. Chem. Soc. 1985, 107, 1590-1595. (34) Wise, M . B.; Jacobson, D. B.; Freiser, B. S. J. Am. Chem. Soc. 1985, 107, 6744. -30-(35) Operti, L . ; Tews, E. C ; Freiser, B. S. J. Am. Chem. Soc. 1988, 110, 3847-3853. (36) Uppal, J. S.; Staley, R. H. / . Am. Chem. Soc. 1982,104, 1229-1234. (37) Chowdhury, A. K.; Wilkins, C. L. International Journal of Mass Spectrometry and Ion Processes 1988, 82, 163-176. (38) Tsarbopoulos, A.; Allison, J. J. Am. Chem. Soc. 1985,107, 5085-5093. (39) Tolbert, M . A.; Beauchamp, J. L . J. Phys. Chem. 1986, 90, 5015-5022. (40) Gord, J. R.; Freiser, B. S. Analytica Chimica Acta 1989, 225, 11-24. (41) Jones, R. W.; Staley, R. H. J. Phys. Chem. 1982, 86, 1669-1674. (42) McLuckey, S. A.; Schoen, A. E. ; Cooks, R. G. J. Am. Chem. Soc. 1982, 104, 848-850. (43) KarraG, S.; Prusse, T.; Eller, K.; Schwarz, H. J. Am. Chem. Soc. 1989, 111, 9018-9023. (44) Kang, H.; Beauchamp, J. L . J. Am. Chem. Soc. 1986,108, 7502-7509. (45) Sharpe, P.; Cassady, C. J. Chemical Physic Letters 1992,191, 111-116. (46) Wieting, R. D.; Staley, R. H.; Beauchamp, J. L . / . Am. Chem. Soc. 1975, 97, 924-926. (47) Prusse, T.; Schwarz, H. Organometallics 1989, 8, 2856-2860. (48) Prusse, T.; Allison, J.; Schwarz, H. International Journal of Mass Spectrometry and Ion Processes 1991, 107, 553-557. (49) Creasy, W. R.; Farrar, J. M . J. Phys. Chem. 1985, 89, 3952-3955. (50) Comisarow, M . B.; Marshall, A. G. Chemical Physics Letters 1974, 25, 282-283. (51) Comisarow, M . B.; Marshall, A. G. Chemical Physics Letters 191 A, 26, 489-490. (52) Amster, I. J. Jounal of Mass Spectrometry 1996, 31, 1325-1337. -31-(53) Marshall, A. G.; Schweikhard, L. International Journal of Mass Spectrometry and Ion Processes 1992,118/119, 37-70. (54) Wanczek, K. P. International Journal of Mass Spectrometry and Ion Processes 1989, 95, 1-38. (55) Asamoto, B. FT-ICR/MS: analytical applications of Fourier transform ion cyclotron resonance mass spectrometry; V C H Publishers: New York, N.Y., 1991. (56) Guan, S.; Marshall, A. G. International Journal of Mass Spectrometry and Ion Processes 1995,146/147, 261-296. (57) Grosshans, P. B.; Shields, P. J.; Marshall, A. G. Journal of Chem. Phys. 1991, 94, 5341-5352. (58) Comisarow, M . B. International Journal of Mass Spectrometry and Ion Physics 1981, 37, 251-257. (59) Comisarow, M.B.J. Chem. Phys. 1978, 69, 4097-4104. (60) Comisarow, M . B.; Grassi, V.; Parisod, G. Chemical Physics Letters 1978, 57, 413-416. (61) Honovich, J. P.; Markey, S. P. International Journal of Mass Spectrometry and Ion Processes 1990, 98, 51-68. (62) Marshall, A. G.; Wang, T. C ; Ricca, T. L. Journal of American Chemical Society 1985, 107, 7893-7897. (63) Bartmess, J. E.; Georgiadis, R. M . Vacuum 1983, 33, 149-153. - 3 2 -Chapter 2 Gas Phase Ion-Molecule Reactions of Cyclopentadienyliron Ion (T | 5 -C 5 H 5 )Fe+ with Alcohols When (C5H6)Fe+ is formed by electron ionization of cyclohexadienoneiron tricarbonyl (r|6-C6H60)Fe(CO)3, along with the ironcyclopentadiene fragment (m/z 122), an ironcyclopentadienyl ion (C5H5)Fe+ (m/z 121), is formed as well. Because the two fragment ions differ only by one mass unit the ironcyclopentadiene ion cannot be isolated completely without some of the ironcyclopentadienyl ion being present in the cell. This complicates the interpretation of spectra. We therefore decided to first study the gas-phase ion molecule reactions of the ironcyclopentadienyl ion with alcohols and then compare the two spectra. This will also allow us to compare how the reactivity of transition metal ions changes when the ligand on the metal centre is varied. Ferrocene was employed as a source of pure ironcyclopentadienyl ions. A 3 0 eV x 2 0 ms electron ionization of ferrocene at a pressure 1.2 x 10" 7 Torr generates three fragment ions, Fe + m/z 56 , (C5H5)Fe+ m/z 121 and (C5H5)2Fe+ m/z 186. For simplicity in writing, Cp is used for a cyclopentadienyl ligand (C5H5). The CpFe + was subsequently isolated by a triple resonance experiment and allowed to react with alcohol. The alcohol was introduced simultaneously with the ferrocene and kept at a pressure about ten times higher in order to suppress self ion molecule reactions of ferrocene. The observed ion-molecule reactions and product distributions for CpFe + with methanol, ethanol, n-, i-propanol and n-, i-, f-butanol are summarized in Table 2 .1 . The numbers in Table 2.1 were determined from the ratio plots for individual alcohols by extrapolating the normalized ion intensity to reaction time zero. -33-alcohol ion formed m/z . T reaction neutrals ion no. no. 153 (2.3) 3 167 (2.6) 4 139 C 2H4 (2.7) 8 149 H 2 0 (2.8) 9 147 H 2 0 + H 2 (2.9) 10 165 H 2 (2.10) 14 139 C 3 H 6 (2.11) 8 163 H 2 0 (2.12) 15 181 (2.13) 16 139 C 3H6 (2.17) 8 163 H 2 0 (2.18) 15 179 H 2 (2.19) 23 181 (2.20) 24 139 C 4 H 8 (2.24) 8 175 H 2 0 + H 2 (2.25) 26 139 C 3H6 (2.26) 8 175 H 2 0 + H 2 (2.27) 28 177 H 2 0 (2.28) 27 139 C 4 H 8 (2.29) 8 177 H 2 0 (2.30) 27 175 H 2 0 + H 2 (2.31) 28 CpFe + + C H 3 O H CpFe + + C 2 H 5 O H CpFe + + n - C 3 H 7 O H 1 Q Q \ » CpFe(CH 3OH) + 3 8 % - CpFe(C 2 H 5 OH) + - ^ t CpFe(H 20) + ^ CpFe(C 2H4)+ CpFe(C 2 H 2 ) + CpFe(OC 2H4)+ CpFe(H 2 0) + CpFe(C 3 H 6 ) + CpFe(C 3 H 7 OH) + 13% 12% 65% 22% 13% CpFe + + i-GjF^OH f2% CpFe(H 20) + 2 ^ CpFe(C 3 H 6 ) + 10% ^ . . . T , . ™ ™ CpFeT CpFe + + / - C 4 H 9 O H pFe( 3 6 ) + CpFe(OC(CH 3)2) + 7% 1 • CpFe(C 3 H 7 OH) + CpFe(H20)" 45% _ : + + n - Q F ^ O H i 5 5 ^ pFe(H 20) +  4. CpFe(C 4 H 6 ) + C pFe(H 2 0) + ^ CpFe(C 4 H 6 ) + 15% CpFe ^ p r c ^ 4 n 6 ; 1 CpFe(C 4H 8)H ?-C 4 H 9 OH | 7 0 y V CpFe(H 2 0) + CpFe(C 4H g)-3% ^ _ T , _ ^ TT V • CpFe(C 4H 6)-Data are expressed in %Xfragments = 100%. Product percentages are reproducible to ±10% of the values reported. The total pressure is kept at 1.2 x 10*6 Torr. - 3 4 -In all reaction systems studied, in addition to the products from the reaction of CpFe + with alcohol, products from self ion-molecule reactions of CpFe + with ferrocene1 were observed as well (reaction 2.1 and 2.2). CpFe+ + CpFe2 *~ Cp2Fe+ + CpFe (2.1) m/z 121 1 m/z 186 CpFe+ + CpFe 2 ^ C p 3 F e 2 + (2.2) m/z 121 2 m/z 307 The charge transfer reaction 2.1 forms the ferrocene molecular ion Cp2Fe + 1, and the clustering reaction 2.2 leads to the formation of Cp3Fe2+ 2. It was proposed that ion 2 has a triple decker structure1 . Both the molecular ion 1 and the cluster ion 2 are inert towards ferrocene and alcohol. 2.1 Reactions with Methanol The variation of ion abundancies with time for the CpFe + /MeOH system is shown in Figure 2.1. Reaction of CpFe + with methanol gives rise to one ion only, CpFe(CH30H) + 3 at m/z 153. It is a product of the bimolecular addition reaction 2.3. The reactions involving loss of alkene or water and/or hydrogen, that are characteristic for higher alcohols, were not observed. CpFe+ + C H 3 O H CpFe(CH3OH)+ (2.3) m/z 121 3 m/z 153 The molecular formula of adduct ion 3 was confirmed by the reactions of CpFe + with CH3OD, CD3OH and CD3OD as illustrated in Figure 2.2. - 3 5 -100 -*s S <u S3 •o s 10 « o CpFe + CpFe(CH 3OH) +(3) ...X X- / I o 1000 Reaction Time (ms) 2000 Figure 2.1 Temporal variation of ion abundancies in the CpFe +/CH30H/Cp2Fe system at a total pressure of 1.2 x 10"6 Torr. Multiple resonance experiment confirmed that the product of reaction 2.3, Cp(CH30H) + 3, reacts slowly with ferrocene according to reaction 2.4. CpFe(CH3OH)+ + Cp 2Fe *~ C p 3 F e 2 + + C H 3 O H (2.4) 3 m/z 153 2 m/z 307 This is also apparent from the variation of ion intensities in Figure 2.1, as well as from the changes in the ratio of C p 2 F e + to Cp3Fe 2 + in the pure ferrocene and in the presence of alcohol. In the absence of methanol the ratio Cp 2 Fe + :Cp3Fe 2 + stabilizes at 80:20 whereas when methanol is introduced into the ICR cell the ratio becomes 30:70. This reversal of the Cp 2 Fe + :Cp3Fe 2 + ratio in the presence of methanol suggests that there is an additional source of Cp3Fe 2 + 2 ion besides reaction 2.2, i.e. reaction 2.4. The increase in the intensity of the C p 3 F e 2 + does not correspond to the decrease in the intensity of -36-Cp(CH30H) + 3. Some of the ion 3 must therefore be lost from the ICR cell at long delay times. As discussed in Chapter 1.3.3, reaction 2.3 follows a pseudo-first order kinetics. The plot of log(normalized intensity) vs. reaction time for the decay of CpFe+ should therefore be a straight line. It is apparent from Figure 2.1 that the decay plot has a curvature. We suggest that this is due to the vibrationally excited CpFe + ions. With increasing reaction time the excess vibrational energy is dissipated radiatively or through the collisions. This explanation is corroborated by the order of appearance of the products in Figure 2.1. At short delay times (less than -200 ms) only the Cp2Fe+ ion is observed. It is formed by a charge transfer reaction 2.1, in which the immediate contact of participating species is not necessary. As the CpFe + loses some of its excess vibrational energy, the products of addition reactions, CpFe(CH30H)+ and Cp3Fe2+, begin to appear. This coincides with the bend in the decay curve at about 200 ms delay time. The increase of the delay time D L l from 5 ms to 50 ms (see Figure 1.3 in Chapter 1.3.1) should allow for better cooling of the reactant ion. This was indeed demonstrated by a decrease in the curvature of the log(normalized intensity) vs. reaction time plot. However, during the delay time D L l the CpFe + ion, besides cooling, also undergoes ion-molecule reactions with both ferrocene and alcohol. This reduces the number of ions available for the reaction during the DL3 and causes a decrease in S/N ratio. Therefore, we decided to set the D L l at 5 ms and measure the slope of the log(normalized intensity) vs. reaction time from the linear portion of the plot. - 3 7 -w z. a _i u a: CpFe+ a) C H 3 O H CpFe(CH3OH)+ 11 *1 > <• 1 1 3 0 . . 1 4 0 1 5 0 MRfif> I N R . M . U . I "1 < '*« ' '^ " 1 6 0 0 1 10 1 7 0 a . a UJ UJ > a _i LU o r in CpFe+ b) C H 3 O D CpFe(CH3OD)+ CpFe(CH3OH)+ 1 9 0 1 4 0 1 5 0 M R S ? IN R . M . U . 1 6 0 170 Figure 2.2a-b Partial triple resonance mass spectra illustrating the reactions of CpFe with a) CH3OH at delay time 200 ms, b) C H 3 O D at delay time 250 ms. -38-C L Q « J I/I 2 11" a ID CpFe + c) C D 3 O H CpFe(CD3OH)+ r 150 r 160 Q 1 10 120 130 140 MRSS IN R . M . U . 170 in z u I -z .in a ID CpFe+ d) C D 3 O D CpFe(CD3OD)+ CpFe(CD3OH)+ 110 120 130 110 150 MRSS. IN. R.M.U IBO 170 Figure 2.2c-d Partial triple resonance mass spectra taken at delay time 200 illustrating reactions of CpFe+ with c) CD3OH, d) CD3OD. -39-The possible structures of ion 3 are shown below. I II III iv -^===r~0+ ^~~^=^~^~l+ *-~~Z===Z?-~-J~l + + Fe Fe Fe Fe / \ y \ y \ | H O C H 3 H O C H 3 H C H 2 O H Q H C H 3 Structures I, II and III result from the oxidative addition of a C-O, a O-H and a C-H bond to the metal centre, respectively. If the CpFe(CH30H) + had structure I, II or III we would expect to observe product ions like Cp3Fe20H+, Cp3Fe2CH3+, Cp 3 Fe20CH 3 + in reaction 2.4 . However, the only product observed in reaction 2.4 is due to displacement of the intact methanol molecule. This result indicates that the CpFe(CH 3 OH) + is an ion-neutral complex IV2> 3 . That is the CpFe + does not form a covalent bond with CH3OH, rather the ion and the neutral are held together by electrostatic attraction. When the CpFe + is reacted with C H 3 O D , in addition to the peak assigned to CpFe(CH30D) + at m/z 154, there is a peak with m/z one unit smaller, assigned to CpFe(CH3OH)+ (Figure 2.2b). Similarly, for the reaction of CD3OD with CpFe+, there is a peak at m/z 157 corresponding to CpFe(CD30D) + and a peak at m/z 156 assigned to CpFe(CD 3 OH) + (Figure 2.2d). We propose that the ions at m/z 153 in spectrum 2.2b and at m/z 156 in spectrum 2.2d arise from the H/D exchange of hydroxy deuterium with background water (reaction 2.5). This H/D exchange likely takes place also in the "free" alcohol before it associates with the CpFe + ion. However, in this case the rate of exchange would be slower as compared to the rate of exchange in reaction 2.5, because two neutral molecules (e.g. C H 3 O D and H2O), rather than an ion and a molecule (e.g. CpFe(CH30D)+ and H2O) are involved in the reaction. CpFe(ROD)+ + H 2 0 ^ CpFe(ROH)+ + HDO (2.5) -40-2.2 Reactions with Ethanol The reactions of CpFe + with ethanol are summarized in equations 2.6 to 2.10. CpFe + + C 2 H 5 O H 38% CpFe(C 2 H 5 OH) + 24%_ CpFe(H 2 0) + + 13% • CpFe(C 2 H 4 ) + + 13% • CpFe(C 2 H 2 ) + + 12% • CpFe(OC 2 H 4 ) + + C 2 H 4 H 2 0 H 2 0 + H 2 H 2 (2.6) (2.7) (2.8) (2.9) (2.10) The major product ions observed are CpFe (C2HsOH) + 4 at m/z 167 , CpFe (H20) + 8 at m/z 139, CpFe(C 2H 4)+ 9 at m/z 149, CpFe(C 2 H 2 ) + 10 at m/z 147 and CpFe(OC 2H 4)+ 14 at m/z 165. Reactions of CpFe+ with isotopomers CH3CH2OD (EtOD), C H 3 C D 2 O H (EtOH-d 2), C D 3 C H 2 O H (EtOH-d 3) and C D 3 C D 2 O D (EtOH-d 6), as illustrated by the spectra in Figure 2.3, were used to confirm the molecular formula of product ions and to elucidate the reaction mechanism. Q . I--u: si 2 LU > _ l LU LT Q 1 10 a) EtOH CpFe+ CpFe+(CH 3 CH 2 OH) + CpFe(H 20) CpFe(C 2H 2)+ + j CpFe(C 2H 4)H l?0 130 140 150 1B0 170 180 M R S S IN R . M . U . Figure 2.3a Triple resonance mass spectrum illustrating the reactions of CpFe + with ethanol at delay time 200 ms. The total pressure is 1.2 x 10"6 Torr. -41 -o 110 CpFe+ b) EtOD C p F e + ( C H 3 C H 2 O D ) + C p F e + ( C H 3 C H 2 O H ) + \ J i?o CpFe<HD0) + C p p e ( C 2 H 4 ) , .4.,-L r..lll 130 110 150 MRSS IN R . M .U . 160 170 180 D 1 ID c) EtOH-d 2 CpFe+ CpFe+(CH 3CD 2OH)+ — ^ CpFe(H 20)+ CpFe(C 2H 2D2)+ \ \ 120 13D 1MD- 150. MRSS IN fl.M-B. 160 170 180 Figure 2.3b-c Triple resonance mass spectra illustrating the reactions of CpFe + with b) C H 3 C H 2 O D at delay time 250 ms, c) CH 3 CD20H, delay time 200 ms. The total pressure is 1.2 x lO"6 Torr. - 4 2 -d) EtOH-d 3 CpFe(CD 3CH 2OH)+ CpFeCHDO)-1 120 ISO 110 150 160 MRSS. IN R . M . U , 170 180 e) EtOH-d 6 — CpFe + CpFe(CX>3CD20D)t. CpFe(CD 3 CD20H) + \ CpFe(D20) + \ 140 150 160 170 180 MRSS. IN. R . M . U . Figure 2.3d-e Triple resonance mass spectra illustrating the reactions of CpFe + with d) CD3CH2OH, delay time 150 ms, d) C2D5OD, delay time 210 ms. Total pressure of neutrals is 1.2 x 10"6 Torr. -43-The variation of ion abundances in the CpFe+/EtOH/Cp2Fe system is displayed in Figure 2.4. 0 200 400 600 800 Reaction Time (ms) Figure 2.4. Temporal variation of normalized ion intensities in the CpFe + /EtOH/Cp 2 Fe system. The curvature in the plot of log(normalized intensity) vs. reaction time for the reactant ion (Figure 2.4) is explained in the discussion related to Figure 2.1 (see page 36). The formation of CpFe(H 20)+ CpFe(C 2 H 4 )+ and CpFe(C 2 H 2 )+ ions can be explained in terms of a mechanism postulated by Allison and Ridge4 shown in Scheme 2.1. Oxidative addition of the C-0 bond to the metal centre (5-R—»6-R, where R=H) is followed by a 8-hydrogen transfer. The resulting olefin complex 7-R undergoes competitive loss (reductive elimination) of C H 2 = C H 2 and H 2 0 to generate ions 8 and 9 (R=H), respectively. The metal centre in ion 9 inserts into a C-H bond of ligated alkene. Subsequent 8-H transfer and elimination of molecular hydrogen yield ion 10. -44-The evidence in support of the above mechanism is based on the presence of hydrated cyclopentadienyl iron ion 8 and its deuterated analogues. With CH3CH2OD (Figure 2.3b) a peak at m/z 140 was observed that was assigned to the CpFe(H-OD)4" ion. A small peak at m/z 139 is due to H/D exchange of hydroxy deuterium with background water (reaction R. R, M + O H 5-R — • || — H M + - O H 6-R O H 7-R 6-H + O H 8 — V M + + H 2 0 9,15,25 ^ / C - H \ C - H M + * V H 1. B - H 2. - H 2 Scheme 2.1 Insertion of a metal centre into a C-0 bond. M = CpFe; ethanol, R=H, n=2; n-propanol, R=CH3, n=3; n-butanol, R=CH3CH2, n=4. Proposed structures of observed ions are in a box. 2.5, where R=H). With C H 3 C D 2 O H (Figure 2.3c) there is only a peak at m/z 139 that corresponds to ion 8. Absence of m/z 140 shows that no scrambling of hydrogens between a-deuteriums and hydroxy hydrogen takes place. With C D 3 C H 2 O H a 6-H transfer in the intermediate 6-R (Scheme 2.1, where R=D) and subsequent loss of C2H2D2 from 7-R results in the formation of CpFe(D-OH)4" ion at m/z 140. Again, no -45-scrambling between 8-deuteriums and hydroxy hydrogen is observed and no exchange of D from CpFe(DOH) + 8 with the H from background water according to reaction 2.5. This suggests that D in ion 8 is bound directly to the metal rather than to oxygen. From the absence of ions due to scrambling of a- or 8-deuteriums we infer that the 8-hydrogen transfer 6-R—>7-R in Scheme 2.1 is irreversible. The peak at m/z 165 in Figure 2.3a is the isotopic peak of Cp 5 6 Fe(C2H50H) + . However, its relative intensity cannot be accounted for solely by the isotopic contribution. The C p 5 4 F e ( C 2 H 5 O H ) + overlaps with the Cp 5 6 Fe(OC 2 H 4 )+ ion 14, a product of the dehydrogenation of ethanol, reaction 2.10. The formation of 14 can be explained by insertion of a metal into a H-O bond as shown in Scheme 2.2. The insertion step (11-R->12-R, where R=H) is followed by a 8-hydrogen transfer (12-R->13-R). Reductive elimination of molecular hydrogen yields CpFe(OC 2H4) + ion 14. R , O H 11-R M + R -O . 12-R H 6-H O H 13-R 14,17 M C 3 H 2 0 + - H 2 O M + - H , ^ H 6-H R O \ M + H 21 20 19 18 Scheme 2.2. Insertion of a metal centre into a H-0 bond. M = CpFe; CpFe(H20); ethanol, R=H; n-propanol, R=CH3. Proposed structures of observed ions are in a box. It is apparent from the above results that the CpFe + ion can insert into either a C-0 bond or into a H-0 bond. Thus, the CpFe(C2H50H) + ion 4 (m/z 167), can have two -46-different structures. These are respectively depicted in complex 6-R in Scheme 2.1 and in complex 12-R in Scheme 2.2. The ratio of the sum of branching ratios for the products that result from oxidative addition of the C-0 bond to the sum of branching ratios for the products that result from oxidative addition of the H-0 bond, 50:13, shows that the oxidative addition of a C-O bond is the preferred mode of interaction of ethanol with CpFe + ion. The reaction of CpFe + with ethanol is similar to that of Fe + with ethanol4'5. The major product ions are due to insertion of the metal centre into a C-O bond. There are some noticable differences, however. 1) The dehydrogenation of ethanol, reaction 2.10, was not observed in previous experiments. 2) Unlike our results, further dehydrogenation of ligated ethene (path 9—»10, Scheme 2.1, R=H) was not observed in earlier studies. 3) Fe + was found to insert into the C-C bond of ethanol5. Products due to CpFe + insertion into a C-C bond were not observed in the present study. The above differences between the reactions of CpFe + and Fe + with ethanol illustrate that the reactivity of the transition metal ions can be modified by a presence of the ligand bound to the metal. 2.3 Reactions with n-Propanol Reaction of n-propanol with CpFe + affords three ions, CpFe(H20)+ 8 at m/z 139, C p F e ( C H 2 C H C H 3 ) + 1 5 at m/z 163 and C p F e ( O H C H 2 C H 2 C H 3 ) + 16 at m/z 181 (reactions 2.11-2.13). CpFe + + n - C 3 H 7 O H CpFe(H 2 0)+ 22% — C p F e ( C 3 H 6 r • CpFe(C 3 H 7 OH) + C 3 H 6 H 2 0 (2.11) (2.12) (2.13) - 4 7 -The mass spectrum in Figure 2.5a illustrates these reactions. The presence of ions 8 and 15, CpFe(H 20)+ and CpFe(CH 2 CHCH 3 )+ can be accounted for by a metal insertion reaction into a C-0 bond (Scheme 2.1, R=CH 3). Evidence in support of this mechanism comes from a reaction of CpFe + with n-PrOD (Figure 2.5b). The presence of CpFe(HOD)+ m/z 140 and CpFe(CH 2=CHCH 3)+ m/z 163 suggests that after the insertion of the metal into a C-O bond (5-R—>6-R, Scheme 2.1) a 8-hydrogen transfer takes place (6-R—>7-R). Subsequent elimination of alkene or water gives rise to CpFe(HOD) + ion (7-R->8) or CpFe(C3H6)+ ion (7-R->15). Peaks at m/z 139 and m/z 181 in spectrum in Figure 2.5b are caused by H/D exchange of a hydroxy deuterium of CpFe(HOD)+ (reaction 2.5, where R=H) and CpFe(C 3 HvOD) + with background water (reaction 2.5, where R=C3H7). The temporal variation of ion intensities for CpFe + reacting with n-PrOH and Cp 2 Fe is displayed in Figure 2.6. - 4 8 -Q -z u UJ cH in a) n-PrOH CpFe(CH 3 (CH 2 ) 2 OH)-CpFe + CpFe(H 2OT CpFe(OC 3H 6)+ CpFe(OC 3 H 2 ) + CpFe(C 3H 6)+ ^ \ 1 2 0 W I ' W i T p i ' K ' <¥'^4^')1'i l l!'t-iY ,w'vrYf)'i lri^'^|*rl4f t t v P r ^ n i ' I S O 1 4 0 1 5 0 ' 1 6 0 M R S S I N . R . M . U . . 170 >7 in z UJ tu DC i n b) n-PrOD CpFeH CpFe(CH 3(CH 2) 2OD)+ — • CpFe(CH 3 (CH 2 ) 2 OH^ CpFe(HDO)+ CpFteCOCjH^ / CpFe(C3H 6 ) + J Cp2Fe+ M R S S ' IN R . M . U . . 180 200 Figure 2.5 Partial triple resonance mass spectra taken at delay time 200 ms illustrating the reactions of CpFe+ with a) n-C 3H70H and b) n-C 3H70D. -49-CpFe(H 20) + (8) CpFe(C3H7OH)+ (16) 0 200 400 600 Reaction Time (ms) 800 1000 Figure 2.6 Temporal variation of ion intensities in the CpFe + /«-PrOH/Cp2Fe system. Total pressure of neutrals is 1.2 x 10"6 Torr. For the sake of clarity the CpFe(OC3H6)+ 17 and CpFe(OC3H2)+ 21 were omitted from the graph. Their intensity is less than 2%. The products of self ion-molecule reactions of ferrocene, ions 1 and 2 are not shown either. As the intensity of Cp(C3H 7OH) + 16 increases with time two minor ions appear in the spectra CpFe(OC 3 H 6 )+ 17 at m/z 179 and C p F e ( O C 3 H 2 ) + 21 at m/z 175 (Figure 2.5a). The ion at m/z 179 is analogous to the m/z 165 ion observed for the reaction of EtOH with CpFe+. It is formed by insertion of the metal centre into a H-O bond (11-R—>12-R, Scheme 2.2, where R=CH 3 ) . Subsequent 8-hydrogen transfer (12-R—>13-R) and elimination of hydrogen yield ion 17. The increased length of alkyl chain allows elimination of up to three hydrogen molecules5. The metal centre in ion 17 inserts into a C - H bond that is 8 to C=0 bond (17—>18). This is followed by a 6-hydrogen transfer (18—»19) that leads to elimination of second molecule of hydrogen. A corresponding ion CpFe(OC3H4) + at m/z 177 was not observed. Loss of a third -50-hydrogen molecule (Scheme 2.2, 19—>20->21) gives rise to the CpFe(OC3H.2)+ ion at m/z 175. Ion 16, CpFe(C3H 7 OH) + , is initially formed in only about 13% yield but with increasing reaction time it becomes a major product ion. Hence, besides a reaction of CpFe + with n-propanol, there must be an additional source for its generation. As can be seen from Figure 2.6 the intensity of CpFe(H20)+ and CpFe(C 3 H 6 ) + decreases with time. It was therefore of interest to us to investigate their reactions using multiple resonance experiment. CpFe(H20)+ ion was formed by ion-molecule reaction 2.11 during a 100 ms delay time between the ionization and ejection events (see Figure 1.3). After its isolation CpFe(H 2 0) + was allowed to react with n-propanol. Figure 2.7b shows the products after an additional 100 ms delay time. The reaction of CpFe(H 2 0) + with n-propanol can be formally written as a substitution of water for alcohol (reactions 2.14 and 2.15). No further experiments were done to find out whether the eliminated water molecule is coming solely from the reactant ion or whether H- and/or HO- groups originate from n-propanol. CpFe(H 2 0) + + n - P r O H -8 m/z 139 CpFe(C 3 H 7 OH) + + H 2 0 (2.14) 16 m/z 181 CpFe(OC 3 H 2 ) + + H 2 0 + 3 H 2 (2.15) 21 m/z 175 The ion at m/z 175 in the spectrum in Figure 2.7b was assigned to CpFe(OC3H2)+. From its presence we infer that the metal centre in the CpFe(H 2 0) + ion inserts into a O-H bond of n-propanol (Scheme 2.2 with M+= CpFe(H20)+). When CpFe(H20)+ inserts into a H-O bond, elimination of water precedes 13-hydrogen transfer. The appearance of ion at m/z 181 in the spectrum in Figure 2.7b could also be explained in terms of C-0 bond insertion. However, this mechanism cannot account for ion at m/z 175. -51 -A similar experiment was performed to study the secondary reactions of ion 15, CpFe(C 3H6) +. It was found that 15 undergoes a substitution reaction with n-propanol in which alkene is displaced by alcohol (Figure 2.8, reaction 2.16). CpFe(C 3H 6)+ + n-PrOH CpFe(C 3H 7OH)+ + C 3 H 6 (2.16) 15 m/z 163 16 m/z 181 52 a) Reaction time = 100 ms CpFe(H20)+ 18D 140 15D n v M 1 i i i i H | » i i i I ' ' i * | r - 1 - 1 T -l BO 170 11 MRSS IN. R.M.U..-190 200 cL in z. LU LU IE , o -b) Reaction time = 200 ms CpFe(CH3(CH2)2CW CpFe(H 2 0) + i i i i i i i i ^ i i i 'i i i i i i i i i • • • • i < • ' • i • ' • • i 1 1 • ' i • o ' I I ' • • • ' I " " " " ' I 1 3 0 140 150 160 170 MRSS. IN. R . M . U . 180 190 200 Figure 2.7 Partial triple resonance mass spectra illustrating reactions of CpFe(H20) + with n-propanol. a) Cp(H20)+ formed by reaction 2.11 and isolated by multiple resonance after a 100 ms delay time, b) Products of the reaction of Cp(H 2 0) + with n-propanol after additional 100 ms reaction time. -53-170 180 MRSS. IN. R . M . U . 200 CpFe(CH3(CH2)20H) —I—i—i—i—i—|—i—i—r-170 180 MRSS IN. R.M.U.. 200 Figure 2.8 Partial triple resonance mass spectra illustrating the reactions of Cp^Hg)"1" with n-propanol. a) CpFe(C3H6)+ formed by reaction 2.12 and isolated by multiple resonance after a 100 ms delay time, b) Products of ion-molecule reactions of CpFe(C3H6)+ with n-propanol after additional 150 ms delay. -54-There are several differences between the reaction of CpFe + with n-propanol and the reaction of Fe + with n-propanol as it was studied by Huang and Gross5. 1. Fe + does not dehydrogenate n-propanol. We observed that CpFe + is able to expel up to three hydrogen molecules giving rise to CpFe(OC3H6)+ 17 and CpFe(OC3H2)+ 21. In principle, CpFe(OC3Hg)+ and CpFe(OC 3H2) + are primary product ions but they do not appear until long delay times when the intensity of CpFe(C3H70H)+ becomes high enough that they are not lost in the baseline noise. 2. Up to 32% of the products are due to insertion of Fe+ into C1-C2 or C2-C3 bonds of n-propanol. We did not observe any of the corresponding reactions with CpFe + . The reason for the different behaviour of CpFe + might be twofold. A lower electron deficiency on the metal centre (ED = 6 for CpFe + vs. ED = 11 for Fe + ) reduces its reactivity. Further, bulky cyclopentadienyl ligand may hinder the access of metal to the carbon atoms. Schwarz et al. 6 studied metastable ion decomposition of the n-propanol/Fe+ complex by tandem mass spectrometry. The only process observed was Fe + dehydration of n-propanol. While a significant portion of hydrogen originates from the 8-position, labeling experiments shown that the hydrogens from all positions were involved to some extent. To explain this observation he proposed a remote functionalization mechanism6. 2.4 Reactions with i-Propanol Reaction of CpFe + with /-propanol affords four product ions, CpFe(H20) + 8 at m/z 139, C p F e ( C 3 H 6 ) + 15 at m/z 163, C p F e ( O C 3 H 6 ) + 23 at m/z 179 and CpFe(C3H70H) +24 at m/z 181. Reactions 2.17 to 2.20 summarize the ion-molecule reactions initiated by CpFe + ion. -55-CpFe + + * - C 3 H 7 O H 62% • 23% 10% 7% CpFe(H 2 0) + + C 3 H 6 CpFe(C 3 H 6 ) + H , 0 v CpFe(OC(CH 3) 2) + + H 2 CpFe(C 3 H 7 OH) + (2.17) (2.18) (2.19) (2.20) Triple resonance FT-ICR mass spectra in Figure 2.9 illustrate reactions of CpFe + with /-propanol as well as with its isotopomers (CH 3 )2CHOD (i-PrOD), ( C D s h C H O H (/-PrOH-d6) and (CD 3 ) 2 CDOD (j-PrOH-dg). The variation of ion intensities with time for the reaction of CpFe + with i-propanol is shown in Figure 2.10. a) i-PrOH CpFeH C p F e ( O C 3 H 6 ) + - ^ CpFe(H20)-1 CpFe(C 3 H 6 ) + / CpFe(C 3HTOH) + - » - C p 2 F e + 1 > LMW*>^^ lC-n l i n 160 H R S 3 I N . R . M . U . 180 200 Figure 2.9a FT-ICR triple resonance mass spectrum of CpFe + reacting with i-propanol taken at a reaction delay time 180 ms. -56-b) i-PrOD CpFeCOLX^HT)-1 CpFe(OC3H6)+ CpFe(C3H6)+ 120 Cp2FeH 140 160 180 MRSS IN, R . M . U . 200 >-\— . in UJ I-I-a ui o— c) i-PrOH-d* CpFe-» / CpFe(OHC3HD6)+-^ Cp2Fe^ CpFe(OC3D6)+ o CpFe(HDO)+ CpFe(C3HD5)+ 1 2 0 14 0 160 180 200 MRSS IN. R . M . U . Figure 2.9b-c FT-ICR triple resonance mass spectra of CpFe+ reacting with i-PrOD and i-PrOH-dg, respectively. Reaction delay times are 100 ms (spectrum b) and 200 ms (spectrum c). -57-d) i - P r O H - d 8 CpFe + Cp2Fe + -^ CpFe(D20)+ CpFeCOCsDe)* CpFe(C 3D6) +-^ \ 120 1 4 0 1 6 0 1 8 0 2 0 0 MRSS IN R . M . U . Figure 2.9d FT-ICR triple resonance mass spectrum of CpFe + reacting with i-PrOH-dg taken at a reaction delay time 250 ms. T 1 1 1 — r 0 200 400 600 800 Reaction Time (ms) Figure 2.10 Temporal variation of ion intensities in the CpFe+//-PrOH/Cp2Fe system. Products of self ion-molecule reactions, ions 1 and 2 are not shown. -58-The reactions of j-propanol with CpFe + have several points in common with the reactions of n-propanol. These include: 1. Experiments with labeled i-propanol confirmed that ions 8 Cp(H20) + and 15 Cp(C3H6) + are formed by insertion of the metal centre into a C-O bond as shown in Scheme 2.3 (see also Figure 2.9). The abundance of ions 8 and 15 is similar in the CpFe+/i-PrOH and CpFe+/n-PrOH systems (see Table 2.1). Y 6-H O H N ~ " H / N O H O H M + H ' N O H + C 3 H 6 + H 9 O Scheme 2.3 Insertion of a metal into a C-O bond. M = CpFe. Proposed structures of observed ions are in a box. 2. Multiple resonance experiments confirmed that both Cp(H.20)+ and Cp(C3H6)+ react further with alcohol according to reactions 2.21, 2.22 and 2.23, respectively. CpFe(H 2 0) + + i-PrOH -8 m/z 139 CpFe(OC 3 H 6 ) + 23 m/z 179 CpFe(C 3 H 7 OH) + 24 m/z 181 H , 0 + FF H 2 0 (2.21) (2.22) CpFe(C 3 H 6 ) + + Z-PrOH ^ CpFe(C 3H 70H)+ + C 3 H 6 (2.23) 15 m/z 163 24 m/z 181 In these reactions water or alkene ligand are displaced by alcohol giving rise to ion 24, CpFe(C 3H70H) + . Secondary reaction of 8 with /-propanol yields also a dehydrogenation product 23 at m/z 179. Similarly as for the reaction of n-propanol with CpFe + , its -59-presence can be explained by the CpFe(H20)+ ion inserting into a H-0 bond rather than into a C-0 bond of alcohol (Scheme 2.4, where M = CpFe(H 20)). When CpFe(H20)+ reacts with /-propanol the 6-hydrogen transfer in Scheme 2.4 is preceded by elimination of water. Scheme 2.4 Insertion of a metal into a H-0 bond. M = CpFe, CpFe(H 20). Proposed structure of observed ion is in a box. Unlike n-propanol, the dehydrogenation of i-propanol, as illustrated in Scheme 2.4, is prominent, especially at long delay times. This is mainly due to reaction 2.21. Reactions of CpFe4" with labeled i-propanol confirmed that one of the hydrogen atoms in the expelled hydrogen molecule in Scheme 2.4 originates from CI carbon of alcohol. The other hydrogen is from a hydroxy group. Reductive elimination of hydrogen molecule gives rise to ion 23, a major product at longer delay times. A comparison of the product distribution for the reaction of Fe + with f'-propanol4 and that for the reaction of CpFe + and CpFe(H 2 0) + with /-propanol is yet another example of how the reactivity changes with a ligation on the metal centre. The major product ions for the former reaction are due to C-0 insertion. No dehydrogenation of alcohol by Fe + ion was observed. Gross et al.5 reported similar results with the exception that Fe(C3H60) + ions formed by a loss of hydrogen was only 2% of the total product distribution. Our results show that both CpFe + and especially CpFe(H20) + are able to dehydrogenate /-propanol to a larger extent. 23 H -60-2.5 Reactions with n-, i- and r-Butanol Equations 2.24 and 2.25 summarize CpFe + induced reactions of n-butanol. 55% CpFe + + n - C 4 H 9 O H | ^ CpFe(H 20) + + C 4 H 8 (2.24) CpFe(C 4 H 6 ) + + H 2 0 + H 2 ( 2.25) 45% Figure 2.11 shows three FT-ICR triple resonance mass spectra that illustrate the reactivity of CpFe + ion towards n-butanol and its isotopomers CH3(CH2)30D (ra-BuOD) and CD 3 (CD 2 )30H (n-BuOH-d9). Equations 2.26 to 2.28 summarize reactions of /-butanol with CpFe+. 68% CpFe + + / - C 4 H 9 O H ^ CpFe(H 2 0) + + C 3 H 6 (2.26) CpFe(C 4 H 6 ) + + H 2 0 + H 2 (2.27) CpFe(C 4 H 8 ) + + H 2 0 (2.28) Figure 2.12 shows a triple resonance mass spectrum for a reaction of CpFe + with i-butanol. Reactions of r-butanol and its isotopomer C(CD 3 )30D, r-BuOH-dio are illustrated in Figure 2.13 and summarized in equation 2.29 to 2.31. CpFe + + f - C 4 H 9 O H CpFe(H 2 0) + ^ CpFe(C 4 H 8 ) + 3% CpFe(C 4 H 6 ) + + C 3 H 6 (2.29) + H 2 0 (2.30) + H 2 Q + H 2 (2.31) -61 -Q-O IU t-z. liJ ID ( L J h i n; a) n-BuOH ^CpFe"1 CpFe(C4H6)+_^ --CpFe(H20)+ 120 lub .lfio isn Cp2Fe + \ CpFe(OHC4H9)+ 9 i«y<>^A1yw^ MRSS JN- R . M . D . 200 Q-O L L . J k J I E b) n-BuOD ^CpFe + Cp2Fe+ CpFe(C4H6)+_^ ^ 120 IMO 160 .180 200 MRSS -IN- R . M . U . \ CpFe(OjOC4H9)+ Figure 2.11a-b Triple resonance mass spectra illustrating reactions of CpFe + with n-BuOH, reaction time 140 ms and b) w-BuOD, reaction time 100 ms. -62-^CpFe-1 CpFe(C4D6)+-^ CpFe(HDO)+ c) n-BuOH-d9 CpFe(OHC4D9)+ X ^Cp 2 Fe + V 1 1 2 0 1 4 0 1 R 0 1 8 0 2 0 0 M R S S I N R . M . U . Figure 2.11c Triple resonance mass spectrum illustrating reactions of CpFe + with n-BuOH-d9 at a delay time 100 ms. .CpFe"1 CpFe(H20)+ Cp2Fe-' CpFe(C4H8)-' CpFt(C4H6)+. 1 2 0 1 4 0 1 R 0 1 8 0 i-BuOH CpFe(OHC4H9)+ 0 M R S S I N R . M . U . 200 Figure 2.12 Triple resonance mass spectrum taken at delay time 150 ms illustrating reactions of CpFe4" with /-BuOH. -63-a) t-BuOH CpFe"1 120 CpFe(C4H8)+-^ —CpFe(H20)-Cp2Fe+ 140 160 180 MPS5 IN R . M . U . CpFe(OHC4H9)+ 200 220 b) t-BuOH-dio CpFe(C 4D 8) +-^ CpFe(D20)+ — Cp2Fe"' 120 140 160 180 MRSS. IN. R . M . U . -T-V-T i -y . I • • | - r - T - f I | 0 200 I D. u 220 Figure 2.13 Triple resonance mass spectra illustrating reactions of CpFe + with a) r-BuOH, reaction delay time 150 ms and b) f-BuOH-dio, reaction delay time 200 ms. -64-As can be seen from Table 2.1, the major primary product ions are CpFe(H20)+ 8 at m/z 139, CpFe(C 4 H 8 ) + at m/z 177 and CpFe(C 4H 6)+ at m/z 175. They are formed by oxidative addition of C-0 bond to the metal centre. The mechanism for the reaction of CpFe + with n-butanol is displayed in Scheme 2.1 (R=CH.3CH2). Insertion of the metal centre into the C-O bond (5-R—>6-R) is followed by a 6-hydrogen transfer (6-R—>7-R) that yields hydrated metal-olefin complex. Subsequent loss of alkene (7-R—>8) or water (7-R—>25) from complex 7-R give rise to ions 8 and 25, respectively. Ion 25 was not observed. The reaction proceeds further by the insertion of the metal into an allylic C-H bond of ligated 1-butene. Subsequent 6-H transfer and loss of hydrogen molecule give rise to ion 26. This indicates that step 25—»26 is faster than step 7-R—»25. Scheme 2.5 shows the mechanism for the insertion of the CpFe + into a C-O bond of j-butanol and r-butanol (step 5-R—>6-R). For clarity ?-butanol is not shown in the scheme but all the reactions following step 6-R yield the same products for /-butanol and r-butanol. Subsequent 6-hydrogen transfer (6-R—>7-R) and a loss of water give rise to ion 27 at m/z 177. The metal in 27 may insert into a C-H bond of ligated alkene. This is followed by a 6-hydrogen transfer and a loss of hydrogen that give rise to ion 28. The organic ligand in 28 forms either a metallacycle or has a trimethylenemethane structure. Unlike CpFe+/n-butanol system, in the CpFe+/i-butanol system step 27—>28 occurs to much smaller extent. The CpFe(C 4H6) + ion 26 in Figure 2.14 and the CpFe(C 4H6) + ion 28 in Figure 2.15 do not undergo any further reactions with neutral ferrocene or alcohol even at long delay times. -65-M + OH 5-R _6_ M+- OH 6-R -H Ti ^ H OH 7-R + C4H8 + H 20 C-H insertion 1.6-H 2. -H2 Scheme 2.5 Insertion of a metal into a C-O bond of /-butanol. M=CpFe. Proposed structures of observed ions are in a box. 100 -G m I 10 o 0 • \ CpFe + CpFe(C 4H 9OH) + (29) I -4^ 4. f * \ CpFe(C 4H 6)+ (26) • ''•> — • + / / X i CpFe(H 20) + (8) * + r 200 400 600 Reaction Time (ms) T 800 1000 Figure 2.14 Temporal variation of ion intensities in the CpFe+/n-BuOH/Cp2Fe system. Products of self ion-molecule reactions of ferrocene, ions 1 and 2 are not shown. -66-0 200 400 600 800 1000 Reaction Time (ms) Figure 2.15 Temporal variation of ion intensities in the CpFe+/?-BuOH/Cp2Fe system. Products of self ion-molecule reactions of ferrocene, ions 1 and 2 are not shown. CpFe + CpFe(C 4H 9OH) + (31) if / x xV if f P\ CpFe(C 4H 8)+ (27) I 1 IB / * / >y f CpFe(H 20) + ( 8 ) \ * CpFe(OC 3H 6) + (32) •\ 0 T 200 400 600 Reaction Time (ms) 800 1000 Figure 2.16 Temporal variation of ion intensities in the CpFe+/f-BuOH/Cp2Fe system. Products of self ion-molecule reactions of ferrocene, ions 1 and 2 are not shown. -67-Resonance experiments confirmed that CpFe(ROH)+ 29 (R=CH 3(CH2)3), 30 (R=(CH3)2CHCH2), and 31 (R=(CH3)3Q are products of a secondary reaction of CpFe(H20)+ 8 and CpFe(C 4H 8)+ 27 with alcohol (reactions 2.32 and 2.33). CpFe(H20)+ + C 4 H 9 O H - CpFe(C 4H 9OH)+ + H 2 0 (2.32) 8 m/z 139 29,30,31 m/z 195 CpFe(C 4H 8)+ + C 4 H 9 O H - CpFe(C 4H 9OH)+ + C 4 H 8 (2.33) 27 m/z 177 29, 30, 31 m/z 195 Reactions 2.32 and 2.33 take place for all three isomers of butanol. The product of C-C bond insertion begins to emerge for the reactions of f-butanol. The mechanism is shown in Scheme 2.6. The insertion of a metal into a C-C bond is followed by a 8-hydrogen transfer. Subsequent elimination of methane gives rise to the CpFe(OC 3 H 6 ) + 32 at m/z 179. Scheme 2.6 Insertion of a metal into a C-C bond of f-butanol. M=CpFe. Proposed structure of observed ion is in a box. 2.6 Summary and Conclusions Table 2.2 summarizes the distribution of primary product ions for the reaction of the CpFe+ ion with the studied alcohols. Two general types of CpFe+ reactions with alcohol were identified: 1. C-O bond insertion and subsequent dehydration (or elimination of alkene) 2. H-0 bond insertion and subsequent dehydrogenation -68-Table 2.2 Primary product distributions for the ion-molecule reactions of CpFe + with alcohols. CpFe(ROH)+ C-O bond insertion H O bond insertion CpFe(H20)+ CpFe(C nH 2 n)+ CpFe(C n H 2 n . 2 )+ CpFe(OC nH 2 n)+ MeOH 100 3 EtOH 38 24 13 13 12 4 8 9 10 14 n-PrOH 13 65 22 * 16 8 15 /-PrOH 5 62 23 10 24 8 22 23 n-BuOH 0 55 45 29 8 26 i-BuOH 0 68 15 17 30 8 27 28 ?-BuOH 0 70 27 3 31 8 27 28 The upper number of each pair is the percentage of the total ion products. The lower bold number is the compound number. A blank indicates that the product ion was not observed. A zero indicates that the ion was observed but as a secondary product. * It is impossible to determine the branching ratio for the CpFe(OC3H6)+ ion at short delay times due to its very low intensity. It was observed only at long delay times. -69-By inspecting data in Table 2.2 it is obvious that the CpFe(H.20)+ and the CpFe(C n H2n) + ions, resulting from the insertion of the metal into a C-0 bond, are the major, if not the only, primary product ions for the reaction of CpFe+ with all alcohols except methanol. Further loss of a hydrogen molecule from the ironcyclopentadienyl-alkene complex CpFe(C nH.2 n) + yields the CpFe(C nH2n-2) + ion. This was observed for ethanol (reaction 2.9), n-butanol (reaction 2.25) and to a smaller extent for j-butanol and r-butanol (reactions 2.27 and 2.31, respectively). The organic ligand in the CpFe(CnH.2n-2)+ ion is ethyne (for the reaction of ethanol), 1,3-butadiene (for the reaction of n-butanol) and trimethylenemethane (for the reaction of i-butanol and r-butanol). The CpFe(OC n H2n) + ion (last column in Table 2.2) is formed by dehydrogenation of the CpFe(ROH)+ complex as it is displayed in Schemes 2.2 and 2.4. It is also formed in secondary reactions of CpFe(H20) + 8 with /-PrOH (reaction 2.21). In all systems studied it is a minor product with the exception of the CpFe + /i-PrOH system, where it becomes the major ion at long delay times. The ratio of the products due to the insertion of the metal into a C-O bond vs. metal insertion into a H-O bond is 50:13 for ethanol, 85:10 for /-propanol and 100:0 for n-, i-, t-butanol. The initial intensity of the ion formed by the insertion of the metal into a H-0 bond of n-propanol is too low to be reliably estimated. This ratio trend suggests that the CpFe + insertion into a C-O bond of alcohol is preferred to insertion into a H-0 bond. As we have pointed out earlier, the reactivity of the metal can be changed by varying the ligands that are attached to it. The secondary reactions of CpFe(H20) + ion offer an example of this modified reactivity. Unlike the reactions of CpFe + with alcohols, where the insertion into the C-0 bond was mainly observed, the major products for the reaction of CpFe(H20)+ with alcohols seem to arise from the insertion of the metal centre into a H - 0 bond. This is illustrated in equations 2.15 and 2.21. In the CpFe+//-PrOH/Cp2Fe system the secondary reaction 2.21 gives rise to a major product ion at long delay times. - 7 0 -The ironcyclopentadienyl-alcohol complex CpFe(ROH) + can be formed in two ways: 1) As a product of primary reactions of CpFe + with alcohol (reactions 2.3, 2.6, 2.13 and 2.20). 2) As a product of secondary reactions of CpFe(H20) + (reactions 2.14, 2.22 and 2.31) and C p F e ( C n H 2 n ) + (reactions 2.16, 2.23 and 2.32) with alcohol. The abundance of CpFe(ROH)+ as a primary product decreases with the increasing length of alkyl chain of alcohol (first column, Table 2.2). For methanol it is the only product formed in reaction 2.3. For the reaction of CpFe + with C4-alcohols the adduct ions 29, 30 and 31 do not even appear as a primary product. They are formed by secondary reactions of CpFe(H20)+ and CpFe(C4Hg)+ ions with alcohol (reactions 2.32 and 2.33). The adduct ions 4,16 and 24 are formed in both ways. In all systems except CpFe +/i-PrOH, the CpFe(ROH) + becomes the major ion at long delay times (Figures 2.4, 2.6, 2.14-2.16) For all the reactions of CpFe+ with alcohols, i.e. 2.3, 2.6-2.10, 2.11-2.13, 2.17-2.20, 2.24 and 2.25, 2.26-2.28 and 2.29-2.31 the total rate constant k R o H was determined from the slope of the log(normalized intensity) vs. reaction time. The procedure is described in detail in Chapter 1.3.3. As it is apparent from Figures 2.1, 2.4, 2.6 and 2.10 the plot of log(normalized intensity) vs. reaction time has a curvature at short delay times. This is especially pronounced in the methanol and ethanol systems. The curvature gradually disappears with the increasing length of alkyl chain and it is absent in butanol systems. We explain this by the presence of vibrationally excited reactant ions that are formed during the electron ionization event. Two experimental observations corroborate this explanation. First, as the reaction time increases, the decay of the reactant ion becomes linear. Longer reaction time allows for sufficient cooling of the CpFe + by radiation (and/or collisions) with neutral molecules. Second, with the increasing size of -71 -alcohol, more vibrational degrees of freedom become available and the excited complex [CpFe(ROH)+]* is more efficiently stabilized. In all cases in which the plot of log (normalized intensity) vs. reaction time for the reactant ion has a curvature, the slope needed for calculation of rate constant was taken from the linear portion of the plot. The experimental rate constants kROH, as well as theoretical rate constants kL and kADO, calculated from the Langevin theory7 and from the Average Dipole Theory (ADO) 8 , respectively, are listed in Table 2.3. The numbers in the last column were calculated as kROH/kADO and represent the fraction of reactive collisions (reaction efficiency). The values of kROH from Table 2.3 are plotted in Figure 2.17. As can be seen from the data in Table 2.3 and from Figure 2.17, the reactivity of the CpFe + /ROH system increases with the increasing length of alkyl group of alcohol. In the CpFe + /MeOH system, only about 10% of the collisions lead to the reaction. The CpFe +/EtOH system is twice as reactive as methanol system and the CpFe+/n-PrOH is four times as reactive as methanol system, with over 40% of the collisions being reactive. The most pronounced increase in the reactivity is between the C1/C2 alcohols and C2/C3 alcohols. After that the reactivity levels off with C4-alcohols being only slightly more reactive than C3-alcohols. This chain-length effect was observed before for the reactions of C o + with 1-chloro alkanes and linear alcohols9. This trend in reactivity can be better understood if we consider the elementary reactions of the general ion-molecule association process o^bs A+ + B • AB+ (2.34) where ko D S is the rate constant for formation of the association product A B + . -72-Table 2.3 Experimental and theoretical rate constants for the reaction of CpFe + with alcohols. Alcohol kROHxlOlO k L x l O l O k A D O x l O l O Jgg^  MeOH 1.7 8.4 16.0 0.10 EtOH 4.3 9.4 15.6 0.28 n-PrOH 6.6 9.6 15.0 0.44 i-PrOH 7.3 9.6 15.0 0.49 n-BuOH 7.0 10.3 15.0 0.47 /-BuOH 7.3 10.3 15.0 0.48 f-BuOH 7.2 10.3 15.0 0.48 All k values are in units cm3molecule"1s"1. Reaction efficiency was calculated as kROH/kADO -1 -1 ilecule s Os 00 n-A i-• t-* * * o o 1-1 © ^ 4 -X o * ** • * * 2 - ** * * T 0 2 3 Number of carbons in alcohol Figure 2.17 The trend in reactivity of CI to C4 alcohols with CpFe + ion. -73-The overall reaction 2.34 consists of the following elementary reactions: A B + (2.35) k s [M] A + + B ^ , (AB + )* k b — • A B + + hv (2.36) C + + D (2.37) where kf and kb represent the rate constants for formation and back-dissociation of the excited complex. It is usually assumed that kf is given by the corresponding capture rate constant calculated according to the ADO model of Su and Bowers10. The excited complex can be stabilized via collisions with bath gas M with a rate constant k s (reaction 2.35), via radiative emission of a photon with a rate constant k r (reaction 2.36) or via dissociation of the complex into some fragment channel other than the reactant channel with a rate constant kd (reaction 2.37). The stabilization of the excited complex by collisions with neutral M is not very likely under our experimental conditions. A higher pressure would be needed. For the reaction of methanol with CpFe + no neutral loss from the adduct CpFe(CH3OH)+ was observed. This means that the excited (CpFe(CH30H)+)* complex can be stabilized by the emission of the photon only (reaction 2.36). The low reaction efficiency indicates that the k r « k D and that most of the excited adduct ion falls apart into the reactant species, the CpFe4" ion and methanol. With higher alcohols, an important reaction channel becomes available due to the presence of 6-hydrogen. If the metal inserts into a C-0 bond the B-hydrogen transfer leads to competitive elimination of water or alkene. Metal insertion into a H-O bond and subsequent 6-hydrogen transfer result in the elimination of a hydrogen molecule. In either case, the neutral carries away the excess vibrational energy and stabilizes the product ion (reaction 2.37). The stabilization of the excited (CpFe(CnH2n+iOH)+)* complex (where n>2) by the fragmentation, in addition to the radiative emission of a photon, accounts for the increased reactivity of C2-C4 alcohols when compared to the reactivity of methanol. Further increase in reactivity between C2 -74-and C3 alcohols is probably related to the "origin" of the B-hydrogen that is being transferred from alkyl chain onto the metal. When the metal inserts into a C-O bond of ethanol, the B-hydrogen is on a primary carbon while in n-propanol, the B-hydrogen is on a secondary carbon (see Scheme 2.1). Lower BDE of a C-H bond when hydrogen is bound to a secondary carbon vs. when it is bound to a primary carbon makes the 8-hydrogen transfer more facile. This is also reflected in the relative abundance of CpFe(C3H.70H)+ to CpFe(C2H 5OH)+ adduct ions (first column, Table 2.3). In the CpFe+M-PrOH system only 13% of the excited adduct complex remained intact, the remaining 87% of it fragmented into ions 8 and 15. While in the CpFe + /EtOH system 38% of the excited adduct complex remained intact. Almost three times as much as in the CpFe +/n-PrOH system. -75 -References (1) Kan, Z . , University of British Columbia, Vancouver, 1993. (2) McAdoo, D. J. Mass Spectrometry Reviews 1988, 7, 363-393. (3) Bowen, R. D. Acc. Chem. Res. 1991, 24, 364-371. (4) Allison, J.; Ridge, D. P. Journal of the American Chemical Society 1979,101, 4998-5008. (5) Huang, S.; Holman, R. W.; Gross, M . L. Organometallics 1986, 5, 1857-1863. (6) Schwarz, H.; Karrass, S.; Prusse, T.; Eller, K. Journal of American Chemical Society 1989, 111, 9018-2023. (7) Gioumousis, G.; Stevenson, D. P. The Journal of Chemical Physics 1958, 29, 294-299. (8) Su, T.; Bowers, M . T. Jounal of Chemical Physics 1978, 69, 2243. (9) Tsarbopoulos, A.; Allison, J. Journal of American Chemical Society 1985,107, 5085-5093. (10) Su, T.; Bowers, M . T. The Journal Of chemical Physics 1973, 58, 3027-3037. -76-Chapter 3 Gas-Phase Ion-Molecule Reactions of Cyclopentadieneiron Ion Csrl6Fe + with Alcohols. In this chapter we will examine the reactions of ironcyclopentadiene ion C5H6Fe+ with methanol, ethanol, n-, f'-propanol and n-, i-, f-butanol. As discussed in Chapter 1.1 the C5H6Fe + ion was proposed to have a 7t-cyclopentadienyl iron-hydride structure 2 shown below. 2 We are interested in whether the reactions of Cs^Fe"1" ion, that is thought to contain a metal-hydrogen bond, differ from the reactions of (r)5-C5H5)Fe+ ion that does not have a metal-hydrogen bond. The C5H6Fe + ion was generated by 30 eV, 20 ms electron beam pulse on cyclohexadienone iron tricarbonyl (r|4-c-C6H60)Fe(CO)3, that was present in the ICR cell at a pressure 1.2 x 10~7 Torr. The Cs^Fe"1" ion was subsequently isolated by a triple resonance experiment and allowed to react with alcohol during a delay time DL3. The alcohol was introduced simultaneously with the organometallic neutral and kept at a pressure about ten times higher in order to minimize the extent of self ion-molecule reactions of C s ^ F e * with its parent neutral1. The products of self ion-molecule reactions, summarized in equations 3.1-3.3, were observed in all studied systems -77-C 5 H 6 Fe+ + (C 6 H 6 0)Fe(CO) 3 - » (C 5H5)(C6H 50)Fe2+ + 3CO + H 2 (3.1) m/z 122 m/z 270 -> (C 5 H 6 ) (C 6 H 6 0)Fe2 + + 3CO (3.2) m/z 272 -> (C5H6)(C6H60)Fe2(CO)+ + 2CO (3.3) m/z 300 During the electron ionization event, along with the C5H-6Fe+ ion at m/z 122, the CpFe + fragment at m/z 121 is generated. Because the ICR frequencies of the two ions are close to each other, the ion of interest, i.e. Csr^Fe"1", cannot be isolated completely without some interference from CpFe + ion. Hence, in addition to the reactions of CsHgFe"1" with alcohol and the self ion-molecule reactions 3.1-3.3, the products of the reactions of CpFe + ion with alcohol and organometallic neutral were detected as well in all systems studied. The reactions of CpFe+ with (C6H60)Fe(CO)3 were studied before by Kan 1 and are summarized below. CpFe+ + C 6 H 6 OFe(CO) 3 -> Cp(C 6 H 6 0)Fe 2 + + 3CO (3.4) m/z 121 m/z 271 - » Cp(C 6H 60)Fe 2(CO)+ + 2CO (3.5) m/z 299 -> CpFe(C 6H 60)+ + Fe(CO) 3 (3.6) m/z 215 The observed ion-molecule reactions and primary product distributions for C5H6Fe + with methanol, ethanol, n-, i-propanol and n-, i-, f-butanol are listed in Table 3.1. The branching ratios in Table 3.1 were determined from the ratio plots for individual alcohols by extrapolating the normalized ion intensity to reaction time zero. -78-Table 3.1 Primary product distribution for the ion-molecule reactions of C5H6Fe+ with alcohols. alcohol ion formed m/z neutrals reaction ion no. no. CsHgFe"1" + CH3OH C 5 H 6 F e + + C 2 H 5 O H C s ^ F e * + J -C 3 H 7 OH C5H6F6+ + n - C 3 H 7 O H C 5 H 6 F e + + ra-QHgOH C 5 H 6 F e + + / - C 4 H 9 O H C s ^ F e * + r - C 4 H 9 O H 100% 100% 60% • 30% 10% 80% • 10% 10% 90% 10% CpFe(OCH3)+ CpFe(OH)+ CpFe(OH)+ CpFe(C 3 H 5 ) + CpFe(OC 3 H 5 ) + CpFe(OH)+ CpFe(C 3 H 5 ) + CpFe(OC 3 H 5 ) + CpFe(OH)+ CpFe(C 4 H 7 ) + 8 0 V c P F e ( O H ) + 20% CpFe(C 4 H 7 ) + 55% • CpFe(C 4 H 7 ) + 15% >- CpFe(OH)+ 152 H 2 (3.7) 7 138 H 2 + C 2 H 4 (3.9) 10 138 C 3 H 8 (3.H) 10 162 H 2 + H 2 0 (3.12) 34 178 2 H 2 (3.13) 39 138 H 2 + C2^€ (3.15) 10 162 H 2 + H 2 0 (3.16) 34 178 2 H 2 (3.17) 44 138 H 2 + C 4 H 8 (3.19) 10 176 H 2 + H 2 0 (3.20) 50 138 H 2 + C 4 H 8 (3.21) 10 176 H 2 + H 2 0 (3.22) 56 176 H 2 + H 2 0 (3.23) 56 138 H 2 + C 4Hg (3.24) 10 Data are expressed in %Xfragments = 100%. Product percentages are reproducible to ±10% of the values reported. The total pressure is kept at 1.2 x 10-6 Torr. -79-3.1 Reactions with Methanol Kan showed1 and we have confirmed that the CsH6Fe + reacts with methanol according to addition-dehydrogenation reaction 3.7 to give the only product CpFe(OCH.3)+ ion 7 at m/z 152. C 5 H 6 Fe+ + C H 3 O H m/z 122 100% CpFe(OCH3)+ + H 2 7 m/z 152 (3.7) The variation of ion abundances with reaction time for the C5H6Fe +/MeOH system is displayed in Figure 3.1. 100 -a CU cu "a S u © 10 1 -* • • •We" CpFe(OCH 3) + (7) • X 0 T 200 400 Reaction Time (ms) 600 800 Figure 3.1. Temporal variation of ion intensities in the C5H6Fe+/MeOH/(C6H60)Fe(CO)3 system. Products of self ion-molecule reactions of C5H.6Fe+, reactions 3.1-3.3 are not shown. - 8 0 -Similarly as in the CpFe + /MeOH system a curvature is observed in the log(normalized intensity) vs. reaction time plot for the decay of the reactant ion (Figure 3.1). We propose that it is caused by vibrationally excited CsHgFe"1" ions. The slope for the calculation of the rate constant was taken from the linear part after the reaction time of about 100 ms. To determine the origin of product hydrogen molecule in reaction 3.7, the experiments with deuterated analogues of methanol, CH3OD (b), CD3OH (c) and CD3OD (d) were performed. Typical spectra obtained in these experiments are shown in Figure 3.2. Peaks labeled in plain letters are products of ion-molecule reactions of the interfering CpFe + ion (see also Chapter 2.1). The results of these experiments are summarized in Table 3.2. 1 3 cr n a o ST ft Ni O n sc o D 2. 53* CD 3 to TO C »1 TO u» to • cr cT O a s o t/5 CD o to s R E L R T I V E I N T E N S I T Y 5 0 1 0 0 8 5 cr CD o 3 n x CD + - 0 R E L R T I V E I N T E N S I T Y 5 0 1 0 D -82-o. o in z UJ C 5 l l 5 l 7 c + - ^ C) C D 3 O H ^ - C 5 H « F e + C 5 H S Fe(CD 2 OH)+_^ C 5 H 5 F e ( O C D 3 ) + / C5H5Fe(CD3OH)+ a 1 110 120 1 3 0 1 1 0 1 5 0 M R S S I N R . M . U . 1 6 0 1 7 0 C L • d) CD3OD C 5 H 5 F e ( C D 2 O D ) + / C 5 H 5 F e ( O C D 3 ) + \ * - C 5 H 6 F e + — • C 5H 5Fe(CD 3OD) + / C 5 H 5 F e ( C D 2 O H ) + \ in z UJ I-z U J " > L U L T 110 1 2 0 1 3 0 1 4 0 „ , , I S O M R S S I N R . M . U . 1 6 0 1 7 0 Figure 3.2c-d Partial triple resonance mass spectra illustrating the reactions of Csli^Fc with c) CD3OH at delay time 180 ms, d) CD3OD at delay time 150 ms. -83-Table 3.2 Ionic products observed for the reaction of C5H6Fe+ with methanol and its isotopologues. A+ Am = A+-P+ * inferred neutral loss CH 3 OH/(C 5 H 6 )Fe+ 3a 154- 152= =2 H 2 CH 3 OD/(C 5 H 6 )Fe+ 3b 155- 153= =2 H 2 155- 152= =3 HD CD 3 OH/(C 5 H6)Fe + 3c 157- 154= =3 HD 157- 155= =2 H 2 CD 3 OD/(C 5 H 6 )Fe+ 3d 158- 155= =3 HD Am is the mass difference between the adduct intermediate A+ and the observed product ion P + . For the reaction of CsHgFe"1" with CD3OD we observed a peak at m/z 155 (spectrum d in Figure 3.2) that corresponds to a loss of HD molecule from the adduct intermediate 3d (m/z 158, Table 3.2). For the reaction of (C5H6)Fe+ with C D 3 O H (spectrum c), the product ion at m/z 155 was assigned to the loss of H 2 from the adduct intermediate 3c (m/z 157). The presence of ions at m/z 155 in spectrum d and c confirms that one of the hydrogen atoms in the eliminated neutral in reaction 3.7 originates on the cyclopentadienyl ring. This indicates that the reacting C5H6Fe+ ion has an olefin-metal-hydride structure 2. The other hydrogen atom in the eliminated neutral can be either from the hydroxy group or from the methyl group of methanol. The presence of two peaks in spectra b and c in Figure 3.2 as well as the data in Table 3.2 indicate that both options are possible. The data in Table 3.2 can be explained by a mechanism described in Scheme 3.1, that uses the reaction of C s ^ F e * with CD3OH (c) as an example. The adduct intermediate from which -84-the hydrogen molecule is eliminated can be formed in two ways. The metal may insert into a H-0 bond of methanol giving rise to the intermediate with structure 3c-l. Subsequent elimination of H 2 yields ion 7c-1 at m/z 155. Another possibility is the insertion of the C5HoFe+ into a C-D bond of methanol that results in the adduct intermediate with structure 3c-2. The elimination of HD from 3c-2 yields ion 7c-2 at m/z 154. Both ions, at m/z 155 and at m/z 154, were observed in spectrum c. The ion at m/z 155 could also be formed from complex 3c-2 by scrambling of the deuterium bound to the iron centre with the hydrogen on the cyclopentadienyl ring (3c-2—»4c—»5c). The dihydrido-cyclopentadienyl complex 3c-2 rearranges to a hydrido-cyclopentadiene complex 4c by a migration of D from the metal centre onto the ring. A similar mechanism, that invokes a rapid equilibrium between the hydrido-cyclopentadienyl complex and the cyclopentadiene complex has been proposed to explain the observed H/D exchanges of CsHgFe4" and C5H6Co + with excess D 2 2 . The incorporated deuterium in complex 4c is in the endo position and remains unscrambled. Complex 4c has to rearrange first via a [1,5] sigmatropic shift into a 5c before the hydrogen can transfer from the cyclopentadiene ring back to the metal centre giving complex 6c. Elimination of H 2 from complex 6c would give ion at m/z 155. If this type of H/D scrambling does occur there should be a deuterium on a cyclopentadienyl ring. However, if the ion at m/z 155 is formed by insertion of the metal into a H-0 bond (3c-l—>7c-l) all deuteriums will be on the methoxy group. The CID of ion at m/z 155 yields only a fragment at m/z 121 that corresponds to CsHsFe"1". This result indicates that the above discussed H/D scrambling in complex 3c-2 does not take place and that the ion at m/z 155 is formed by the insertion of iron into a H-0 bond of methanol. -85-F e + C D . O H H C-D H i Fe / I \ H 3c-l O C D 3 1 + Fe / I \ H D C D 2 - O H 3c-2 - H , - H D 7c-l m/z 155 7c-2 m/z 154 Hexo 1 + Dendo 1 Fe , " Fe / \ [1,5] / \ H CD2 - O H H C D 2 - O H 5c 4c n + - H 2 C D 2 - 0 H 6c m/z 155 Scheme 3.1 Proposed mechanism for the interaction of a metal centre in the CsHgFe"1" ion with methanol. Proposed structures of observed ions are in a box. -86-In Chapter 2.1 we postulated that the CpFe + ion reacts with methanol to give only an ion-neutral complex CpFe(CH30H) +. The experiments with isotopomers of methanol (Table 3.2) show that the CsHgFe"1" ion reacts with methanol by inserting into both a H-0 and a C - H bond of alcohol (Scheme 3.1). Allison and Ridge studied the reactions of atomic metal ions with alkyl halides and alcohols by ion cyclotron resonance3- 4 . They found that the only ionic product of the reaction of Fe + with methanol is FeOH + . In a later study Huang and Gross found Fe + to be unreactive towards methanol5. The reaction of F e + with methanol observed by Allison and Ridge has been attributed to electronically excited iron cations6. Schwarz et al.6 found that the FeCH3 + ion inserts into the H-0 bond of methanol. The results of our experiments as well as those of Huang and Gross, and Schwarz demonstrate that the inherent reactivity of gas phase metal ions as well as the site of bond activation can be altered by the ligands bound to the metal centre. This has been shown before for the reactions of metal ions with alkenes and alkanes. For example, while the bare metal ions Fe + , C o + and N i + are unreactive towards ethene, the hydrides FeD+, C o D + and N i D + are reactive producing D / H exchange and (except for FeD + ) dehydrogenation products7. In sharp contrast to the bare metal ions which add nearly exclusively across C-C bonds of linear alkanes, the corresponding metal hydrides favour initial oxidative addition across C-H bonds7. 3.2 Reactions with Ethanol The only primary product observed for the reaction of C s ^ F e * with ethanol is CpFe(OH)+ 10 at m/z 138, reaction 3.9. C 5 H 6 F e + + C 2 H 5 O H » CpFe(OH)+ + H 2 + C 2 H 4 (3.9) m/z 122 10 m/z 138 -87-Ion 10 reacts further with ethanol according to reaction 3.10. CpFe(OH)+ + C2H5OH -> CpFe(OC 2H 3)+ + H 2 0 + H 2 (3.10) 10 m/z 139 17 m/z 164 Reactions of C5H 6Fe+ with isotopomers C H 3 C H 2 O D (EtOD, b), C H 3 C D 2 O H (EtOH-d2, c), C D 3 C H 2 O H (EtOH-d 3 , d) and C D 3 C D 2 O D (EtOH-d6, e) were used to confirm the molecular formula of product ions and to elucidate the reaction mechanism. Typical spectra are shown in Figure 3.3. Peaks labeled with asterisk are from the ion-molecule reactions of CpFe + with ethanol, as discussed in Chapter 2.2. The variation of ion abundances in the C5H6Fe+/EtOH/(C6H60)Fe(CO)3 system is displayed in Figure 3.4. O-o a L U CE c r i in CpFe + -^ o 110 a) E t O H - - C 5 H 6 F e + CpFe(OH) +-^ CpFe(OC 2 H 3 ) + -^ 120 130 140 150 160 170 180 MRSS IN R . M . U . • r> ' ' ' 1 160 170 Figure 3.3a Partial triple resonance mass spectrum illustrating the reactions of CsHoEe"1" with ethanol at delay time 120 ms. Total pressure of neutrals is 1.2 x 10 - 6 Torr. -88-Q tn z UJ r-Z cr _i UJ ec C p l V O 110 b) EtOD ^ C 5 H 6 F e + CpFe(OD)+_^ CpFe(OC 2 H 3 )+ \ 120 130 140 150 160 170 1B0 M R S S I N R . M . U . in •z LLI LT LU IT o-CpF Q 110 c) EtOH-d2 * - C 5 H 6 F e + CpFe(OH)+-1 2 0 C 5 H4D(OC 2 H3) + * C p F e ( O C 2 H 3 ) ^ J 1 3 0 1 4 0 1 5 0 M R S S I N R . M . U . 1 6 0 170 180 Figure 3.3b-c Partial triple resonance mass spectra illustrating the reactions of CsH6Fe+ with b) CH3CH2OD at delay time 100 ms and c) CH3CD2OH at delay time 100 ms. Total pressure of neutrals is 1.2 x 10"6 Torr. Peaks marked with * are from the reactions of CpFe + with alcohol. -89-CJ. Cj >-t-w z LU 1-LT _ l LU LT C p F V d) E t O H - d 3 ^ - C 5 H 6 F e + CpFe(OH)+-^ 110 CpFe(OC 2 D 3 ) + \ 1 2 0 1 3 0 1 4 0 1 5 0 M R S S IN R . M . U . 160 170 180 8-in z UJ CpFc + LU LU SE e) EtOH-dtf * - C 5 H 6 F e + CpFe(OD)+ \ 120 130 CpFe(OC 2D 5)+ \ CpFe(OC 2 D 3 ) + X L4Jb ISO 1BQ ITf. MRSS IJN _R . i l .D ISO 190. Figure 3.3d-e Partial triple resonance mass spectra illustrating the reactions of CsHoEe4" with d) CD3CH2OH at delay time 100 ms and e) C2D5OD at delay time 150 ms. Total pressure of neutrals is 1.2 x 10"6 Torr. Peaks marked with * are from the reactions of CpFe + with alcohol. -90-100 -ii 10 T3 a N s-o Z 1 -C 5 rLFe + CpFe(OH)+ (10) V* X'" X CpFe(OQH 3) + (17) CpFe(OC 2H 5) + (14) I 0 ~ l 1 100 200 Reaction Time (ms) 300 i 400 Figure 3.4 Temporal variation of ion intensities in the C5H6Fe+/EtOH/(C6H60)Fe(CO)3 system. Products of self ion-molecule reactions of C5>H6Fe+, reactions 3.1-3.3, are not shown. The data from the reaction of C s ^ F e * with isotopomers of ethanol, that are pertinent to the formation of CpFe(OH) + 10, are summarized in Table 3.3. The mass difference of 30 mass units between the adduct intermediate 8a and the observed product CpFe(OH)+ (spectrum a in Figure 3.3) may be explained by 1) a consecutive loss of hydrogen and ethene (in this order or reversed one), 2) loss of ethane, 3) loss of hydrogen and carbon monoxide, or 4) loss of formaldehyde. The last two possibilities can be excluded since they cannot account for the Am=33 between the adduct 8d and the observed product in spectrum d, and the Am=35 between the adduct 8e and the observed product in spectrum e. The Am=35 for the reaction of C5H6Fe+ with EtOH-d6 indicates that one of the hydrogens eliminated from 8e originates on the cyclopentadiene ring (last row of Table 3.3). This shows that the reacting CsHgFe"1" has a 7t-cyclopentadienyl iron hydride -91 -Table 3.3 Ionic products observed for the reaction of C 5 H 6 F e + with ethanol and its isotopologues. A+ Am = A+-P+ * inferred neutral loss C H 3 C H 2 O H / C 5 H 6 F e + 8a 168-138=30 C 2 H 6 or H 2 + C 2 H 4 CH 3 CH 2 OD/C5H6Fe + 8b 169-139=30 C 2 H 6 or H 2 + C 2 H 4 CH 3 CD 2 OH/C5H6Fe + 8c 170-138=32 C 2 H 4 D 2 or H 2 + C 2 H 2 D 2 HD + C 2 H 3 D CD 3 CH 2 OH/C5H6Fe + 8d 171-138=33 C 2 H 3 D 3 or H 2 + C 2 H D 3 HD + C 2 H 2 D 2 C D 3 C D 2 O D / C 5 H 6 F e + 8e 174-139=35 C 2 H D 5 or HD + C 2 D 4 Am is the mass difference between the adduct intermediate A+ and the observed product ion P + . structure 2 rather than cyclopentadieneiron structure. The Am=30 between the adduct intermediate 8b and the product ion CpFe(OD) + observed in spectrum b (Figure 3.3) shows that the hydroxy hydrogen is not involved in the eliminated neutrals in reaction 3.9, but remains on complex 8x (x=a, b, c, d, e). The consecutive loss of hydrogen molecule and ethene or the loss of ethane can be explained in terms of mechanism depicted in Scheme 3.2. The metal in the C5H6Fe+ ion inserts into a C-O bond of ethanol (a) giving rise to an adduct intermediate 8a. This can decompose by a competitive loss of alkane (8a—>10) or loss of water (path 8a—>9a). Transfer of hydrogen from the metal centre onto the alkyl chain ligated to the metal and subsequent reductive elimination of alkane -92-(8a—>10) has been proposed as one of the steps for catalytic hydrogenation of alkenes and alkynes in the solution8. Elimination of water molecule (8a—»9a) gives rise to cyclopentadienyliron-alkyl complex 9a at m/z 150. This complex was not observed for the reactions of C5H6Fe+ with ethanol, EtOD and EtOH-d6- However, corresponding peaks of low intensity (ca. 7% of the base peak) were observed for the reaction of CsHgFe"1" with EtOH-d2 and EtOH-d3. Peak at m/z 152 in spectrum c (Figure 3.3) was assigned to C p F e ( C D 2 C H 3 ) + ion (9c) and peak at m/z 153 in spectrum d was assigned to CpFe(CH 2CD 3)+ (9d). Another path that might lead to ion 10 is a consecutive loss of hydrogen and ethene. This could be explained by a 6-hydrogen transfer from the alkyl group onto the metal centre in complex 8a (path 8a—»11a). Subsequent elimination of hydrogen (11a—> 12a) and alkene (12a—>10) gives rise to a cyclopentadienyliron-hydroxo complex 10. However, our experimental evidence is insufficient to allow us to determine unambiguously through which of the two proposed paths the CpFe(OH) + 10 is formed. The results of the experiments with labeled ethanol relevant to the formation of the CpFe(OC 2H3) + 17, the product of reaction 3.10, are summarized in Table 3.4. -93 -I Fe H C-0 OH I Fe ' I ^ H 1 HO 8a n 6-H 11a - H 2 10 12a Scheme 3.2 Proposed mechanism for the insertion of the metal centre in the CsHgFe ion into a C-0 bond of ethanol. Proposed structures of observed ions are in a box. -94-Table 3.4 Ionic products observed for the reaction of CpFe(OH)+ with ethanol and its isotopologues. A+ Am = A+ _p+ * inferred neutral loss CH 3 CH 2 OH/CpFe(OH)+ 13a 184- 164= =20 H 2 0 + H 2 CH 3CH 2OD/CpFe(OD)+ 13b 186- 164= =22 D 2 0 + H 2 or HDO + HD CH 3CD 2OH/CpFe(OH)+ 13c 186- 164= =22 H 2 0 + D 2 or HDO + HD 186- 165= =21 HDO + H 2 or H 2 0 + HD CD 3CH 2OH/CpFe(OH)+ 13d 187- 167= =20 H 2 0 + H 2 CD 3CD 2OD/CpFe(OD)+ 13e 191- 167= =24 D 2 0 + D 2 191- 168= =23 D 2 0 + HD or HDO + D 2 Am is the mass difference between the adduct intermediate A + and the observed product ion P + . The mass difference Am=23 between the adduct intermediate 13e and the product ion at m/z 168 observed in spectrum e (last row of Table 3.4) shows that one of the neutrals eliminated from 13e contains a hydrogen atom from the cyclopentadiene ring. The other two hydrogen atoms eliminated in reaction 3.10 must therefore originate on alcohol. Their elimination may involve any combination of the insertion of the metal into a C6-H bond, C a - H bond and H-0 bond of alcohol. When C H 3 C H 2 O D reacts with CpFe(OD)+ a product ion at m/z 164 appears in spectrum b (Figure 3.3). This shows that the hydroxy deuterium is lost from the adduct intermediate 13b (Table 3.4). CpFe(OH) + reacts with C D 3 C H 2 O H to give a product at m/z 167 in spectrum d (Figure 3.3). This corresponds to -95-a loss of H2O and H2 from the adduct 13d (Table 3.4). The absence of deuterium in the eliminated neutrals indicates that either the metal does not insert into a CG-H or some scrambling takes place before the elimination. The deuterium could scramble either within the alcohol or with the hydrogens on the cyclopentadienyl ring (see Scheme 2.1 in the second chapter). The CID of ion at m/z 167 results in the only fragment ion at m/z 121. This excludes the possibility of scrambling of D on a CG carbon with the hydrogens on the cyclopentadienyl ring. The H/D scrambling could take place in the alkyl chain of alcohol. If it does we should see two kinds of ions in spectrum d, the ones where the scrambling did occur and the ones without any scrambling. This is not the case. The results of the experiments with isotopomers of ethanol illustrated in the spectra in Figure 3.3 and summarized in Table 3.4 indicate, that the CpFe(OH) + ion reacts with ethanol by inserting into both a H-0 bond and a Ca-H bond. However, the order in which the insertions take place cannot be readily distinguished since, as it is shown below, both sequences H-O—»Ca-H and Ca-H—»H-0 give rise to a product ion with the same m/z. The proposed mechanism for the insertion of the CpFe(OH)+ into a H-O bond followed by the insertion into a C a - H bond of C H 3 C D 2 O H (c) is depicted in Scheme 3.3. Oxidative addition of the metal centre across a H-O bond of ethanol followed by reductive elimination of water molecule gives rise to an intermediate 14c. Since the CpFe(OH)+ is a primary product ion, presumably it has a smaller excess of internal energy than CsHgFe"1". This is reflected by the fact that the complex 14x (x=a,b,c,d,e) can be observed at m/z 166, 166, 168, 169 and 174 in spectra a-e, respectively. Transfer of G-D from the ethoxide ligand in 14c yields complex 15c. The migration of a G-H in a metal alkoxide complex has been observed in solution9. The metal in complex 15c inserts into a C-D bond of ligated -96-HO 10 1 + Fe + CH 3 CD 2 OH H-0 1 Fe / I ^ HO „ O C D 2 C H 3 H 13c - H 2 0 Fe' n + \ 0CD2CH3 14c 1 O Fe II C / \ H 3 C D 19c 18c 15c 20c 16c C H 3 1. C-D 2. -HD _ _ 1 + Fe \ ^ ° C 1 1 C H 3 m/z 165 Scheme 3.3 Proposed mechanism for the insertion of the metal centre in theCpFe(OH)+ ion into a H-0 and a C a - H bond of ethanol. -97-aldehyde. Subsequent elimination of D 2 from 16c results in the product 17c at m/z 164 (spectrum c in Figure 3.3). Sonnenfroh and Farrar 1 0 studied the decarbonylation of acetaldehyde by Fe + and C r + . They proposed that the metal inserts into an aldehydic C-H bond. This mode of interaction, i.e. metal insertion into an aldehydic C-H bond rather than into a C-H bond of the methyl group of ligated acetaldehyde, also explains our results for the reaction of CpFe(OH) + with C D 3 C H 2 O H (d). The sequence of reactions 10—»13d—»14d (Scheme 3.3, where EtOH-d2 is replaced by EtOH-d3) gives rise to complex 15d, where the ligated aldehyde is C D 3 C H O . Insertion of the metal into an aldehydic C - H bond and subsequent loss of H2 from 16d gives rise to an observed product at m/z 167 (see Table 3.4, and spectrum d in Figure 3.3). The insertion of the metal into a C-D bond in complex 15d would result in a loss of HD and a product at m/z 166. The absence of ion at m/z 166 in spectrum d (Figure 3.3) is another evidence in support of the mechanism proposed in Scheme 3.3. As mentioned above, the same product ions could be observed if the metal insertion took place in the reversed order. A possible mechanism for the insertion of the metal into a C a - H bond followed by the insertion into a H-O bond of alcohol is displayed in Scheme 3.4. Oxidative addition of the metal into a C a - H followed by a reductive elimination of HOD molecule from complex 21c results in the formation of complex 22c. This can decompose either by a B-H transfer from alkyl group (61-H) or from hydroxy group (62-H). Because C-H bond dissociation energy is slightly lower than that of H-O bond ( D ° 2 9 8 ( H - C H 2 0 H ) ~ 9 6 kcal/mol vs. D°298(H-OC 2 H 5 )~104 kcal/mol1 1) we would expect 6 1 - H transfer to be favoured. Path Gi-H-»25c-»26c (Scheme 3.4) would explain the presence of the ion at m/z 165 observed in spectrum c (Figure 3.3). However, our results with C D 3 C H 2 O H (d) indicate that the Bi-H transfer in 22d does not take place. Bi -H transfer in complex 22d (Scheme 3.4, where EtOH-d2 is replaced by EtOH-d3) and -98-Fe 1 + + C H 3 C D 2 O H 1 + H O 10 25c I H-0 "I + • H 2 1' 1 + 1 O Fe -— II / C / \ H 3 C 23c |c-D D 1 + -HD m/z 165 m/z 164 Scheme 3.4 Proposed mechanism for the insertion of the metal centre in the CpFe(OH)+ ion into a C a - H and a H-0 bond of ethanol. -99-subsequent insertion of the metal centre in complex 25d into a H-O bond would result in the loss of HD molecule from 26d and an ion at m/z 166 in the spectrum d (Figure 3.3). This was not observed. Because of the above discussed inconsistency between our experimental observations and the mechanism in Scheme 3.4 we propose that the formation of CpFe(OC2H3)+ ion in reaction 3.10 be explained in terms of a mechanism in Scheme 3.3. That is, first the insertion into a H-O bond of alcohol takes place followed by the insertion into a C a - H bond. It remains to explain the presence of ions at m/z 165 and 168 in spectra c and e, respectively. For the reaction of C2H5OD no ions due to H/D scrambling were observed. From this we can conclude that the H/D scrambling does not take place in the adduct intermediate 13x (x=b,c,d,e Scheme 3.3). From the data in Table 3.4 it appears that H/D scrambling takes place whenever there is deuterium on C a carbon (i.e. CH3CD2OH and CD3CD2OD). This means that the scrambling can take place only after the C a deuterium has migrated to the metal centre, i.e. when complex 15c or 15e has been formed. The H/D scrambling between the deuterium on the metal centre and the hydrogens on the cyclopentadienyl ring, as explained in detail in Chapter 3.1, gives rise to complex 20c or 20e (Scheme 3.3). Final elimination of HD molecule from 20c or 20e results in the formation of ion at m/z 165 and m/z 168, respectively. The CID of ions at m/z 165 and m/z 168 should give a fragment at m/z 122 that would correspond to C 5 H 4 D F e + . Unfortunately, the intensity of these ions is too low for this experiment to be done. - 100-3.3 Reactions with i-Propanol For didactic purposes we will first describe the reaction of isopropanol with C5H6Fe + and then discuss the C5H6Fe + /n-C3H70H system. CsHgFe"1" reacts with /-C3H7OH to give three ions, CpFe(OH)+ 10 at m/z 138, CpFe(C3H5)+ 34 at m/z 162 and CpFe(OC 3H 5)+ 39 at m/z 178 (reactions 3.11-3.13). C 5 H 6 F e + / - C 3 H 7 O H 60% 30% 10% CpFe(OH)+ + CpFe(C 3 H 5 ) + + CpFe(OC 3 H 5 ) + + C 3 H 8 H 2 0 + H 2 2H-, (3.11) (3.12) (3.13) The CpFe(OH) + 10 reacts with i-propanol according to reactions 3.14a and 3.14b. CpFe(OH)+ + / - C 3 H 7 O H 10 m/z 138 CpFe(OC 3 H 7 ) + 36 m/z 180 CpFe(OC 3 H 5 ) + 39 m/z 178 + H-,0 H , 0 + Hn (3.14a) (3.14b) The products of reaction of C s ^ F e * with j'-propanol and its isotopomers J-C3H7OD (b), j-(CD3)2CHOH (c) and J - C 3 D 7 O D (d) are illustrated in the spectra in Figure 3.5. Figure 3.6 shows the variation of ion abundances with time for C5H6Fe+ with 1.0 x 10"6 Torr of i-propanol. - 101 -CpFe+ \ ^ C S H 6 F e + CpFe(OC 3 H 5 ) + -^ C PFe(OH) + CpFe(C 3 H 5 ) + X r T 120 l«iO 160 MOSS IN P . M . U . 180 200 C p F e ( O C 3 H 5 ) + - ^ C 5 H 6 F e + Cpl-e+ X CpFe(C 3 H 5 ) + CpFe(OD)+ "X \ 120 U O J,.it. 160 MRSS IN R . M . U . 180 200 Figure 3.5a-b Partial triple resonance mass spectra illustrating the reactions of CsH6Fe+ with a) 1 -C3H7OH at delay time 150 ms, b) 1 - C 3 H 7 O D at delay time 130 ms. Peaks marked with * are from the reactions of CpFe + with alcohol. - 102-CpFc + c) i-PrOH-d6 - C 5 H 6 F e + CpFe(OC 3 D 5 ) + X -«-CpFe(OH) + CpFe(C 3 D 4 H) + 'J, I ' '—i—i r 'i i i 120 140 160 180 MOSS IN R . M . U . 200 Cple + \ - C 5 H 6 F e + d) i-PrOH-d8 C 5 H 4 D ( O C 3 D 5 ) + - ^ CpFe(OD)+ \ CpFe(C3D5)+_ fe T—*j 't T ' i r r ' ]' i r *r i | — i — i i T 11—J— 120 140 160 MRSS IN R . M . U . 180 200 Figure 3.5c-d Partial triple resonance mass spectra illustrating the reactions of C5>H6Fe+ with c) HCD3)2CHOH at delay time 100 ms, d) 1-C3D7OD at delay time 150 ms. Peaks marked with * are from the reactions of CpFe + with alcohol. - 103-100 -C 5 FLFe + CpFeCOQrL/ (39) 1 -0 100 200 300 Reaction Time (ms) 400 500 Figure 3.6 Tempora l variation of ion intensities in the C5H6Fe+/i-C3H70H/(C6H60)Fe(CO)3 system. Products of self ion-molecule reactions of C5H.6Fe+, reactions 3.1-3.3, are not shown. formation of CpFe(C3H.5)+ 34, the product of reaction 3.12, are summarized in Table 3.5. The product ion at m/z 162 observed in spectrum b (Figure 3.5) shows that hydroxy deuterium is eliminated from the adduct intermediate 27b (see Table 3.5). No peaks due to H/D scrambling were observed in spectrum b. With the deuterium on the alkyl chain the ions that were assigned to be due to H/D scrambling begin to emerge. They appear at m/z 165 and 167 in spectrum c and at m/z 166 and 168 in spectrum d (Figure 3.5). For the reaction of CsH6Fe+ with /-C3D7OD the ion at m/z 167 corresponds to a loss of HDO + D 2 or D 2 0 + HD (the order may be reversed) from the adduct intermediate 27d. This shows that one of the hydrogen atoms in the eliminated neutrals is from the cyclopentadiene ring The results of the experiments with labeled isopropanol that are relevant to the - 104-Table 3.5 Formation of CpFe(C 3H5) + observed for the reaction of C 5 H 6 F e + with /-propanol and its isotopologues. A+ Am = A+-P+ * inferred neutral loss (CH 3 )2CHOH/C 5 H 6 Fe + 27a 182- 162=20 H 2 0 + H 2 (CH 3)2CHOD/C 5H6Fe+ 27b 183- 162=21 HDO + H 2 or H 2 0 + HD (CD 3 )2CHOH/C 5 H 6 Fe + 27c 188- 166=22 H 2 0 + D 2 or HDO + HD 188- 165=23 HDO + D 2 188- 167=21 H 2 0 + HD or HDO + H 2 ( C D 3 ) 2 C D O D / C 5 H 6 F e + 27d 190- 167=23 HDO + D 2 or D 2 0 + HD 190- 166=24 D 2 0 + D 2 190- 168=22 HDO + HD or D 2 0 + H 2 * Am is the mass difference between the adduct intermediate A+ and the observed product ion P + of the reactant ion. The loss of water and hydrogen molecules and the observed H/D scrambling can be explained by the mechanism in Scheme 3.5, where the reaction of C5H6Fe+ with i'-PrOH-d6 (c) is used as an example. Insertion of the metal in the C s ^ F e * ligand into a C-O bond of alcohol gives rise to an intermediate complex 27c. Similarly as in Scheme 3.2, 27c can decompose by a competitive loss of alkane or loss of water. In the former case, reductive elimination of propane, C3D6H2, from 27c yields directly ion 10. Transfer of a hydrogen from the metal centre onto the alkyl chain ligated to the metal and subsequent reductive elimination of - 105 -alkane has been proposed as one of the steps for catalytic hydrogenation of alkenes and alkynes in the solution8. In the latter case, reductive elimination of water gives complex 28c. Subsequent 15-D transfer in complex 28c and insertion into an allylic C - D bond (29c—>30c) lead to a loss of D 2 and result in the formation of ion 34c at m/z 166 (spectrum c in Figure 3.5). It is possible that H/D scrambling takes place in steps 28c<-»29c. This would result in loss of HD in step 30c—>34c and generation of ion at m/z 167. The path 27c—»28c—>29c—»30c—»34c, while not unreasonable, cannot completely account for the observed H/D scrambling. Therefore, we propose that the elimination of water (27c—»28c) and alkane (27c—> 10) is preceded by a 6-H transfer that yields an intermediate 31c. The H/D scrambling takes place in steps 27c<->31c and is described in detail in Scheme 3.6 below. With the postulated presence of the intermediate 31c a second pathway that could explain the observed product ions 10 and 34 emerges. Loss of hydrogen molecule from 31c, that is ca. 20 kcal/mol less endothermic than loss of water, results in complex 32c. Further reactions of 32c are similar to those described in the previous chapter in Scheme 2.2. Competitive loss of alkene (32c—»10) or water (32c—> 34c) gives rise to the observed product ions CpFe (OH) + at m/z 138 and CpFe(C3D4H)+ at m/z 166, respectively (spectrum c in Figure 3.5). The allyl ligand in complex 34x (where x=a, b, c, d) can form either a o bond (rji -CsHs) or a % bond (rj3-C3H5) with the metal. It is very likely that after the loss of water the allyl undergoes T)1—>rj3 rearrangement. This would result in the increase of the number of electrons on the metal by two and an increased stability of the product ion. - 106-+ -C3D6H2 28c 6-D 29c allylic C-D D C D r C< CD2 \ 30c H 27c -HD C D 3 31c C 3 D 5 H n + C D 2 HO H C D 3 32c allylic C-D C D 2 HO ' CD 2 -C D \ 33c H 34c m/z 166 Scheme 3.5 Insertion of the metal centre in the C5H6Fe+ ion into a C-0 bond of i-propanol. - 107-Scheme 3.6 explains in more detail the H/D scrambling that takes place in steps 27c<->31c (Scheme 3.5). The hydrogen ligated to the metal centre (H) in the intermediate complex 31c migrates back onto the alkyl chain giving rise to 27c-H. Depending on a lifetime of complex 27c-H, there are two possible paths that can lead to ion at m/z 165. First, 27c-H may react along the path 27c-H->28c->29c-»30c—>34c (as described in Scheme 3.5) that corresponds to a consecutive loss of HDO, 8-D transfer, insertion of the metal centre in complex 29c into an allylic C-D bond and loss of D 2 . Second, if the lifetime of complex 27c-H is long enough, it may rearrange again into a complex 31c-H by a 6-D transfer from the alkyl chain to the metal as shown in Scheme 3.6. Ensuing loss of D 2 from 31c-H gives complex 32c-H. Insertion of the metal centre into an allylic C-D bond and subsequent loss of HDO yields the observed ion 34c-H at m/z 165. The scrambling of deuterium ligated to the metal centre (D) in 31c can be described by a similar sequence of reaction steps (31c-»27c-D—»31c-D—>32c-D, Scheme 3.6). This gives ion 34c-D at m/z 167 (spectrum c, Figure 3.5). To reduce cluttering the loss of alkene from complexes 32c, 32c-H and 32c-D that generates ion 10 is not shown in Scheme 3.6. In the /-C3D70D/C5H6Fe+ system, the above mechanism proposed for H/D scrambling between the metal centre and the alkyl chain of alcohol accounts for the presence of ion at m/z 166 in spectrum d (Figure 3.5, see also Table 3.5). However, it cannot explain the ion at m/z 168 in spectrum d. As can be seen from Table 3.5 the neutrals eliminated from 27d contain two hydrogen atoms. They must both originate on the cyclopentadiene ring of the reactant ion, since z'-propanol is perdeuterated. The H/D scrambling between the metal centre and the cyclopentadiene ring, as described in Scheme 3.1, must therefore take place, in addition to the H/D scrambling between the metal centre and alkyl chain. - 108-~ i H 1 C D 2 - C H D - C D 3 HO 27c-D H 1 H C D 3 D HO 31c-D 32c-D 1. allylic C-D 2. -HDO ~ i + 1 1 Fe \ ^ C D 2 C D 2 - C ^ D 34c-D m/z 167 D bJO a o a -HD I C D 2 HO C D 3 H 32c H C D 2 H C D 3 D 1 D C D 2 H H HO 32c-H 1. allylic C-D 2. -HDO 34c-H m/z 165 Scheme 3.6 Proposed mechanism for H/D scrambling observed in the C5H6Fe +//-PrOH system. Proposed structures of observed ions are in a box. -109-Table 3.6 summarizes the results from the experiments with labeled i'-propanol pertinent to the formation of CpFe(OC3H5)+ 39, the product of reaction 3.13. Table 3.6 Formation of CpFe(OC3H5)+ observed for the reaction of C 5 H6Fe + with i-propanol and its isotopologues. A + Am = A+-P+ * inferred neutral loss (CH 3 )2CHOH/C 5 H6Fe + 35a 182-178=4 2 H 2 (CH 3 )2CHOD/C 5 H 6 Fe + 35b 183-178=5 HD + H 2 (CD 3 )2CHOH/C 5 H6Fe + 35c 188-183=5 HD + H 2 188-184=4 2 H 2 188-182=6 HD + HD or H.2 + D 2 (CD 3 )2CDOD/C 5 H 6 Fe + 35d 190-184=6 HD + HD or H 2 + D 2 190-183=7 HD + D 2 * Am is the mass difference between the adduct intermediate A+ and the observed product ion P + . As can be seen from Table 3.6, the CpFe(OC3H5)+ ion is formed by a double dehydrogenation of adduct intermediate 35x (where x=a, b, c, d). Experiments with ( C H 3 ) 2 C H O D and ( C D 3 ) 2 C D O D show that a hydroxy hydrogen and a hydrogen atom from the cyclopentadiene ring of the C5H6Fe+ ion are involved in this process (see Table 3.6). From the presence of an intense peak at m/z 183 in spectrum c in Figure 3.5 (see also Table 3.6) we conclude that the other two hydrogens originate on C a and CG carbons. Low intensity peaks at m/z 184 and m/z 182 observed in spectrum c as well as a peak at m/z 184 in spectrum d indicate that some H/D scrambling takes place in this system. - 110-The data in Table 3.6 can be accounted for by a mechanism presented in Scheme 3.7, where a reaction of Cs^Fe"1" with (CD3)2CHOH (c) is used as an example. 1 + Q5~ H-0 + (CD3)2CH-OH • Fe / I \ / H u OCH^ CD3 2: L=H 10: L=OH H 35c m/z 188 CD3 -HD C D 2 C // \ O C D 3 C D 3 39c m/z 183 38c -LH C-D 1 + OCH \ 36c m/z 186 C D 3 6-H 37c Scheme 3.7 Proposed mechanism for the insertion of a metal centre in the CsH^Fe4" and CpFe(OH)+ ions into a H-0 bond of /-propanol. Proposed structures of observed ions are in a box. The insertion of the metal centre of the reactant ion into a H-O bond of i-propanol followed by a reductive elimination of first H 2 gives an intermediate complex 36c at m/z 186. This complex can be observed in spectrum c. The migration of a C a hydrogen in 36c from the alkoxide ligand onto the metal results in the formation of a ketone ligand in complex 37c. The metal in complex 37c inserts into a C-D bond of ligated acetone. Ensuing loss of a HD molecule yields a product ion 39c at m/z 183. The observation of - I l l -two peaks in spectra c and d in Figure 3.5 suggests that H/D scrambling occurs. We propose that this H/D scrambling takes place in steps 36c<->37c. The H/D scrambling in the alkyl chain would result in the loss of D 2 in step 38c—»39c and a product at m/z 182. To explain the ion at m/z 184 in spectra c and d, we have to invoke a H/D scrambling in the cyclopentadiene ring as described in Scheme 3.1 (3c-2—>4c—>5c). It is rather unexpected that the ion at m/z 184 formed by this type of H/D scrambling in spectrum d has much higher intensity than the ion at m/z 183. The CpFe(OC3H5) + 39 is also a product of a secondary reaction of CpFe(OH)+, reaction 3.14b. Its formation can be explained in terms of the mechanism in Scheme 3.7, where 2 is replaced by 10, CpFe(OH) +, and water is lost instead of hydrogen in the first elimination step 35c—>36c. Since the CpFe(OH) + is a product of a primary reaction of C5H6Fe+, presumably, it has a smaller excess of internal energy as compared to C5H6Fe+. This is reflected in the fact, that in reaction 3.14 that proceeds through the mechanism in Scheme 3.7, we can also observe an intermediate complex 36x (where x=a, b, c, d) at m/z 180, 180, 186 and 187 in spectra a, b, c and d in Figure 3.5 (see also Figure 3.6). 3.4 Reactions with n-Propanol Reaction of n-propanol with C5H6Fe+ affords three ions, CpFe(OH) + 10 at m/z 138, CpFe(C 3H 5)+ 34 at m/z 162 and Cp(OC 3H 5)+ 44 at m/z 178 (reactions 3.15-3.17). C 5 H 6 F e + n - C 3 H 7 O H 80% 10% 10% CpFe(OH)+ + H 2 + C 2 H 6 (3.15) CpFe(C 3 H 5 ) + + H 2 + H 2 0 (3.16) CpFe(OC 3 H 5 ) + + 2 H 2 (3.17) - 112-Partial triple resonance spectra that illustrate the reaction of CsHgFe"1" with H-C3H7OH (a) and n-CsHjOD (b) are shown in Figure 3.7. The products of reactions 3.16 and 3.17 are stable within the studied time scale. Multiple resonance experiment confirmed that the CpFe(OH) + 10 is consumed in a secondary reaction with n-propanol, reactions 3.18a and 3.18b. CpFe(OH)+ + n-PrOH 10 m/z 138 CpFe(OC 3 H 7 ) + + H 2 0 42 m/z 180 (3.18a) CpFe(OC 3 H 5 ) + + H 2 0 + H 2 44 m/z 178 (3.18b) Figure 3.8 shows the variation of ion abundances with time for CsHgFe"1" with 1.0 x 10"6 Torr of n-propanol. - 113-o tn z UJ > cr _J UJ cc Cpl-"e+-^ -C 5 H 6 Fe+ CpFe(OC 3 H 5 ) + -^ -«-CpFe(OH) + CpFe(C 3 H 5 ) + a) n-PrOH 110 120 130 140 150 160 170 180 190 200 MRSS IN R . M . U . D_ O LO z u LT IS) CpFe+-0 rt 110 120 - * - C 5 H 6 F e + -«-CpFe(OD) + CpFe(C 3 H 5 ) + \ i T T ' T ' J i i i i [ t i T i ' j - i ' i f i | t i T i -j I ' I f i | 11 T i * | T i n ' ; V n r ' | f i 11 | i r i 130 140 150 160 170 MRSS IN R . M . U . 180 190 200 Figure 3.7 Triple resonance mass spectra illustrating the reaction of CsHgFe"1" with a) /X-C3H7OH at delay time 150 ms b) n-C3JH70D at delay time 90 ms. The total pressure of neutrals is 1.2 x 10'6 Torr. Peaks marked with * are from the reactions of CpFe + ion with alcohol. - 114 -100 -C cu • - i 10 -O cu a © Z 1 -C 5 rLFe + CpFe(OC3H5)+ (44) CpFe(OC3H7)+ (41) -E CpFe(C 3H 5)+ (34) ^ I 200 400 Reaction Time (ms) 1^  600 I 800 Figure 3.8 Tempora l variation of ion intensities in the C5H6Fe+/n-PrOH/(C6H60)Fe(CO)3 system. Products of self ion-molecule reactions of C5H.6Fe+, reactions 3.1-3.3, are not shown. The presence of ions 10 and 34 formed in reactions 3.15 and 3.16 can be accounted for by a mechanism in Scheme 3.5, where z'-propanol is replaced by n-propanol. The intermediate 31x (x=a,b) has the same structure for both n-propanol and j'-propanol. Hence we assume that all reaction steps that follow its formation are identical for n-propanol and /-propanol and yield product ions with the same structure in both systems. A small amount of the CpFe(OC3H5) + 44 is formed in a primary reaction of C5H6Fe+, reaction 3.17, but most of it, together with the CpFe(OC3H7)+ 41 is formed in a secondary reaction of CpFe(OH)+, reactions 3.18a and 3.18b. The generation of 41 and - 115-44 can be explained by the insertion of the Cs^Fe-1" or CpFe(OH)+ into a H-O bond of n-propanol as shown in Scheme 3.8 (R=CH3). i Fe + «-C 3 H 7 OH L 2 L=H a 10 L=OH CO H-O Fe ' I ^ L R OCH 2 CH 2 R 40a -LH CH 2 -R 44a-l, R=CH 3 62a-l, R = C H 2 C H 3 CH 2 -R 43a-l i + Fe \ OCH 2 CH 2 R 41a m/z 180 8-H I Fe 1 + O H C / \ C H 2 H I R 42a 44a-2, R=CH 3 62a-2, R=CH 9 CH 3 43a-2 Scheme 3.8 Proposed mechanism for the insertion of the metal in the Cs^Fe"1" 2, and CpFe(OH)+ 10 into a H-0 bond of n-propanol, R=CH 3 ; n-butanol, R = C H 2 C H 3 . - 116-Reductive elimination of L H neutral, (LH = H 2 for the reaction of ion 2, L H = H 2 O for the reaction of ion 10) yields ion 41a. This reacts further by a B-H transfer from the alkoxide ligand that results in the formation of an intermediate 42a. The metal in 42a may insert either into an aldehydic C-H bond (path 1) as was observed for a similar reaction with ethanol (see Scheme 3.3), or it may insert into a C8-H bond (path 2). Subsequent loss of the second hydrogen molecule yields a product ion 44a at m/z 178 (spectrum a) that can have either structure 44a-1 or 44a-2 depending on into which of the above mentioned C - H bonds the metal inserts. The experiments with C H 3 C H 2 C D 2 O H and/or C H 3 C D 2 C H 2 O H that could unambiguously identify the structure of 44 were not performed. The metal insertion into a CB-H bond in complex 42a is expected to be more facile than in complex 15c, where it did not occur at all (Scheme 3.3), since CB is a secondary carbon in the former but a primary carbon in the latter. 3.5 Reactions with n-, i-, l-Butanol Two product ions were observed for the reaction of the C s ^ F e * with n-butanol, the CpFe(OH)+ 10 at m/z 138 and the CpFe(C4H7)+ 50 at m/z 176 (reactions 3.19 and 3.20, respectively). C 5 H 6 F e + + n - C 4 H 9 O H \90%*> CpFe(OH)+ + H 2 + C 4 H 8 (3.19) 10% 4- CpFe(C 4 H 7 ) + + H 2 + H 2 0 3.(20) The products of the above reactions with n-butanol and its isotopomers CH3(CH2)30D (b) and C D 3 ( C D 2 ) 3 O H (c) are shown in the mass spectra in Figure 3.9. - 117-Reaction of C5H.6Fe+ with /-butanol yields two products, CpFe(OH)+ 10 at m/z 138 and CpFe(C4H7)+ 56 at m/z 176 (reactions 3.21 and 3.22). Figure 3.10 shows a triple resonance mass spectrum that illustrates these reactions. C 5 H 6 F e + + i - C 4 H 9 O H CpFe(OH)+ + H 2 + C 4 H 8 (3.21) 20% • CpFe(C 4 H 7 ) + + H 2 + H 2 0 (3.22) Equations 3.23 and 3.24 summarize reactions of r-butanol induced by CsH6Fe + that yield product ions CpFe(OH)+ 10 at m/z 138 and CpFe(C4H7)+ 56 at m/z 176. Mass spectrum in Figure 3.11 illustrates these reactions for f-butanol and (CD3)3COD. C 5 H 6 F e + + ?-C 4 H 9 OH \ 5 5 % * CpFe(C 4 H 7 ) + + H 2 + H 2 0 (3.23) CpFe(OH)+ + H 2 + C 4 H 8 (3.24) The variation of ion abundances with time in the C5H6Fe+/n-, /-, r-butanol systems is shown in Figures 3.12-3.14, respectively. Similarly as in the ethanol, n-propanol and /-propanol systems, multiple resonance experiments confirmed that the CpFe(OH) + 10 reacts further with n-butanol, /-butanol and r-butanol according to reaction 3.25. In addition, n-butanol and /-butanol react with C5H6Fe+ to give also the product of reaction 3.26. CpFe(OH)+ + C 4 H 9 O H 10 m/z 138 CpFe(OC 4 H 7 ) + + H 2 + H 2 0 (3.25) 62, 63 m/z 192 CpFe(C 3 H 5 ) + + C H 3 O H + H 2 0 (3.26) 61 m/z 162 - 118-CpFe+ a) n-BuOH ^ C 5 H 6 F e + CpFe(OC 4 H 7 ) + _^ « - C P F e ( O H ) + CpFe(C 3 H 5 ) + -* ft u v U. a. U 120 140 160 MRS5 IN R . M . U . b) n-BuOD l BO 2 0 0 CpFe+ \ ^ C 5 H 6 F e + CpFe(OD)+ CpFe(C 3 H 5 ) + X 120 140 160 MRSS IN R . M . U . Figure 3.9a-b Partial triple resonance mass spectra taken at delay time 150 ms illustrating the reactions of C5>H6Fe+ with a) n-BuOH and b) n-BuOD Peaks marked with * are from the reactions of CpFe + with alcohol. - 119-in 2 C J -,in i-a UJ c) n-BuOH-dp C S H 6 F e + C p F e ( O C 4 D 7 ) + - ^ CpF« \ CpFe(OH)+ * + Q u U. 120 14 0 160 180 200 MRSS IN R . M . U . Figure 3.9c Partial triple resonance mass spectrum taken at a reaction delay time 150 ms illustrating the reactions of C5H.6Fe+ with n-BuOH-d9. in IU U l u > UJ CpFe+ i rwn 120 i -BuOH * - C s H 6 F e + -*-CpFe(OH)+ CpFe(C 4 H 7 ) + \ CpFe(C 3 H 5 ) + \ CpFe(OC 4 H 7 ) + U i O 160 MRSS IN R . M . U . 180 200 Figure 3.10 Partial triple resonance mass spectrum taken at a delay time 120 ms illustrating reactions of C5HoFe+ with i-BuOH. - 120 -a. a 2 Ld LT o-C p R + \ * - C 5 H 6 F e » C p F e ( C 4 H 7 ) + - ^ CpFe(OH)+ V 1 2 0 1 4 0 1 6 0 1 8 0 2 0 0 MRSS IN R . M . U . a) t-BuOH CpFe(OC 4H 7)+ 220 z UJ UJ LT Cpl< ^ C 5 H 6 F e + CpFe(C 4 D 6 H) + _^ CpFe(OD)+ 120 140 b) t-BuOH-dio -*-CpFe(C 4D 7)+ 160 180 MRSS IN R . M . U . 200 220 Figure 3.11 Partial triple resonance mass spectra illustrating reactions of Cs^Fe"1" with a) f-BuOH at delay time 120 ms and b) f-BuOH-dio at delay time 100 ms. Peaks marked with * are from the reactions of CpFe + with alcohol. - 121 -I 0 C 5 rLFe + CpFe(OC 4H 7) + (62) L v .X-. C p F e ^ H g f (61) -CpFe(OH)+ (10) CpFe(C 4H 7)+ (50) ~ l 1 1 1— 100 200 300 400 Reaction Time (ms) 500 Figure 3.12 Tempora l variation of C 5H6Fe+/n-BuOH/(C6H 60)Fe(CO)3 system. 7 I 600 ion I 700 intensities in the 100 -S * H 10 ns <u S t-o Z 1 -CpFe(OC 4H 7) + (63) / CpFe(OH)+ ( 1 0 ) V r * I "T " 1 ~"b-CpFe(C 4H 7) + (56) CpFe(C3H5)+ (61) T 0 100 200 300 Reaction Time (ms) ion Figure 3.13 Tempora l variation of C5H6Fe+/;-BuOH/(C6H60)Fe(CO)3 system. T 400 intensities T 500 in the - 122 -100 E \ CpFe(C 4H 7)+ (56) B © CpFe(OH)+ (10) 0 200 400 Reaction Time (ms) 600 800 Figure 3.14 Tempora l variation of ion intensities in the C5H6Fe+/?-BuOH/(C6H60)Fe(CO)3 system. The formation of ions 10 and 50 in the CsHoPe+M-butanol system and ions 10 and 56 in the C5H6Fe+/i-butanol and C5H6Fe+/?-butanol systems can be explained by the C-O insertion/6-H transfer sequence. Scheme 3.9 shows the details of this mechanism for «-CD3(CD2)30H (c). Insertion of the C5H6Fe+ into a C-0 bond generates complex 45c that may react further either by a loss of butane to give ion 10, or loss of water to give ion 47c. A 6-D transfer from a butyl ligand in 47c and subsequent insertion of the metal centre in complex 48c into an allylic C-D bond lead to a loss of D 2 molecule and formation of a product 50c observed at m /z 183 in Figure 3.9c. Similarly to the reactions of deuterated /-propanol (see Scheme 3.6), ions due to H/D scrambling have been observed with n-CD3(CD2)30H and ?-(CD3)3COD. As a result of H/D scrambling of the hydrogen on the metal centre and deuteriums in the alkyl chain of alcohol that takes place in steps 45c-»46c a 45c-H ion is generated. - 123 -2 + n-C4D9OH c C-0 50c 50c-H m/z 183 m/z 182 Scheme 3.9 Proposed mechanism for the reaction of C 5 H 6 F e + with n-butanol. - 124-The decomposition of this ion is identical to that of ion 45c and gives a product ion 50c-H observed at m/z 182 in spectrum c in Figure 3.9. Scheme 3.10 explains the insertion of the C5H.6Fe+ into a C-0 bond of /-butanol and ^-butanol, where the reaction with r-C4D90D (b) is used as an example. The intermediate complex 54b has the same structure for both alcohols therefore the product ion 56b at m/z 183 has the same structure in both systems. An ion 56b-H at m/z 182 observed in spectrum b (Figure 3.11) is due to H/D scrambling between the hydrogen on the metal centre and the deuterium on the alkyl chain of f-butanol that takes place in step 51b—>52b and generates complex 51b-H. The product of reaction 3.26, CpFe(C3H5)+ is formed by the insertion of the metal in the CpFe(OH) + ion into a C1-C2 bond of n-butanol and /-butanol as displayed in Scheme 3.11. Ions that would result from the insertion of CsH6Fe + into a C-C bond of ^-butanol were not observed. -125-CD 2H 55b-H 56b m/z 183 56b-H m/z 182 Scheme 3.10 Proposed mechanism for the reaction of C 5 H 6 F e + with r-butanol. - 126-+ H O C-C ~ i + C H 2 - O H 6-H 57a H O 58a -CH3OH 61-ry Scheme 3.11 Proposed mechanism for the insertion of the CpFe(OH) + into a C-C bond of /-butanol. Proposed structures of observed ions are in a box. The formation of the CpFe(OC4H7)+ ion, product of reaction 3.25, can be explained by the insertion of the CpFe(OH)+ ion into a H-O bond of n-butanol (ion 62) and /-butanol (ion 63) as illustrated in Scheme 3.8 (where the reactant ion 2 is replaced by 10). 3.6 Summary and Conclusions Primary product ions for the reaction of Cs^Fe"1" with methanol, ethanol, n-, i-propanol and n-, r-butanol are listed in Table 3.7. In all systems but methanol, the only product ions observed result from the insertion of the CsHgFe"1" into a C-O bond of alcohol. For the reaction of CsHgFe"1" with n- and i-propanol, a small percentage of ions -127-due to the insertion of the metal centre into a H-O bond of alcohol was observed as well. The absence of adduct ions C5H6Fe(ROH)+ in MeOH, EtOH and n-, /-PrOH and n-, i-, t-BuOH systems indicates that the lifetime of the adduct intrermediate [C5HgFe(ROH)+]* is shorter than that of the [CpFe(ROH)+]* intermediate. Principal structural difference between the [C5HgFe(ROH)+]* and the [CpFe(ROH)+]* is presence of a hydrogen atom on the metal centre in the former ion. Due to this hydrogen atom the rate of reductive elimination of water (if the metal inserts into a C-O bond of alcohol) is faster than the rate of radiative stabilization of the [C5H 6Fe(ROH) +]* and the adduct is not observed. The CpFe(H20) + and the CpFe(C n H2 n ) + ions observed for the reactions of CpFe + ion with alcohols are formed from a common intermediate by a competitive loss of alkene or water, respectively. In the C5H6Fe +/ROH system, the CpFe(OH) + and the CpFe(C n H2 n -l) + ions are their analogues. However, in addition to the reductive elimination of water or alkene, loss of a hydrogen molecule is observed, that can either precede or follow the loss of water/alkene as it is depicted in Schemes 3.2 and 3.5. Reductive elimination of hydrogen can be explained if we postulate that the reactive form of the C5H6Fe + ion has a Tt-cyclopentadienyl iron-hydride structure 2, rather that cyclopentadiene structure 1, formed by the equilibrium below. H 1 2 If the C5H6Fe + ion did not have a metal-hydrogen bond we would expect to observe the same products as in the CpFe + /ROH system with a m/z ratio one mass unit - 128-Table 3.7 Primary product distributions for the ion-molecule reactions of CsfJ^Fe"1" with alcohols. C-O H-O H-O/C-H bond insertion bond insertion bond insertion CpFe(OH)+ C p F e ( C „ H 2 n - i ) + CpFe(OC n H 2 n - i ) + CpFe(OC n H 2 n + i )+ MeOH 100 7 EtOH 100 10 n-PrOH 80 10 10 34 10 44 z-PrOH 60 10 30 34 10 39 n-BuOH 90 10 10 50 Z-BuOH 80 10 20 56 f-BuOH 55 10 45 . 56 The upper number of each pair is the percentage of the total ion products. The lower bold number is the compound number. A blank indicates that the product ion was not observed. - 1 2 9 -higher. The C n H 2 n - i ligand in the CpFe(C n H2 n -i) + ion is allyl in C 3 alcohols, 3-methyl allyl in n-butanol and methyl allyl in /-, ^ -butanol. The allyl ligand can form either a a-bond (r^-aUyl) or a K-bond (r|3-allyl) with the metal. We propose that it forms a 7t-bond. This would decrease the electron deficiency on the metal centre by two and contribute to the increased stability of the product ion. For all the reactions of C5H6Fe+ with alcohols, i.e. reactions 3.7, 3.9, 3 .11-3 .13, 3 .15-3.17, 3 .19 and 3.20, 3.21 and 3.22, 3.23 and 3.24, the total rate constant K R O H was determined from the slope of the log(normalized intensity) vs. reaction time. The procedure is described in detail in Chapter 1.3.3. Similarly as for the reactions of alcohols with CpFe + ion, the decay plot of the reactant ion C5H6Fe+ has a curvature that is most obvious for the reaction with methanol (Figure 3.1) and disappears as the size of alcohol increases. In all cases where the bend in the decay plot of CsHeFe4" ion was observed, the rate was determined from the linear portion of the plot at long delay times. The experimental rate constant ICROH* as well as theoretical rate constants ICL and 1CADO> calculated from the Langevin theory12 and from the Average Dipole Orientation Theory (ADO) 1 3 , respectively, are listed in Table 3.8. The fraction of reactive collisions (reaction efficiency) was calculated as kROH/kADO- Figure 3 .15 shows the experimental rate constants listed in Table 3.8 plotted as a function of the length of alkyl chain of alcohol. The C5H6Fe + exhibits similar trends in reactivity towards alcohols as the CpFe + (Figure 2 .17). That is, the reactivity increases with the increasing length of the alkyl chain of alcohol. There is a striking difference, however, when the reactivity of CpFe+ and C5H6Fe+ ions towards methanol is compared. The C5H6Fe+ ion shows a twofold increase in the reaction efficiency. More than 2 0 % of the collisions of C5H6Fe+ with methanol lead to the reaction, whereas in the CpFe+ system, only 1 0 % of the collisions are reactive. As discussed in Chapter 2 .1 , our results indicate that the CpFe4" forms only an ion-neutral - 130-Table 3.8 Experimental and theoretical rate constants for the reaction of C5H6Fe+ with alcohols. Alcohol reaction kROH x 10 1 0 k L x l 0 1 0 kADO x 10 1 0 efficiency MeOH EtOH n-PrOH i-PrOH n-BuOH /-BuOH r-BuOH 3.7 3.8 6.1 7.6 7.7 7.3 6.9 8.4 9.4 9.6 9.6 10.3 10.3 10.3 16.0 15.6 15.0 15.0 15.0 15.0 15.0 0.23 0.24 0.41 0.51 0.51 0.49 0.46 All k values are in units cm3molecule _ 1s _ 1 . Reaction efficiency was calculated as kRQH/kADO-A n-A i- • • t-* * * * * * "a* 3 u mm © S "a 7-6-5 -4 -X o 0 1 2 3 4 Number of carbons in alcohol Figure 3.15 The trend in reactivity of CI to C4 alcohols with CsHgFe"1" ion. - 131 -adduct with methanol, no neutral loss was observed. Therefore, most of the excited (CpFe(CH30H)+)* complex reverts back to the reactant species, and the small percentage of the CpFe(CH30H)+ product observed is due to stabilization by radiative emission of a photon. Unlike CpFe+, experiments with isotopically labeled methanol showed that the C5H.6Fe+ inserts into both a C-H and a H-O bond of methanol with subsequent loss of a hydrogen molecule. The elimination of hydrogen is probably a driving force and contributes to an increased reaction efficiency. Kan 1 suggested that the reaction of organometallic ions with methanol be used to probe the metal-hydrogen bond in the gas phase. We studied the reactions of two organometallic ions with methanol and higher alcohols to see if, and how these reactions might differ from each other and to investigate the potential usefulness of higher alcohols in the probing of metal-hydrogen character of organometallic ions. As to the observations related to our first goal, we found out that although the reactions of CpFe + and C5H6Fe+ are similar in the sense that 1. the reactivity of organometallic ion increases with the increasing length of the alkyl chain 2. for both ions the preferred mode of interaction is activation of a C-0 bond of alcohol, with the exception of methanol there are some notable differences. These include: 1. In the C5H6Fe+ system a loss of hydrogen takes place in addition to the loss of alkene or water. This accounts for the formation of the CpFe(OH) + and CpFe(C n H2 n - i ) + ions, respectively that are products of a C-O bond activation . Loss of alkene or water accompanied by a loss of hydrogen can be explained only if the C s ^ F e - 1 - ion has a metal-hydrogen bond. This is in agreement with the previous conclusions of Kan 1 . - 132-The experiments with labeled alcohols confirmed that one of the hydrogen atoms in the lost neutral originates on the C5H6Fe+ ion. 2. Unlike the CpFe + /ROH system, an extensive FI/D scrambling was observed for the reactions of the C5H6Fe+ ion with alcohols. 3. Methanol gives only an ion-neutral product with CpFe+, whereas C5H6Fe+ was found to activate both a C-H and O-H bonds. Potentially, all studied alcohols could be used to probe the metal-hydrogen bond in the gas phase, since in all C5H6Fe+/ROH systems studied, loss of a hydrogen molecule, in addition to the loss of water or alkene that are characteristic in CpFe + /ROH system, occured. From a practical point of view, however, methanol appears to be the most suitable reagent. Its reaction with an organometallic complex that contains a metal-hydrogen bond would be expected to result in a loss of neutral hydrogen molecule only. Isotopic labelling could relatively easily determine the origin of the eliminated molecule. With higher alcohols several bond activations can be involved and the determination of the source of hydrogen and other eliminated neutrals may not always be unambiguous. Further, with higher alcohols, extensive H/D scrambling was observed. This lowers the intensity of the peak corresponding to the addition-dehydrogenation product ion, and produces additional ions that complicate the interpretation of spectrum but have no informational content with regard to the presence/absence of a M - H bond. The results of other studies, as reviewed in Chapter 1.1, indicate that methanol is fairly unreactive with many organometallic ions. It is conceivable that it might be inert toward some organometallic ions, even though the metal-hydrogen bond is present. It is in such cases that the reactions with higher alcohols could be employed. - 133-References (1) Kan, Z . , University of British Columbia, Vancouver, 1993. (2) Jacobson, D. B.; Freiser, B. S. Journal of American Chemical Society 1983,105, 7492-7500. (3) Allison, J.; Ridge, D. P. Journal of American Chemical Society 1976, 98, 7445-7446. (4) Allison, J.; Ridge, D. P. Journal of the American Chemical Society 1979,101, 4998-5008. (5) Huang, S.; Holman, R. W.; Gross, M . L. Organometallics 1986, 5, 1857-1863. (6) O.Blum; Schwarz, H.; Stockigt, D.; Schroder, D. Angew. Chem. Int. Ed. Engl. 1992,37, 603-604. (7) Freiser, B. S.; Carlin, T. J.; Sallans, L . ; Cassady, C. J.; Jacobson, D. B. Journal of American Chemical Society 1983,105, 6320-6321. (8) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry; Interscience Publishers: New York, N.Y., 1982, p.789. (9) Bryndza, H. E. ; Tarn, W. Chemistry Review 1988, 88, 1163-1186. (10) Sonnenfroh, D. M . ; Farrar, J. M . Journal of American Chemical Society 1986, 108, 3521-3522. (11) Company, C. R. Handbook of Chemistry and Physics, 74 ed.; C R C Press: Clevelend, Ohio, 1993-1994. (12) Gioumousis, G.; Stevenson, D. P. The Journal of Chemical Physics 1958, 29, 294-299. (13) Su, T.; Bowers, M . T. The Journal Of chemical Physics 1973,58, 3027-3037. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059552/manifest

Comment

Related Items