.xouC4c 352% WIWIIIHIIWWIHWNIH”!!!(IIHHIIHIUHIIIHI Date 0-7639 This is to certify that the thesis entitled ION CYCLOTRON RESONANCE STUDIES OF THE GAS PHASE REACTIONS OCCURRING IN MIXTURES 0F TRI'CARBONYLNITROSYLCOBALT(0) AND AMINES presented by Barbara Diane Radecki has been accepted towards fulfillment of the requirements for m‘ 5' degreein CLeM'a-r‘l 0 Major professor 5// 1/5; I MSU LIBRARIES m RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. ION CYCLOTRON RESONANCE STUDIES OF THE GAS PHASE REACTIONS OCCURRING IN MIXTURES OF TRICARBONYLNITROSYLCOBALT(O) AND AMINES by Barbara Diane Radecki A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1982 ABSTRACT ION CYCLOTRON RESONANCE STUDIES OF THE GAS PHASE REACTIONS OCCURRING IN MIXTURES OF TRICARBONYLNITROSYLCOBALT (0) AND AMINES by Barbara Diane Radecki The gas phase reactions of the ions formed by electron impact on tricarbonylnitrosylcobalt(0) with methyl, ethyl, n-propyl, isopropyl, n-butyl, t-butyl, sec-butyl and isobutylamine are studied using ion cyclotron resonance spectrometry. The precursors of the ion products formed in these reactions, as determined by ion cyclotron double resonance, are reported. Reaction mechanisms are proposed. The significance of this study is discussed in terms of the basic chemis- try of the isolated Co+ metal center and cobalt containing fragment ions and in terms of the utility of Co(C0)3NO as a chemical ionization reagent. This thesis is dedicated to Pat and Alicia and my Mother and Father without whose love and support I would have not made it this far. 11' ACKNOWLEDGMENTS I wish to express my sincere gratitude to my research preceptor, Dr. John Allison, for his invaluable direction and assistance in my research and the writing of this thesis. iii TABLE OF CONTENTS Chapter Page LIST OF TABLES ............................................ V LIST OF FIGURES ........................................... vi I. INTRODUCTION ......................................... I II. BACKGROUND ........................................... 2 A. Principles of Operation of ICR .................. 2 B. Overview of Gas Phase Transition Metal Chemistry. 5 III. EXPERIMENTAL AND RESULTS ............................. 15 IV. DISCUSSION ........................................... 27 LIST OF REFERENCES ......................................... 54 iv TABLE Noam-noon) (I) ICDR Results ICDR Results ICDR Results ICDR Results ICDR Results IDCR Results ICDR Results of l: of l: of l: of l: of l: of l of l :2 :1 LIST OF TABLES Mixture Mixture Mixture Mixture Mixture Mixture Mixture Co(CO)3NO to Co(C0)3N0 to Co(CO)3NO to Co(C0)3N0 to Co(C0)3N0 to Co(CO)3N0 to of Co(CO)3NO Methylamine ....... Ethylamine ........ n-Propylamine ..... Isopropylamine.... n-Butylamine ...... Isobutylamine ..... to sec-Butylamine ............................................. ICDR Results of l: l Mixture of Co(C0)3N0 to t-Butylamine... Summary of Reactions ....................................... Page l7 T8 19 20 27 23 25 26 28 LIST OF FIGURES Figure Page Figure 1 Block Diagram of ICR Spectrometer for Single Resonance Experiment.... .......................... 2 Figure 2 The ICR Cell ...................................... 3 Figure 3 Amine Bond Strengths .............................. 43 vi I. Introduction Ion cyclotron resonance (ICR) spectrometry was developed as a mass spectrometric technique for the study of bimolecular reactions in the gas phase at low pressures. The study of chemical reactions under "mass spectrometric conditions" allows one to examine reactive systems in the absence of solvent thereby eliminating complications from the interference of solvent molecules which occur in solution. This technique was utilized to study the reactions between the electron impact fragments of Co(C0)3NO and methyl, ethyl, n-propyl, isopropyl, n-butyl, sec-butyl, isobutyl and t-butylamine. The electron impact fragments of Co(CO)3NO include 00*, CoC0+, Co(c0);, CoCON0+, Co(C0)2- NO+ and Co(CO)3N0+. Before presenting the results, the basis for their interpretation must be established in an understanding of the basic principles of operation of an ICR spectrometer. This understanding, coupled with a knowledge of the pertinent research already carried out in the areas of both gas phase and condensed phase organometallic che- mistry, allows one to extract and interpret the structural, mechanis- tic and thermodynamic information contained in the results. This study not only allows one to examine the basic chemistry of the isolated Co+ metal center with the amine series and determine how this chemistry changes as available orbitals on the metal become occupied, but also allows one to examine Co(C0)3N0 as a chemical ionization reagent in the development of metal ion chemical ionization mass spectrometry (MICIMS). II. Background A. Principles of operation of ICR A block diagram of an ICR spectrometer is shown in Figure l. The mass spectrometric functions (ion formation, manipulation, detection) occur in the ICR "cell". The ICR cell is contained within a stainless steel vacuum chamber and is located between the poles of an electro- magnet° The cell consists of three sections which include an ion source, analyzer and collector. It is shown in Figure 2. The instru- ment is basically a mass spectrometer. Thorough reviews of the theory and principles involved in ICR spectrometry have been presented."6 Basically, ions are formed in the source by electron impact (70eV). The presence of the magnetic field, B, causes the ions to travel in a circular orbit in the XY plane at a DRIFT & TRAPPING VOLTAGES I: IAMPLIFIER MODULATION MAGNET REFERENCE 33:23:01: POWER OSCILLATOR . EMISSION f SUPPLY f ‘ * * CURRENT ' - J - PHASE- L CONTROL I SWEEP . SENSITIVEH XY RECORDER CONTROL DETECTOR J L f Fig. l. Block Diagram of ICR Spectrometer for Single Resonance Experiment. I Source 'Analyzer ' Collector] If from . I I Marginal Oscillator l I l | I Drift Plate (Analyzer) l I Ion Current Drift Plate (Source) Collector ~. ‘ “//\ ” L -:r Trapping Voltage Fig. 2. The ICR Cen6 characteristic frequency. This frequency is known as the cyclotron frequency, on? and it is dependent upon the mass to charge ratio, m/q, of the ion. This relationship is expressed in equation (l) B mpg; (I) (where c is the speed of light). A trapping potential, Vtrap’ applied to plates 2 and 4 restricts the motion of the ions in the z direction by creating a potential minimum near the center of the cell. Thus we have iOns trapped by crossed electric, E, and magnetic, B, fields. To move the ions out of the source, it is necessary to apply a voltage difference across the top and bottom plates 1,5,3 and 6 thus producing a static electric field. With an electric field perpendicu- lar to the magnetic field (an E'x B field), the ions drift into the analyzer in a circular manner. The top and bottom plates are also connected to a marginal oscil- lator detector. As the magnetic field is scanned, the cyclotron frequency of the ion constantly changes (IDC «:8). When the cyclotron frequency of the ion equals the frequency of the oscillating electric field of the marginal oscillator, E0, the ion absorbs power and is detected° Typically, the frequency of the marginal oscillator is set at 153 KHz because this frequency corresponds to lOO Gauss/amU. The collector consists of plates 7,8,9 and 10. These plates are at ground potential and are connected to an electrometer. Ions drift into the collector and the total ion current is then measured by an electrometer. The data obtained is a plot of absorption versus magnetic field. This plot is essentially a mass spectrum and is linear in mass. Ion—molecule reactions occur as a result of ion-neutral collisions 6 torr. These in the cell when the pressure is increased above 5 x 10' bimolecular reactions occur within an order of magnitude of the colli- sion frequency. This is an indication that these reactions must have a very small or no activation energy. Thus they must be exothermic or thermoneutral reactions. This assumption allows one to determine limits on heats of formation of products and limits on bondstrengths. Product ions show up as "new" peaks in the mass spectrum as the pressure is raised from very low "collision free" pressures. To identify reaction sequences involving these product ions, ion cyclotron double resonance (ICDR) is utilized. ICDR allows one to unambiguously assign the precursors of a product ion under study. The cyclotron frequency of a product ion is set equal to the frequency of the marginal oscillator so that the intensity of this peak is monitored continuously. A second oscillator is connected to the source and its frequency is scanned such that the frequency of this second oscillator always corresponds to the cyclotron frequency of some other ion. In this manner, ions of a given mass absorb energy, their translational energy increases, and they are quickly ejected from the cell. When the ejected ion corresponds to a precursor, a decrease in the intensity of the product ion occurs. Because the magnetic field is now constant, the relationship expressed in equation (2) now holds. B (2) = = m wProductfibroduct(: wPrecursor PrecursorC q q Thus, the mass of the precursor is readily determined by equation (3) when the frequency ratio is known. wPrOduct = rr'lPrecursor (3) “Precursor mProduct 1,2 Detailed descriptions of the ICDR experiment and thorough discussions concerning the interpretation of double resonance6’7 have been presented in the literature. B. Overview of gas phase transition metal studies. The earliest mass spectrometric investigations of transition metal complexes focused on volatile species such as M(CO)x [where M=Ni, x=4; M=Fe, x=5; and M=Cr, Mo, W, x=6]. Fragmentation patterns containing peaks corresponding to the masses of the positive fragment ions result- ing from 70eV electron impact on a metal complex were obtained. Values for the ionization potential Of M(CO)x and the appearance potentials of the remaining fragment ions were determined. This information was used to determine the energetics involved in the ionization and disso- ciation of the fragment ions.9’10 Transition metal carbonyls were of interest due to the discovery that the metal carbonyl entity played an important structural role in many organometallic complexes. Gas phase studies allowed one to examine the instrinsic properties of the metal complex in the absence of complications from the interference of solvent. Bimetallic ions were first Observed in the mass spectra of mono- nuclear organometallic complexes by Schumacher and Taubenest.1] They identified these bimetallic ions as the products of ion-molecule reac- tions, e.g. I NiCp+ + Nisz —————> NiZCp3+ (CD = C5H5l (4) 16 He Later Mfiller investigated two types of ion-molecule reactions. reported the formation of polynuclear metal complexes from ion-molecule reactions for CpV(CO)4, Cr(C0)6, Ni(PF3)3, Cr(C6H6)2, CpCrCONO and (C6H6)Cr(CO)3° For example, under suitable operating conditions he Observed + + He also reported ion-molecule reactions occurring between the ionized . + + transition metal complexes CpMn(CO)3+, CpNTNO , CpCo(CO)2 and CpV(CO)4+ and a number of o and n donor molecules including H20, R20, NH3, NHRZ, H28, acetylene and benzene. The general equation for reac- tions observed is LM++L'————-)L L'M++ L (6) n n-x X where L=CO or NO. Emerging from this investigation was the realization that ion-molecule reactions of organometallic complexes in the gas phase could provide much infOrmation on the formation and stability of metal-metal and metal-ligand bonds. As MUller was conducting his studies, the technique of ICR was being developed. Soon the capability of ICR for the study of gas phase ion-molecule reactions was realized.12 Clearly ICR had the potential to serve as a useful tool in the characterization of the thermodynamic and intrinsic chemical properties Of ions and molecules in the absence of solvent. One could readily examine the strength and nature of metal- ligand and metal-metal bonds using this technique. Foster and Beauchamp were the first to utilize ICR in the study of the reactions of ions formed on electron impact Of Fe(CO)5 alone and in a binary mix- 13 They reported ture with molecules such as CH3F, H20, NH3 and HCl. the formation of binuclear iron species from ion molecule reactions such as Fe+ + I=e(CO)5 —-———> Fe2(CO)4+ + CO (7) The precursor ions of the binuclear iron complex were unambiguously determined by ICDR. From the binary mixtures studied, bond energies and relative rates of ligand substitution in Fe(CO)n+ (n=0-5) were estimated. Since the initial ICR studies dealing with the formation of poly- nuclear clusters and ligand substitution reactions in transition metal complex reaction systems, it has been discovered that, in addition, many of these transition metal complexes exhibit rich chemistries with organic molecules in the gas phase.18'20’ 22'48 Work on gas phase bimolecular organometallic chemistry is now being conducted by the research groups of Ridge (Delaware), Beauchamp (Cal. Tech.), Freiser (Purdue), Allison (MSU) and until recently, Staley (MIT). Ridge and co-workers have examined a number of transition metal species in conjunction with organic molecules.23’28'32’34 They were the first to report the formation of metal to carbon bonds in the reaction of the electron impact fragments of CO(C0)3N0 and Fe(CO)5 with alkyl halides.18 CO(C0)3N0 and other metal carbonyls were chosen to study because they are volatile complexes and CO and NO are important ligands in transition metal chemistry. To explain the formation of metal to carbon bonds, they reported the oxidative addition of M+ formed from electron impact on MLn (M=CO, Fe, Ni, Cr) to C-X bonds (X=halide, OH) in saturated polar organic neutrals which included small alkyl halides and alcohols.23’32 Fe+ + CH + + I (8) I -—————9 [CH3-Fe-I]i ea: FeCH 3 3 + , i FeI + CH3 (9) Also included in this mechanism was the shift of a hydrogen atom from a carbon 8 to the metal, onto the metal followed by the elimination of a small stable neutral such as HX (X=halide, OH). D + + DI (l0) + Fe + C20 I -—-—-—9 FeC2 4 5 Thus Fe+, Co+ and Ni+ were found to dehydrate alcohols and dehydrohalo- genate alkyl halides according to the following general mechanism a: M+ + Mao—NE Brie—9 x—M ...:H )-——9M+...\“’ + HX(11) /\ /\ where X= OH, F, Cl, I. Ridge and co-workers have also studied mixtures of TiCl4 with 27 30 Studies of small olefins and oxygen containing organic compounds. mixtures of TiCl4 with organic molecules in the gas phase are of interest due to TiCl4's catalytic participation with aluminum alkyls in the Ziegler-Natta scheme for the polymerization of olefins in solu- tion. Gas phase studies of TiCl4 may be reflective of condensed phase titanium organometallic chemistry. The electron impact fragments of TiCl TiCT4 (TiCl 3 , TiCT2 , TiCl+, Ti+) exhibited specific Chemis- 4+ 9 try with organic compounds. Ti+ and TiCl+ eliminated H2 in four chain olefins, e.g. + + v TiCl + ——+ cum; +H2 (12) . + TTCl2 bonyls e.g. and TiCl3+ eliminated HCl in olefins and in small organic car- 4- TiCl3+ +3] ——->C12Ti---3 +HCL (13) + O TiCT3 +JL ___—. TTCl2C3H50 +HCl (l4) TiCl+, TiClz+, Tic13+ eliminated smaller olefins in olefins and in oxygen containing organic compounds as is illustrated in reactions 15 and l6: lO Cl -C4H8 Cl I4- TiCl3+ +\/\/\ -——> Cl- Tiv —————9 Cl-Ti ———a Cl Cl--H Que) CT Cl-Ti+----" (15) Cl TiCl2 +,JH\,,,\ -—————9'TiClz(C2H 40) + C 3H6 (l6) TiCl4+ was found to be inactive in all cases. In species containing carbon-oxygen single bonds, cleavage of the C-0 bond was observed, e.g. Cl I T + ROR' ——-> Cl-Ti+-Cl -————-> TiClZ(OR)+ + R'Cl (17) [31) L R. TTCl3 The mechanism consistent with these results involved the oxidative addition of the metal ion fragments of TiCl4 to the double bond (C=C, C=O) or to the carbon-oxygen single bond fOllowed by the reductive elimination of small neutral molecules or bond cleavage in the case Of the single C-O bond. Reactions 15 and l7 above illustrated this mechanism. Furthermore, the position of the double bond and the struc- ture of the carbon skeleton were significant in determining the reaction sequence. This is evident in reaction 18 where the driving force is the formation of the methylcyclopentadiene ring which can strongly com- ll plex with Ti+z + 1.1+ + \[\ ———§[Ti+'>(\]i —_.....)T1’: + 2H2 (l8) Results obtained from studies done on alkanes by Ridge and co- workers34 indicated that Fe+ formed from electron impact on Fe(CO)5 could add oxidatively to carbon-carbon and carbon-hydrogen bonds. The . . + . following reaction sequence was observed for Fe and isopropane: / H H Fe+ +>— -——+ CHB-Fe+——€ ——+ Fe+o - . [L—a Fe+--' It CH4 H3 (l9) n\ H ’l/a H \ .___, H--Fe+-—->< —-) /Fe+---)K ———> Fe+--jL + “2 H (20) ICDR results, calculated heats of formation of reactions and studies employing labeled organic compounds for each mixture Ridge examined are all consistent with the structures and the mechanism proposed. Beauchamp and co-workers have also contributed significantly to the area of gas phase transition metal chemistry. They have examined 20 22 the basicity of iron pentacarbonyl,19 ferrocene and nickelocene in the gas phase using ICR and have reported the proton affinities of a number of organotransition metal complexes.44 They have determined the metal-ligand bond dissociation energies of thirty n-donor ligands to 21 cyclopentadienylnickel cation and have described the methodology in- volved in the determination of metal-hydrogen, metal-carbon and metal— 35 ligand bond dissociation energies in the gas phase using ICR. These studies have been undertaken in a effort to gain information relating l2 to the intrinsic reactivity of transition metal complexes in the ab- sence of solvent. Furthermore, Beauchamp has reported the decarbonylation of aldehydes by (nS-C5H5)Ni+ in the gas phase24 e.g. CpNi+ + RCHO ————> CpNiCO+ + RH (21) This reaction was found to occur if RCO+ was relatively stable. The mechanism proposed by Beauchamp involved a reaction intermediate such as [CpNi \RH occur. The decarbonylation of aldehydes has been Observed in solution 49 ] from which the competitive elimination of RH or CO could and on solid surfaces. Transition metal carbenes have also been studied by Beauchamp41’42. Interest in metal carbenes arises from the fact that they may be inter- mediates in the Fischer-Tropsch synthesis and olefin metathesis reac- tions. Beauchamp and co-workers examined cobalt and chromium carbenes in an effort to characterize these metal carbenes in the gas phase. They determined the (Co+-CH2) and (Cr+-CH2) bond energies. Beauchamp has also employed ion beam techniques to produce energy 38 selected cobalt metal ions. He then examined the reactions of CO+ with alkanes containing four or less carbons. His results indicated the occurrence of an oxidative addition of Co+ to carbon-hydrogen and carbon-carbon bonds. These results were consistent with the mechanism proposed by Ridge for the reactions of Fe+ with alkanes. Recently, Beauchamp and co-workers have expanded on these CO+ ion beam studies. 43 45 They examined the reactions of seventeen alkanes, twelve alkenes 46 and the four smallest cycloalkanes with CO+ produced from ion beam 13 techniques. With alkanes they reported once again the oxidative addi- tion of Co+ to a C-C or C-H bond followed by 8 hydrogen abstraction and the reductive elimination of a small neutral molecule (accompanied by the formation of a cobalt alkene ion). Initial oxidative addition was fOund to occur at the weakest bond. The mechanism describing the reaction of alkenes with Co+ involved the oxidative addition of the metal to an allylic C-C or C-H bond and was consistent with the mech- anism proposed for alkanes. The double bond was found to be capable of directing the Co+ addition to the allylic bonds e.g. H\ Co+ + i \ ——-—> Co+...)k/ z H-CO+E —-) :CO. ———-) H H + Co(C4H6) + H2 (23) For cycloalkanes, Beauchamp and co-workers have Observed products from dehydrogenation and ring cleavage implying Co+ insertions into C-C and C-H bonds followed by ring cleavage. This is illustrated in reactions 24 and 25: CH + 2 Co“ + A —— C03 2 if —-» COCH2+ + C2H4 (24) ”\ + Co+ + D ———a H-Co+‘<> ———» /Co+--[[:] ——-—> CO(C4H6) + H2 H (25) . + Staley and co-workers have examined the gas phase chemistry of Cu produced by the volatilization/ionization of copper metal by a pulsed 39 laser with alkyl chlorides. They observed halide transfer onto Cu+and l4 dehydrochlorination. They use the following mechamism to account for their results: I-Cu+ H —.¢ Cl H i _ CJ + = + C“ + H -—> c3. (H --> \ C.“ , (26) y / L‘ CuCl + H HCl Cu+ +>:::-\/ >_-—".\ + HCl air Staley gtggl, also examined the gas phase reactions Of Ti+ produced by laser desorption/ionization with halomethanes, alkylchlorides, chloroethylenes and chlorobenzene.40 They reported that the most pre- valent reactions of Ti+ with halomethanes and alkylchlorides involved chloride transfer and oxidative halogen transfer to yield TiX+. Also, Ti+ dehydrochlorinated vinyl chloride and chlorobenzene. In a recent study Staley gt_gl, determined an order of the rela- +.47 TiCl + was tive binding energies of neutral molecules to TiCl3 3 produced from electron impact on TiCl4. Freiser and co-workers have examined the gas phase reactions of 48 They observed the cleavage of C-C, C-H Fe+ with ketones and ethers. and C-0 bonds to produce alkyl, acyl and alkoxide Fe+ species in oxida- tive addition reactions. Decarbonylation was observed with a few small ketones. Dehydrogenation was reported for unbranched ketones and re- ductive elimination of methane occurred in ketones which were branched at the a carbon. Fe+ was produced by a pulsed laser. From the above overview it is evident that many transition metal 15 ions and transition metal complex ions exhibit rich chemistries with the various types of organic molecules in the gas phase. III. Experimental and Results All data were Obtained on a "drift-type" ICR Spectrometer of conven- tional design8 which was constructed at Michigan State University. A standard marginal oscillator detection system50 was employed. A Wave- tek Model l44 sweep generator was used as the secondary oscillator in double resonance experiments. Tricarbonylnitrosylcobalt(0) was obtained from Alfa Inorganics. n-Butyl, isobutyl, sec-butyl and t-butylamine were purchased from Chem-Service. Isopropyl and ethylamine were purchased from Kodak, n- propylamine was Obtained from Aldrich and methylamine was obtained from Matheson. All samples were degassed three times by freeze- pump-thaw cycles. After degassing, the vapor of a liquid sample (typically around 7 torr) was admitted into an evacuated glass bulb and used without further purification. Sample bulbs were connected to the instrument via a two-sample inlet system. The pressure of a sample gas entering the cell was controlled by a Varian 95l-5106 leak valve and was measured by a Veeco RG-lOOO ionization gauge. Data were acquired in the following manner. First a spectrum of CO(C0)3N0 (molecular weight 173) was obtained up to a mass of 250 at 6 5 x lO' torr. All ions formed by electron impact of CO(C0)3N0 at 70eV were determined. The pressure was then increased to l x lO'5 torr and any ions formed by ion-molecule reactions in Co(CO)3NO alone were determined. Then the leak valve to the cobalt compound was Closed and the system was evacuated to l x lO'7 torr using a 4" diffusion pump and l6 an Ultek lSO ma ion pump. Following this, an amine sample was intro- duced into the cell. The molecular weight of the amines varied from 3l for methylamine to 73 for the butylamines. The ion fragments produced by electron impact on the amine at low and high amine pressures were determined in the same manner as for the cobalt compound. Next a binary mixture containing C0(C0)3N0 and an amine was intro- duced into the cell. These mixtures were either l:2 or l:l (by pressure) Co(CO)3NO to amine. In each case the total pressure was l x lO'5 torr and spectra up to a mass of greater than 250 (173 + amine :_250) were obtained. In interpreting the spectrum of each mixture, first those peaks characteristic of the cobalt compound alone and those characteristic Of the amine by itself were determined. The remaining peaks were assumed to correspond to product ions formed as a result of ion-molecule reactions. Double resonance was performed on all product ions so deter- mined. Precursors were identified in each case. These double resonance results are tabulated for each mixture in Tables l thru 8. (Note that if A+-——)rB+ and B+-——a C+, then the precursors of C+ as determined by ICDR will be A+ and 3+). l7 Table I. ICDR Results of l:l Mixture Co(C0)3NO to Methylamine Product Stoichiometry Precursors m/z 88 Co(CH3N)+ Co+ 90 Co(CH5N)+ Co+, COCO+, Co(CO)2+ ll8 CoCO(CH5N)+ COCO+, Co(CH5N)+, Co(CO)2+ + + + 120 CoNO(CH5N) COCONO , CO(CO)2NO 121 Co(CH5N)2+ COCO+, Co(CH5N)+, CO(CO)2+, CoCO(CH5N)+ + + + l48 CoCONO(CH5N) CHSN , Co(CO)3NO + + + 151 CoNO(CH5N)2 CoCONO(CH5N) , CO(CO)3NO 179 COZNO(CH5N)+ Co+, CoNO+, CO(CO)3NO+, C02(CO)2NO+ 18 Table 2. ICDR Results of l:2 Mixture C0(C0)3N0 to Ethylamine Product Stoichiometry Precursors m/z 88 Co(CH3N)+ Co+ 9O Co(CH5N)+ Coco+ , 102 Co(C2H5N)+ Co+, CoCO+ + ‘04 C0(C2H7N) Co+, CoC0+, Co(CO)2+ ll6 CoCO(CH3N)+ CoCO+ CoCO(C2H5)+ + + ll8 CONO(C2H5) C2H5 + + CoNO(CH3N) C02(CO)2NO 120 CoNO(CH5N)+ COCONO+, Co(CO)2NO+ + + + + + l2l Co(CH5N)2 Co , COCO , C0(C2H7N) , Co(CO)2 + + + + + 130 COCO(C2H5N) Co , COCO , CO(C2H5N) , COZCO + + + + 132 COCO(C2H7N) COCO , CO(C2H7N) , Co(CO)2 134 CoNO(c2H7N)+ COCON0+, Co(CO)2NO+, Co(CO)3NO+ + + + + + 149 Co(C2H7N)2 COCO , CO(C2H7N) , Co(CO)2 , CoCO(C2H7N) 152 COCONO(C2H7N)+ CO(CO)2NO+, Co(co)3NO+ 179 CoNO(C2H7N)2+ Co+, COCO+, CoNO(C2H7N)+, Co(CO)2NO+, COCONO(C2H7N)+, Co(CO)3NO+ + + + + + 193 C02N0(C2H7N) CO , COCO , COZCONO , COZ(CO)2NO 19 Table 3. ICDR Results of l:l Mixture Co(CO)3NO to n-Propylamine Product Stoichiometry Precursors m/z 90 Co(CHsN)+ Co+, Coco+ 114 Co(C3H5N)+ Co+, Coco+ ll6 Co(C3H7N)+ Co+, Coco+ 118 Co(C3H9N)+' Co+, COCO+, Co(CO)2+ 146 CoCO(C3H9N)+ COCO+, Co(CO)2+ + + + + 148 CoNO(C3H9N) COCONO , Co(CO)2NO , Co(CO)3NO + + + + + 149 CoNO(C3H1ON) Co , COCO , COZCO , Co(CO)3NO Co(C H N )+ 4 14 2 + + + l77 Co(C3H9N)2 COCO , Co(CO)2 207 C02N0(C3H9N)+ Co+, CoCO+, Co(CO)2NO+, Co(CO)3NO+ + CoNO(C3H9N)2 20 Table 4. ICDR Results of 1:2 Mixture Co(CO)3NO to Isopropylamine Product Stoichiometry Precursors m/z 102 Co(CszN)+ C0+, Coco+ 116 Co(C3H7N)+ Co+, COCO+, Co(C0)2+ ll8 Co(C3H9N)+ Co+, COCO+, Co(CO)2+ 146 CoCO(C3H9N)+ COCO+, Co(CO)2+ l48 CoN0(C3H9N)+ COCONO+, Co(CO)2NO+, Co(C0)3N0+ 158 Co(C6H13N)+ CoCO+ 160 Co(C6H15N)+ CoCO+, CO(CO)2+ Co(CO)2(C2H5NH2)+ 162 CoCON0(C2H5NH2)+ COCONO+, Co(CO)2NO+, Co(C0)3N0+ 207 C02N0(03H9N)+ 00+, CoCO+, Co(CO)3N0+ + CONO(C3H9N)2 2T Table 5. ICDR Results of lzl Mixture of Co(CO)3NO to n-Butylamine Product Stoichiometry Precursors m/z 88 Co(CH3N)T CoT 90 Co(CHsN)+ CoT, CoCOT + + 102 Co(CzHSN) Co 104 Co(C2H7N)+ CoT, CoCOT + + ll3 Co(C3H4N) Co + + 114 Co(C3H5N) CH4N ll6 Co(C3H7N)T CoT, CocoT 128 Co(C4H7N)T CoT, CoCOT 130 Co(C4H9N)+ CoT, CoCOT + + + + 132 Co(C4H11N) Co , COCO , Co(CO)2 + COCONO(C3H5) l58 CoCO(C H N)T Coco+ Co(C H N)+ Co(CO) NOT 4 9 ’ 4 9 ’ 2 160 CoCO(C H N)T CoCOT, Co(CO) T 4 11 2 l62 CONO(C4H]]N)+ CoCONOT, Co(C0)2NOT, Co(CO)3N0T + + + 163 CONO(C4H]2N) Co(CO)2NO , Co(CO)3NO + + 190 COCONO(C4H]]N) Co(CO)3NO + + + + + 205 Co(C4HHN)2 Co , COCO , Co(CO)2 , Co(C4H11N) , + CoCO(C4H11N) Table 5, con't. 22 Product Stoichiometry . Precursors m/z + + + + 221 C02N0(C4H]]N) CO , COCO , C02(C0)2N0 235 C N0(C H N) T C (CO) NOT C N0(C H N)T C (CO) NOT o4112 02’0411’03 23 Table 6. ICDR Results Of l:2 Mixture Co(CO)3NO to Isobutylamine Product Stoichiometry Precursors m/z 90 Co(CHsN)T CoT, CoCOT + + ll4 Co(C3H5N) Co 116 Co(C3H7N)T CoT, CoCOT + + 118 CoCO(CH5N) Co(CHSN) + + l28 Co(C4H7N) Co 130 Co(C4H9N)+ CoT, CoCO+ + + + + 132 Co(C4H11N) Co , COCO , Co(CO)2 144 CoCO(C3H7N)T CoT, CoCOT + + + + 158 CoCO(C4H9N) COCO , Co(C4H9N) , Co(C0)2NO + COCONO(C3H5) + + + + 160 CoCO(C4H11N) COCO , Co(CO)2 , Co(C4H11N) + + + + l62 CONO(C4H]]N) COCONO , CO(CO)2NO , Co(CO)3NO l63 CoNO(C4H]2N)+ CoCOT, CoNOT, CoCONOT, Co(CO)2N0T, Co(C0)3N0T + + 190 COCONO(C4H1]N) Co(CO)3NO 24 Table 6, can't. Product Stoichiometry Precursors m/z + + + + 205 Co(C4HnN)2 Co , COCO , Co(CO)2 + + + + 22l C02N0(C4H11N) Co , COCO , C02(CO)2NO + + + + 235 C0N0(C4HHN)2 Co(CO)2NO , CoNO(C N) , Co(CO)3NO 4”11 25 Table 7. ICDR Results of 1:1 Mixture of Co(CO)3NO to sec-Butylamine Product Stoichiometry Precursors m/z 90 Co(CHSN)+ CO+ 102 Co(CzHSN)+ CoT, CoCOT 104 Co(C2H7N)T CoT, CoCOT + + ll3 Co(C3H4N) CO + + l28 Co(C4H7N) CO 130 Co(C4H9N)T CoT, CoCOT 132 Co(C H N)+ CoT, CoCOT, Co(CO) T 4 11 2 158 C CO(C H N)T C (CO) NOT ° 4 9 ° 3 + CoCONO(C3H5) + + + 160 CoCO(C4H11N) COCO , Co(CO)2 162 CoN0(C4HHN)+ CoCONOT, Co(CO)2NO+, Co(CO)3N0T 190 CoCONO(C H N)T Co(CO) NOT 4 11 3 + + + + 205 Co(C4HnN)2 Co , COCO , Co(CO)2 + + 22l C02N0(C4H11N) Co 235 CONO(C H N) T Co(CO) NOT CONO(C H N)+ Co(CO) NOT 4 11 2 2 . 4 11 . 3 26 Table 8. ICDR Results of l:l Mixture of Co(CO)3NO t-Butylamine Product Stoichiometry Precursors m/z ll6 Co(C3H7N)T CoT, CoCOT + + + l32 Co(C4H11N) COCO , Co(CO)2 T C H NT 158 CoCONO(C3H5) 3 8 + T T C (C H N)T 160 CoCO(C4H11N) COCO . Co(Co)2 . o 4 11 + + + + l62 CoNO(C4H]1N) CoCONO , Co(CO)2NO , Co(CO)3NO 190 CoCONO(C H N)T Co(CO) NOT 4 11 3 + + + + 205 Co(C4HnN)2 Co , COCO , Co(CO)2 + + + + 221 C02N0(C4H11N) Co , COCO , COZCONO 235 CONO(C4H]]N)2+ CoNOT, CoCONOT 27 IV. Discussion This discussion represents a continuation in the quest for the characterization of the gas phase reactions occurring between the elec- tron impact fragments of transition metal complexes and organic mole- cules using ICR. Previously, the major emphasis has been placed on the study of the reactions of the bare metal ion with organic compounds. In this study, reactions of all the ion fragments produced by electron impact on Co(CO)3NO (CO+, COCO+, Co(CO)2+, COCON0+, Co(CO)2NO+, Co(CO)3NO+) with methyl, ethyl, n-propyl, isopropyl, n-butyl, isobutyl, sec-butyl and t-butylamine have been examined. A summary of the reac- tions is presented in Table 9. To interpret the results, it is neces- sary to formulate a mechanism consistent with this data. First the chemistry of Co+ with amines will be examined. Then we shall determine how this chemistry is altered as the number and nature of the ligands in the reacting metal-containing ion changes. In alkyl halides and alcohols, metal insertion into the polar C-X bond has been observed followed by the elimination of HX23 e.g. CoT + >——Cl —-i>-CoT—CT —> CoT-olL + HCl (27) + + CoT + >—OH —-)>-Co —OH —-> CO --[L + H20 (28) Since amines also contain a polar bond, one would expect to see the elimination of NH3 from amines analogous to reactions (27) and (28): CoT + >—-NH2 ——-—> CoT-nlL + NH3 (29) Instead, reaction (29) was not observed for isopropylamine or for any 28 : IX .o.z .o.z P _ N eu + NzN + +27 eN: N_ hooveo .. o o o u NIN + +zp-eN=euoneveu .. N N ooN + N: + +zp+eNz UN-oneveu .4 _ N _ _ P F .o.z N cu + N: + +zp+eN=eu_-onuveN .4 o F.o o o o _.o on m N: + ;zp+eN Io xfiooveu .4 .o.z _ _ F F < ON + N-eN=F-eo + +zmxu_-xlouvee .4 o o o o < N- eN:_- :0 + +2 mze xaooveo .+ I mzzp+cmzcu + +x38ou -pcmp -omm Tome 1: Tome 1: waxy :owpommm V": M": N": F": ocxm mcoBoemm ..B 385.5 .m eeeee 29 N.A N.A N.A N.A N.A N.A N.A Nux A so + +NI zA+eNIe oA-onuveN .4 A A.o A.o A.o A.o A.o A.o A o-x A +NIz AINIe IN onuAeu .4 o I NINN + 4zm-INI m- -e :ouveu .4 A .o.z . I 00+ IINQ 4 +Im-ININ- -I NA-onuveN .4 o o . I . . eINu.+ 42m- IN IN -e :ooveo -. A . . I . oo 4 IINQ.+ +zA-ININ-eeA-onuveu 4 o A , I . eINu + +zA-ININ-INonuveu .4 o I III 4 NI 4 +zN-INI: A-e xiouveu .4 A A . A A A I ON 4 III 4 +zA-INI A- -I uA-onuveN .4 o . o A.o o o A.o I . IIN 4 42A- INI A- I fiAouveU .4 cowuommm -ugmu -umm -omA -: -omA -: waxy on: mu: Nu: Au: .cxz AA.IeeV I eAIeA 30 A A A4. I II 4 +NANIIA+INII IIII .4 . X I I I-x I + AIIVNANII ;+IN IIII .4 TI NIIAINII I. :IIAIN II 338 I N N N N N N . I . III 4 +NIzA+INIIIIzm-xAIIVeI .4 m.N m.N . m.N I.N N N N.N N.Nux I IIN + +NIzA+INII III N-xAIIIeI .4 A A A A A A N.A A-x I II + +NIIA+INII IIIA-xAIIIeI .4 Ami-.... NIzA+INIII + +szAIIIeI N N N N N N N th A . IIN + +NIzA+INE -xAIIIeI .4 cowuummm -ugmu numm uomA -c -omv -: IIAH cu: nu: ma: Pu: .cxm . AI.IIIV I IAIIA 31 um>gmmno uoz u .o.z A+INIIIIIIII .4 A+INIIIII IIA-xAIIIoI +.+ANII N T 23815 + + II IuuuuuuuuuunnnIumuuuuu4nuuuufuuuuJnuunuuuuuuuuuuunnnuuunnuuunuuunnnunnnnuuunuu IIN + +NIIA+INIIIIINII .4 fl II + +NIIA+INIIIIzmII..4 .muuuu NIzA+INIII + +szAIIINII II + +NANIIA+INIIIIIIII .4 or +NANIIA+INIIIIIIII .4 TI NIzA+INIII + +ANIIA+INIIIIszAIIIII nusww I: e": waxy .cxm cowuummm AI.IIII I IAIIA 32 of the amines under study. Since reactions (27) and (28) have been shown to be the result of a meta] insertion, B-hydrogen atom shift mechanism, our observation indicates that the meta] does not appear to insert into the carbon-nitrogen bond. In fact, the reactions observed for amines (H2 eTimination, CH4 eTimination, olefin eTimination) very cToseTy resembTe those which Beauchamp has observed for saturated hydro- carbons.43 Why are amines atypicaT of poTar saturated organic moTecuTes in this respect? That is, why do they "Took Tike" aTkanes to a metaT center during reaction? The answer to these questions Ties in an examination of the processes which do occur in the reaction of Co+ with amines. As indicated in Table 9, reaction types A, B, C and D were observed + . for Co and four or more of the amines: (30+ + anZnHNH2 > Co(CHsN)+ + Cn_]H2n_2 (A) .___, Co(CnH2n+]N)+ + H2 (B) --——-> Co(CnH2n_1N)+ + 2H2 (C) '——-> Co(Cn_1Hén_1N)+ + CH4 (D) Proposed Mechanism for Reaction Type A Reaction type A, Teading to the formation of Co(CHSN)+, was ob- H2 . served for a/\IINH2, ./\f\NH2, >>"~NH2 and ”\r’N . There are five possibTe structure for this species: 33 H c:--N-CH H N-c +-CH H c + CH NH CH NH c + H . 3 2 ° 3 '°' 2 2 3 '°' H (I) (II) (III) (IV) CH—C+° NH 2— ° 3 Note that the CHSN group of atoms corresponds to that for methylamine. No simple fragmentation of n-propyl, n-butyl, isobutyl or sec-butyl amine could give rise to structure I, which is a coordination complex of Co+ to methylamine, or to IV or V. However, there are two possible mechanisms which can give rise to the formation of either II or III. First, n-propyl, n-butyl and isobutyl each contain an alkyl group with a primary carbon attached to the amine functional group. A mechanism consistent with the previously observed metal insertion into a carbon- carbon bond (in alkanes) followed by a B-hydrogen atom shift onto the metal can explain the formation of Co(CHSN)+ with the configuration of III from these three amines: Co+ + RCHZCHz-NHZ -—> HZC-Co+- CH2--NH2 ——9 RM Hzc-\,§:R H (30) HgCo+-CH2-NH2 -—9 C'o+-CH2NH2 + H2C=ER H 34 In support of the above mechanism is the fact that we don't see the for- mation of Co(CHSN)+ from ethyl, isopropyl or t-butyl amine. If Co+ inserted between the analogous two carbons in ethyl, isopropyl or t- butyl amine we could not get the formation of Co(CH5N)+. Ethylamine has no hydrogen atoms on a carbon which are 8 to the metal and both iso- propyl and t-butyl amine have a methyl group attached to the carbon which is a to the amine functional group. Sec-butylamine also has a methyl group attached to the carbon a to the amine functional group but it was observed to react with Co+ to form Co (CH5N)+. A mechanism involving the formation of a five-membered ring intermediate can explain this observation,e.g. : o I Co++ \2\NH ——> 3 -‘+ —-——+ + Co+ (31) 2 «.50 g H C' 3 CH3 II Work being done by Lombarski and Tsarbopoulos in our laboratory on the chemistry of Co+ with multifunctional molecules strongly indicates the preference of Co+ to react via a S-membered ring intermediate in reac- 5] Note that the mechanism tions which don't involve metal insertion. presented in reaction (31) could also be applicable to propyl, n-butyl and isobutyl amine because they each can form a 5-membered ring inter- mediate. Ethyl, isopropyl and t-butylamine are structurally prohibited from forming such an intermediate. Thus the mechanism proposed in reaction (3l) is consistent with all results. Mixtures of ammonia and Co(C0)3NO have been examined in the gas phase by ICR.28- The only reactions reported involved the displacement 35 of up to two carbonyl ligands by NH3 molecules in Co(C0)xN0+ ( x=l,2,3) and Co(CO): (x=l,2) e.g. Co(C0)N0+ + NH3 ———--9 CoNO(NH3)+ + co This implies that metal insertion into nitrogen-hydrogen bonds does not occur suggesting that configuration IV is unlikely. Therefore, the experimental evidence which has been discussed to this point suggests the formation of Co+ ("methylamine") through either a 5-membered ring intermediate gr_via a Co+ insertion into the R-CHZNH2 bond. However, the ring mechanism should be energetically very similar to Co+ inser- tion into a C-N bond (i.e. formation of Co-NH2 and Co-R bonds) which apparently does not occur. Proposed Mechanism for Reaction Type B Reaction type B involving H2 elimination was observed for CH3NH2, /\NH2, /\/NH2, >—NH2, /\/\NH2, >’\NH2 and kaH2. These results can be accounted for if one considers a mechanism involv- ing metal insertion into a carbon-hydrogen bond followed by a shift of a hydrogen atom B to the metal, onto the metal and then an H2 elimina- tion as proposed for alkanes.34’43 T Co+ + *3 i' /H T'JI\H Co + R-C— c— N\ ——————-> n-c-c-N< ' ' H 4m H H H / 36 I It \4 Co (9° H ’4‘ I} R- c-E‘c—N< R- c- (3)-N I I H I l \H H H H H Co 3 90" R— QéCHNHZ R-CH-é- CH-A-‘=NH H (32a) (32b) Note that for methylamine there is only one type of carbon-hydrogen bond into which the metal can insert suggesting that the only possible ion product formed from methylamine is an imine as is illustrated in reac- tion (32b). For ethylamine, and all of the other amines for which this reaction is observed, there is the possibility for the formation of an Co+ (R==\ NH 4. alkene amine (32a) or an imine (R—w-‘co ) (32b) metal NH 2 complex. To determine which of these products would be more favorable, 37 one could calculate the heat of reaction, AHRxn.’ involved in going from an alkylamine to an alkene amine or toaniimine and H2. The reaction entailing the smallest AHRxn would then be considered to be the most favorable energetically. (This, of course, neglects the relative stabi- lities of the cobalt-complex products). Unfortunately, these thermo- chemical data could not be calculated at present due to the lack of in- formation concerning the heat of formation of the carbon-nitrogen double bond (and for imines in general). Hydrogen molecule elimination was not observed for t-butylamine. t-Butylamine has no available hydrogens on the a carbon in which to insert. Also, there are not two adjacent atoms in the molecular skele- ton which contain H atoms, so any complex elimination of H2 from t-butylamine would have to either involve significant skeletal rearrangement gr_the formation of a diradical product. ’3 H-C-H H l8 H ' / H-C———-—-C N I a \H H I ”-9-” t-butylamine H This observation supports the mechanism proposed in reactions (32a) and (32b). Furthermore, the fact that H2 elimination is observed for ethylamine (a”“~NH2) and isopropylamine ( :>-NH2) [which is analo- gous to an ethylamine with a methyl group substituted for a hydrogen atom on the carbon a to the amine group] but not for t-butylamine [which is analogous to an ethylamine with two methyl groups substitu- 38 ted for the two hydrogens on the carbon a to the amine group], implies a simple l,2 elimination also consistent with the mechanism proposed in (32). The formation of an alkene and H2 cannot be explained by the 5- membered intermediate mechanism suggested in reaction (3l). Therefore, we propose that H2 elimination occurs following the initial mechanis- tic step of Co+ insertion into a C-H bond on the a-carbon atom. Proposed Mechanism for Reaction Type C Reaction type C involves the elimination of two hydrogen molecules. This reaction is observed for n-propyl, n-butyl, isobutyl and sec- butylamine. Upon loss of two hydrogen molecules, these amines apparen- tly form a conjugated system of imine and olefin functionalities. The other amines studied were too small to form such systems. After n- propyl, n-butyl, isobutyl and sec-butylamine lose one H2 via the mechanism of reaction (32b) a second H2 is lost. The driving force of this reaction is the formation of a conjugated system which can complex strongly to the cobalt: ’5‘. i‘ \‘ o :as‘ _ 's\ ‘2‘ ’ll;;H 4 .’ NH ’NH . - NH ° + Co+ Co+ Co+ C0 \ ' All of the proposed product structures are isoelectronic with butadiene and methylbutadiene complexes which are known to be very stable. Taking n-propylamine as typical, we are proposing the following mechanism for .__.§__;. Co+...| -————e> 2 (325) reaction type C: C0+ + \/\ NH 39 /H2 NH Co Proposed Mechanism for Reaction Type D The elimination of methane occurs in reaction type D. This reac- tion was observed for Co+ and ethyl, isopropyl, n-butyl, isobutyl, sec- butyl and t-butyl amine. The mechanism proposed for this reaction involves the insertion of the metal ion into a methyl-carbon bond followed by the shift of a hydrogen atom which is B to the metal onto the metal, thus leaving the components of methane [(CH3)(H)] on the metal. These components can then be eliminated in a manner similar to H2 elimination in reaction type B. This mechanism is also analogous to that observed for alkanes in which metals apparently insert into C-C bonds. Thus, the mechanism which we propose for methane elimination is shown in (33) using sec-butylamine as an example: H-C- CH3 "“2 40 H + + Co -CH ""—€> Co ... + CH HG} g\2 A 4 (33) 3 ~C-CH u 3 HZN CH3 NH 2 For ethylamine, there is only one carbon-carbon bond into which the Co+ can insert. The only 8 hydrogen available is that on the nitrogen and once this hydrogen is shifted to the metal and removed, the products formed are the Co+ (imine) complex and CH4. Therefore, for ethylamine Co+ insertion into the carbon-carbon bond cannot yield Co(CHSN)+ and Co(CHSN)+ was not observed. This finding is consistent with the mech- anism proposed in reaction (30). Thus for CZHSNHZ’ (30) predicts CH4 elimination in lieu of Co+(CH5N) formation. t-Butylamine contains three terminal methyl groups and isopropylamine, isobutylamine and sec- butylamine each contain two. Methane elimination was observed for each of these species. Methylamine contains no carbon-carbon bonds. There- fore, mechanism (33) is not operative for this compound and such pro- ducts were not of service. The data for n-propylamine indicates that CH4 elimination does not occur. This seems to be inconsistent because n-propylamine does con- tain a terminal carbon-carbon bond. To explain this it is necessary to consider the energetics involved in the proposed mechanisms. These considerations will also allow one to answer the initial question posed concerning the atypical behavior of amines compared to other polar compounds. 41 Consider how a reaction occurs in the gas phase under low pressure conditions where only bimolecular collisions occur. For metal insertion into an already existing bond to be feasible, the energy required to break the bond must be compensated for by the energy released upon the formation of the two new bonds to cobalt. In general, the energy needed to break a bond in a reaction must always be compensated for by the energy released upon forming new bonds. Both the Ridge and Beauchamp groups concurred that initial oxidative addition of a metal into the C-C or C-H bond of an alkane was occurring preferentially at the weakest bond.34’43 This makes sense energetically because breakage at the weakest bond requires the least amount of energy. In alkyl halides and alcohols, the polar bonds (R-X) which are attacked are strong bonds e.g. D((CH3)2CH-Cl) = 82 kcal/mole D((CH CH-OH) = 92 kcal/mole 3)2 but nonetheless insertion occurs. In general for favorable metal ion insertion into R-X, the following expression must be true: D(R-Co+) + D(X-Co+) > D(R-X) [Note that D(A-B) is the dissociation energy of the A-B bond]. The fact that metal insertion into the C-N bond is never observed probably indi- cates that the (Co+-NH2) bond is weak. Thus making insertion into C-N energetically inaccessible i.e. 42 + + Co + CH3NH2 €X9——;v CH3-Co ~NH2 '. D(Co+-CH3) + D(Co+-NH2) < D(CH -NH2) 3 D(Co+-CH3) = 61 kcal/mole52 D(CH3-NH2) = 80 kcal/mole '. D(Co+-NH2) must be less than l9 kcal/mole Therefore our results place a lower limit on the (Co+-NH2) bondstrength. Figure 3 shows the bondstrengths of the bonds in the eight amines under study. Note that C-C and C-N bond strengths are within 1.4 kcal of each other. Thus when Co+ approaches an amine, it finds all of the skeletal bonds of equal strength. It cannot insert into C-N, so it preferentially attacks C-C or C-H bonds and these are the bonds into which Co+ inserts. If one takes the Co+-CH bond as a typical Co+-C bond (6l kcal/ 3 mole), Co+ could insert into any C-C bond (2 x 6l kcals of energy are greater than any carbon-carbon bondstrength). Also since the Co+-H has 52 C-H bonds are also accessible to a dissociation energy of 52 kcal/mole, Co+ insertion. Thus, it is probably the energetic limitations which lead to the formation of reaction products in amines which resemble those observed for alkanes and not those observed for other polar organics. Consider again reaction type D (CH4 elimination) for n-propylamine. From Figure 3 it is evident that the middle C-C bond in n-propylamine is the weaker of the two C-C bonds. When Co+ interacts with n-propylamine it appears to insert preferentially into this weaker C-C bond. Thus, we do not observe the elimination of CH4 as a result of insertion between 43 79. CH2 —— NH ‘94 4 94.6 Methylamine H H CH3 JE— CH2 "'3 NH2 Ethylamine CH3§2—'—3—CH2-7—9—'—1—CH2§-9—=—LNH2 n-Propylamine CH3 8.6 ’//CH 76'3 NH2 isopropylamine H30 78.6 CH 3 H C | 3 78.5 7l .6 \ 73.13 CH HZC C ——L NH2 CH—7g'3‘ 2\ NH 95.7 {-75.96 H3C 2 H CH3 isobutylamine t-butylamine Figure 3 Amine Bond Strengths (kcal/mole) 44 CH2 CH2 NH2 78.8 80.8 95.7 /////’82:>\\ ///// \\\ ’////;:-4 H CH2 CH2 n-butylamine CH 3 I CH3 CH2 82.9 75.3 73.9 CH2 NH 96.9 H sec-butylamine Figure 3 (con't) 45 the carbon of the terminal methyl group and the adjacent carbon, rather, insertion into the other C-C bond occurs and one observes the formation of Co(CHSN)+ and C2H4 as was illustrated in reaction (30). Reaction type E was observed for Co+ and isobutylamine: + Co + >’\NH2————-) I +\ NH + CH4 + H2 (34) Co This reaction would proceed first through the mechanism of reaction (33) which involved metal ion insertion into a terminal carbon-carbon bond followed by B-hydrogen atom shift and CH4 elimination. The driving force for H2 elimination following CH4 elimination is the formation of a conjugated system which can strongly complex to Co+. Reaction (34) would not be expected to occur for methyl, ethyl, n-propyl, isopropyl, sec-butyl or t-butyl amine due to structural considerations and was not observed for these amines. n-Butyl amine has the potential to elimin- ate CH4 and H2 in a manner similar to that for isobutylamine but it was not observed to do so. Reactions F, G, and H can also be explained by a mechanism involv- ing Co+ insertion into a C-C bond followed by elimination of CZH4 or a small alkane. These reactions were observed only for the butylamines. Reactions F and G were observed exclusively for n-butylamine. NH Co+ + W 2 * 41C0(C2H7N)+ + C2H4 (35) + Co(CH3N) + C3H8 (36) Depending on which C-C bond into which the Co+ inserted, we would observe the following reaction sequences for (35) and (36): 46 ’:::3CH2 HZC-o -Co -CH2-CH2 NH2 H2C‘~Co+'———CH2-CH2NH2 C0 4' WNHZ --) If '—) H/ H2 C- H 4' ( 35) + H H C-CH -CH H3C’CH2'CH2 + 3 2 2 H Co +vaH2—> C\o+\ —-) i0 -—--> H (:———N/H -\\ N 2 2 \\ \H H Co+ (36) E + C3H8 H C=NH 2 Reaction (35) is identical to reaction (30) except that in (35), Co+ has inserted instead into the adjacent C-C bond in the chain. Reac- tion 6 was observed for both n-butyl and sec-butyl amine and can be explained by a similar mechanism: H3C-CH2 Co+ + NH ——) Co+ \v’\v’ 2 Ik‘-¥ HZC— CH -— NH2 H3C-ir2 Co+ NH "32 1 2 COLH =/ (37a) . + H c-—-CH-—-NH C2H6 CH \u——1~NH --)» I 3 ?”2 --e» 47 H C-H C CH3 Co-——.CH / \‘l —>HN =/ + C2H6 (37b) H «NH 0.... 0 Therefore, the fact that we observe reactions arising from all possible cases of Co+ insertion into C-C bonds in n-butylamine indicates that all the C-C bondstrengths are similar. Also G reveals that the molecular structure plays a role in determining the reaction products. This metal ion insertion type mechanism has been able to explain the observed results in reactions A-H. Thus the mechanism of reaction between Co+ and amines appears to be analogous to that reported for Co+ and alkanes. In most cases the addition of one carbonyl ligand to the metal ion center did not affect the reactivity of the metal as is evident in Table 9. Thus, the reaction with CoCO+ would proceed as it did with Co+ alone and the carbonyl ligand would either be retained in the pro- duct ion or it would be lost as a neutral molecule. This fact is illustrated by ethylamine: Co+ + C H NH + 2 5 2 ————-———>C0(C2H5N) +H2 (38) + CoCO+ + CZHSNHZ CoCO(CzHSN) + ”2 (39a) + CO(CZH5N) + C0 + H2 (39b) 48 Reactions analogous to (38) and (39) were not observed in five cases for amines. In the reaction of methylamine with Co+, H2 elimina- tion was observed. But in the reaction of CoCO+ with CH3NH2, it was not observed. Methylamine is the smallest of the amines. The presence of the carbonyl group on Co+ may be enough to deter this reaction with this small amine. The significance of this observation in the other four cases is not readily determined. For example, Co+ reacts with isobutylamine to eliminate two hydrogen molecules but CoCO+ does not. However, both Co+ and CoCO+ react with isobutylamine to eliminate CH4. Thus there may be a series of competing reactions between certain amines and CoC0+. Perhaps the reactions that are observed are the ones which are more favored energetically. In the intermediate illustrated /50 * + CoCO+ + amine-————) Co ——>products amine J usually the lowest energy process is Cleavage of the weak Co+-C0 bond. The dissociation energy of this bond is not known, but we do know that the following reaction occurs: + + CoCO + C2H4 ————-> Co(C2H4) + .CO and D(Co+-C2H4) = 37 1.2 kcal/mole. Hence, we would expect D(Co+-CO) to be less than 37 kcal/mole. An interesting reaction was observed for ethylamine: CoC0+ + ANHZ —-> Co(CHsN)+ + "czH20" (40) Note that Co+ only reacted with CH3CH2NH2 by H2 elimination. A mechanism 49 involving CoCO+ insertion into a carbon-carbon bond followed by the shift of a hydrogen atom B to the metal onto the metal and then elimina- tion of H2C=CO can explain this result: 0 o Coco+ + /\NH2 ——> H2073” §\Co+-CH2NH2 —--) HzC-éC’ /~._.«" .fto+-CH NH H ’,r 2 2 ”I H C=C-o + C +-CH NH 2_ I022 H This is further support for the concept of a necessary H atom on a car- bon which is B to the metal after insertion. A similar mechanism has been proposed for the reaction of CoCO+ with isopropyl chloride.53 When the number of carbonyls on the metal ion increased from one to two, ligand substitution became the dominating process over chemical "reactions“. In ligand substitution an amine molecule substitutes for a carbonyl on the metal ion. The following ligand substitutions were observed for all of the amines: (Note Am stands for the parent amine) Coco+ + Ami-——-—-> Co(Am)+ + co (4l) + + CO(C0)2 Am C: COC0(Am) + C0 (42) Co(Am)+ + 2C0 (43) when a nitrosyl ligand was present on the metal ion, the only processes observed were ligand substitutions: 50 CoCONo+ 4 Am __, CoN0(Am)+ + C0 (44) Co(C0)2N0+ + Am CoCONO(Am)+ + CO (45) —l:: CoN0(Am)+ + 2C0 (46) Co(C0)3N0+ + Am CoC0N0(Am)+ + 2C0 (47) _l:: CoN0(Am)+ + 3C0 (48) Note that all ligand substitutions involved the displacement of the carbonyl ligand by the parent amine. The energy needed to displace the carbonyls comes from the energy released in the formation of the new bond formed between the metal and complexing amine. Therefore, the number D(Co -C0) (49) Reaction types I and J in Table 9 refer to the ligand substitutions just discussed. 6 At pressures greater than 5 x 10' torr, one observes the presence of Co CONO+ and C02(C0)2N0+ in the mass spectrum of Co(CO)3N0. These 2 species are formed in ion-molecule reactions of the ion fragments pro- duced by electron impact on Co(C0)3NO and neutral Co(CO)3N0. The ligand substitutions observed for these ions are listed under reaction type M in Table 9. Up to two carbonyls were observed to be diSplaced by an 51 amine in this case. Reaction types K, L and N are interesting because they are ligand substitutions of product ions. In each case carbonyl ligands are dis- placed by parent amines. Also observed were direct attachment reactions. In a direct attachment reaction, an amine "attaches" directly to the metal without displacing any ligands. The following direct attachments were observed for amines: Co+ + Am1-————-9 CoAm+ + + CoAm + Am ————9 Co(Am)2 CoNOAm+ + Am ————> CoN0(Am)2+ Thus not only have a characteristic series of ligand substitutions been observed but also characteristic attachment reactions have been observed to occur in gas phase mixtures of Co(CO)3N0 and amines. Direct attach- ment may occur in a three-body collision although these are unlikely under ICR conditions. Chemical Ionization (CI) Mass Spectrometry is a technique in which sample molecules undergo ion-molecule reactions with a reagent ion and thus become ionized. Because the energy transferred to the sample mole- cule in an ion-molecule reaction is less than that transferred upon ionization via conventional 70eV electron impact, CI is considered a "soft" ionization method. As a result, the sample molecule is less likely to undergo fragmentation. Therefore, the molecular weight of the sample can be determined. Furthermore, unlike 70eV electron impact, reagent ions which react very specificially with sample molecules can 52 be Chosen. CI reactions frequently yield structural information on the sample molecule. For a species to be considered a useful CI reagent, one must establish a basis for predicting the CI mass spectra of the reagent for analogous sample molecules. Freiser gt_gl, has shown that the Cu+ metal ion may serve as a useful CI reagent in his study of Cu+ with esters and ketones.36 The above discussion has illustrated the utility of Co(C0)3N0 as a CI reagent. It has been shown that Co+ produced from 70eV electron impact reacts specifically with amines. The mechanism consistent with the formation of product ions involves oxidative addition of Co+ to a C-C or C-H bond followed by B-hydrogen atom transfer to the metal and then the elimination of a small neutral molecule such as H2, CH4, C2H4, C H or C H Knowledge of this mechanism allows one to interpret the 3 6 3 8‘ mass spectrum of amines assigning structures to all the product peaks. Thus one can formulate specific structural information on the amine based on this proposed mechanism. Also a characteristic series of ligand substitutions and direct attachment reactions were observed. These reactions yield molecular weight information. For example, every amine directly attached to Co+. Thus the peak corresponding to the molecular weight of the amine plus 59 was seen in the spectrum of each amine. In conclusion, this study has allowed one to examine the basic chemistry of the isolated Co+ metal center with the amine series and determine how this chemistry changed as the available orbitals on the metal became occupied with C0 and N0 ligands, and how the RNH2 chemistry 53 depended on the structure of R. A mechanism was proposed consistent with the observed results. The potential of Co(C0)3N0 as a CI reagent was illustrated. LIST OF REFERENCES 10. 11. 12. 13. 14. 15. 16. LIST OF REFERENCES J.L. Beauchamp, Ann. Rev. Phys. Chem., 33, 527 (1971). T.A. Lehman and M.M. Bursey, Ion Cyclotron Rosononcc Spectroscopy, J. Wiley and Sons, New York (1976). J.D. Baldeschwieler, Science, 133, 263 (1968). €.D. Baldeschwieler and 5.5. Woodgate, Acc. Chem. Res., 3, 114 1971 . C.J. Drewery, G.C. Goode, K.R. Jennings, "Ion Cyclotron Mass Spec- trometry," MTP International Review of Science-Physical Chemistry, Series One, Volume Five, Mass Spectrometry, A. Macoll, Ed., Univer- sity Park Press, Baltimore, 183 (1972). J.M.S. Henis, "Ion Cyclotron Resonance Spectrometry", Ion-Molecule Reactions, Vol. 2, J.L. Franklin, Ed., Plenum Press, New York, 395 1972 . G.C. Goode, A.J. Ferrer-Correia, and K.R. Jennings, Int. J. Mass Spec. Ion Phys., 3, 229 (1970). D. Wobschall, Rev. Sci. Instr., 33, 466 (1965). G.A. Junk, H.J. Svec, Z. Naturforsch, 33, 1 (1968). R.E. Winters, R.W. Kiser, J. Inorg. Chem., 3, 699 (1964). E. Schumacher, R. Taubenest, Helv. Chim. Acta, 41, 1525 (1964). R.C. Dunbar, J.F. Ennever, J.P. Fackler, Inorg. Chem., 13, 2734 (1973). M.S. Faster, J.L. Beauchamp, J. Am. Chem. Soc., 33, 4924 (1971). D.F. Hunt, J.W. Russell, and R.L. Torian, J. Organometal Chem., 33, 175 (1972). W.P. Anderson, N. Hsu, C.W. Stranger, and B. Munson, J. Organo- metal. Chem., 33, 249 (1974). J. Muller, "Ion Molecule Reactions of Organometallic Complexes", Advances in Mass Spectrometry, Vol. 6, A.R. West, Ed., Appl. Sci. Publishers, Ltd., England, 823 (1974). 54 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 31. 32. 33. 34. 35. 36. 55 3.5. Kraihanyel, J.J. Conville, and J.E. Sturm, Chem. Comm., 133, 1971 . J. Allison, D.P. Ridge, J. Organometal. Chem., 33, C-11 (1975). M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc., 31, 4808 1975 . M.S. goster and J.L. Beauchamp, J. Amer. Chem. Soc., 31, 4814 1975 . R.R. Corderman, J.L. Beauchamp, J. Am. Chem. Soc., 3998, (1976). R.R. Corderman and J.L. Beauchamp, Inorg. Chem., 13, 665 (1976). J. A11ison, D.P. Ridge, J. Amer. Chem. Soc., 33, 7445 (1976). R.R. Eorderman and J.L. Beauchamp, J. Amer. Chem. Soc., 33, 5700 1976 . R.D. Bach, A.T. Weibel, J. Patane and L. Kevan, J. Amer. Chem. Soc., 33, 6237 (1976). R.V. Hodges and J.L. Beauchamp, Anal. Chem., 33, 825 (1976). J. Allison, D.P. Ridge, J. Am. Chem. Soc., 33, 35 (1977). G.H. Weddle, J. Allison, D.P. Ridge, J. Amer. Chem. Soc., 33, 105 (1977). T.G. Dietz, D.S. Chateliier, D.P. Ridge, J. Am. Chem. Soc., 133, 4905 (1978). J. Allison and D.P. Ridge, J. Am. Chem. Soc., 133, 163 (1978). . Allison, R. Kinser, T.G. Dietz, M. deAngelis, D.P. Ridge, . Amer. Chem. Soc., 133, 2706 (1978). Allison and D.P. Ridge, J. Am. Chem. Soc., 191; 4998 (1979). CJC—ILI R.C. Burnier, T.J. Carlin, W.D. Reents, Jr., R.B. Cody R.K. Lengel, B.S. Freiser, J. Am. Chem. Soc., 131, 7127 (1979). J. Allison, R.B. Freas, D.P. Ridge, J. Amer. Chem. Soc., 131, 1332 (1979). J.L. Beauchamp, A.E. Stevens, R.R. Corderman, Pure & Appl. Chem., 31, 967 (1979). R.C. Burnier, G.D. Byrd, B.S. Freiser, Anal. Chem., 33, 1641 (1980). 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 56 P.B. Armentrout, J.L. Beauchamp, Chem. Phys., 33, 37 (1980). (13° Armentrout, J.L. Beauchamp, J. Am. Chem. Soc., log, 1736 80 . R.W. Jones and R.H. Staley, J. Am. Chem. Soc., 133, 3794 (1980). J.S. Uppa1, R.H. Sta1ey, J. Am. Chem. Soc., 133, 4144 (1980). P.B. Armentrout, J.L. Beauchamp, J. Chem. Phys., 13, 2819 (1981). L.F. Halle, P.B. Armentrout, J.L. Beauchamp, J. Am. Chem. Soc., .133, 962 (1981). P.B. Armentrout, J.L. Beauchamp, J. Am. Chem. Soc...lQ§4 784 1981 . A.E. Stevens, J.L. Beauchamp, J. Am. Chem. Soc., 133, 190 (1981). P.B. Armentrout, L.F. Halle, J.L. Beauchamp, J. Am. Chem. Soc., 133, 6624 (1981). 3.8. Armentrout, J.L. Beauchamp, J. Am. Chem. Soc., 133, 6628 1981 . J.S. Uppal, D.E. Johnson, R.H. Staley, J. Am. Chem. Soc., 133, 508 (1981). R.C. Burnier, G.D. Byrd, B.S. Freiser, J. Am. Chem. Soc., 133, 4360 (1981). J. Halpern in "Organotransition Chemistry", Y. Ishii and M. Tsutsui, Plenum Press, New York, N.Y., p. 109, 1975. A. Wagnick, L.R. Anders, T.E. Sharp, Rev. Sci. Instrum., 33, 929 1974 . J. Allison, private communication. P.B. Armentrout, L.F. Halle, J.L. Beauchamp, J. Am. Chem. Soc., 133, 6501 (1981). J. Allison, Ph.D. Dissertation, University of Delaware, 1976.