MSU LIBRARIES m \— RETURNING MATERIALS: P1ace in book drop to remove this checkout from your record. [lflgg w111 be charged if book is returned after the date stamped be1ow. METAL ION CHEMICAL IONIZATION —- THE GAS PHASE CHEMISTRY OF COBALT CONTAINING IONS BY Michael Lombarski A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1984 This is dedicated to my family for all the love and support they have given me in my life. This is also dedicated to my undergraduate advisor, the late Dr. Al Steyermark. -11- ACKNOWLEDGEMENTS I would like to thank Dr. John Allison for being my advisor the past years. I have learned much from him, both about chemistry and about peOple. Thanks goes to Mike Nowak for the help on purifying the substituted acids. Also for being a friend. Thanks to all the peOple associated with the Aromatics and Free Radicals, these peOple know how to make softball enjoyable. I want to give special thanks to my friends from back home. Dave Deininger, whose letters and phone calls always were good for a laugh or two around here. I don't know anyone else who has such strange taste in music. A very special thank-you goes to Diane Siedlecki, one person who always believed in me, even at times when I didn't believe in myself. Whenever I found myself needing to talk to someone she was there. I know that in her I will always have a friend. I'd like to thank all the support personnel in the electronics, machine and glassblowing shops, without their assistance the ICR and myself would not be where we are today. ~iii- ABSTRACT Metal Ion Chemical Ionization - The Gas Phase Chemistry of Cobalt Containing Ions The ion/molecule reactions of Co+ with thiols, butenone, butenone, carboxylic acids, allylic compounds and 2-chloro- and 2-bromoethanols are reported. Primarily, Co+ reacts with thiols via the metal insertion/B-H atom shift/competititve ligand loss mechanism that has been used previously to eXplain the chemistry of gaseous transition metal ions with organic molecules containing other functional groups. With ketones, Co+ initially attacks at the carbonyl oxygen from which insertion into either C-CO bond may occur, followed by alkyl or 8-H atom shifts. In carboxylic acids, the Co+ reaction products are indicative of metal insertion into either the C-CO or the C-OH bond as the first step, with insertion into the C-OH bond predominating. Co+ prefers to attack at the substituted group of the allylic compounds; no reactions typical of olefins were observed. In the chemistry of Co+ with 2-chloroethanol attack at the halides occurs to eliminate HCl. Also observed is the elimination Lombarski, Michael of ethene which is explained through the formation of a five-membered ring intermediate. Similar reactions are observed in 2-bromoethanol. Also investigated was the ability to predict reaction products of bifunctional organic compounds based upon the known chemistry 0f each functional group. The ability to predict products varied from 0% for 2-brom0propanoic acid to 100% for other substituted acids and butanones. From the branching ratios the following trend for site of attack by Co+ is suggested: -COOH > -Br > -C1 > -OH > C=O > -H CHAPTER TABLE OF CONTENTS LIST OF TABLES. . . . . . . . . . . . . . . . LIST OF FIGURES. . . . . . . . . . . . . . . INTRODUCTION. . . . . . . . . . . . . . . . . EXPERIMENTAL. . . . . . . . . . . . . . . . . A. Principles of Ion Cyclotron Resonance. . B. Signal Intensities. . . . . . . . . . . . C. Ion Cyclotron Double Resonance. . . . . . D. Experimental Set-Up. . . . . . . . . . . THE CHEMISTRY OF COBALT-CONTAINING IONS WITH THIOLS. . . . . . . . . . . . . . . . A. The Chemistry of Co+ with Thiols. . . . . l. Ethanethiol. . . . . . . . . . . . . 2. Propanethiols. . . . . . . . . . . . 3. Butanethiols. . . . . . . . . . . . . B. The Chemistry of Co(CO)x+ and Co(CO) NO+ Thiols. The Chemistry of Cobalt-Containing Ions With Butanone and 3-Butenone. . . . . . . . . . A. (X'=l, 2; y'=0- 3) Ions With Ligand Effects. . . . . . . . . The Chemistry of Co+ With Butanone and 3-Butenone. . . . . . . . . . . . . . . . l. Butanone. . . . . . . . . . . . . . 2. 3- Butenone. . . . . . . . . . B. The Reactions of CoLn +. with Butanone and Butenone. . . . . . . . . . . . . . . THE CHEMISTRY OF COBALT- CONTAINING IONS WITH UNSUBSTITUTED CARBOXYLIC ACIDS. . . . . . . A. The Chemistry of Co+ with Unsubstituted Carboxylic Acids. . . . . . . . . . . . . Formic Acid. . . . . . . . . . . . . . Acetic Acids. . . . . . . . . . . . . Propanoic Acid. . . . . . . . . . . . . Butanoic Acid. . . . . . . . . . . . . 3- -Butenoic Acid. . . . . . . . 2, 2- -Dimethylpropanoic Acid. . . . . Ligand Effects. . . . . . . . . . . \lO‘U‘I-waP-J -iv— PAGE vi .10 .16 .17 .26 .36 .52 .54 .58 .58 .64 .64 . 64 .66 .70 . 75 . 79 .82 .82 ix CHAPTER 6 10 PAGE THE CHEMISTRY OF COBALT CONTAINING IONS WITH ALLYL-X COMPOUNDS. . . . . . . . . . . . . 86 A. The Chemistry of the Cobalt Ion with Allyl Compounds. . . . . . . . . . . . . .87 B. The Chemistry of CoLn+ Ions with Allyl Compounds. Ligand Effects. . . . . . . . . 98 THE CHEMISTRY OF COBALT-CONTAINING IONS WITH ' 2-CHLORO- AND 2-BROMOETHANOL. . . . . . . . . .103 1. Z-Bromoethanol. . . . . . . . . . . . .110 THE CHEMISTRY OF COBALT- -CONTAINING IONS WITH SUBSTITUTED BUTANONES. . . . . . . . . .115 A. The Chemistry of Co+ with 3- Substituted Butanones. . . . . . . . . . . . . . . 117 B. The Chemistry of CoLn+ Ions with 3- Substituted Butanones. . . . . . . . . . 130 THE CHEMISTRY OF COBALT-CONTAINING IONS WITH SUBSTITUTED CARBOXYLIC ACIDS. . . . . . . 135 A. Reaction of Co+ with Substituted Carboxylic Acids. . . . . . . . . . .135 1. The Chemistry of Co+ with Substituted Acetic Acids. . . . . . . .135 2. The Chemistry of Co+ with 2- Substituted Propanoic Acids. . . . . . 139 3. Reactions of Co with 3- Substituted Propanoic Acids. . . . . . 141 B. The Chemistry of CoLn +with Substituted Carboxylic Acids. Ligand Effects. . . . . 145 1. The Chloro-Acids. . . . . . . . . . . .145 2. The Bromo—Acids. . . . . . . . . . . . 151 3. The Hydroxy-Acids. . . . . . . . . . . 156 4. The Mercapto-Acids. . . . . . . . . . 161 SUMMARY AND CONCLUSIONS. . . . . . . . . . . . 167 APPENDIX A-PERTINENT PROTON AFFINITIES. . . . 173 APPENDIX B-—WATER GAS SHIFT REACTION. . . . . .174 APPENDIX C-USING THE ZX 81 TO CONTROL TRAPPED ICR EXPERIMENTS. . . . . . 181 LIST OF REFERENCES. . . . . . . . . . . . . . .190 —V- TABLE 10 ll 12 13 14 LIST OF TABLES Reactions of the ions derived from Co(CO)3NO with ethanethiol. . . . . . . . Reactions of the ion derived from Co(CO)3NO with 1-propanethiol. . . . . . . . Reactions of the ions derived from Co(CO)3NO with 2-propanethiol. . . . . . . . Reactions of the ions derived from Co(CO)3NO with 1-butanethiol. . . . . . . . Reactions of the ions derived from Co(CO)3NO with 2-butanethiol. . . . . . . . Reactions of the ions derived from Co(CO)3NO with 2-methyl-1-propanethiol. . Reactions of the ions derived from Co(CO)3NO with 2-methyl-2-propanethiol. . . Reactions of the ions derived from Co(CO)3NO with butanone. . . . . . . . . . . Reactions CoCl +c H I (35%) (1) 3 2 3 7 ——> CoC H + +HCl (60%) 3 6 + —> CoHCl +C3H6 (5%) + + Co + (CH3)2CHOH —> CoOH +C3H7' (12%) (2) + ——> CoC3H6 +1120 (48%) + —> Cono +C3H6 (40%) These reactions can be explained through a common mechanism (Scheme I). The metal ion inserts into the C-X (X=Cl,OH) bond, followed by either a charge transfer process to form C3H7T, or a 8-H atom shift to the metal center to give the Co(HX)(olefin)+ complex. Next the HX and olefin ligands compete for sites on the metal in a competitive ligand loss process. Other halides and alcohols have also been observed to react like 2-chloro-propane and 2-pr0panol. The Co+ ion reacts with ethyl iodide via insertion into the C-I bond followed by either a bond cleavage to give CoI+ (11%), or a 8-H atom shift to give (ethene)Co+(HI) complex which then undergoes competitive ligand loss to give CoHI+ (11%) +XIoU A I/ aim /=+ 00 AllllllIX ..... +00 ..... =\AIII +00 ..... AIIIHIIXIII+00IIA TX||A++00 \ x xx- H mEmzum -4- and CoC2H4+ (78%). Co+ reacts with ethanol via the same + (87%). mechanism to give CoH20+ (13%) and CoC2H4 One area of research in our laboratory deals with the chemistry of metal ions with nitrogen containing compounds. The bare cobalt ion, Co+, shows a rich chemistry with 2-nitropropane. There are three reaction products that are formed via the same mechanism as the chloride and alcohol to give the observed products [6]: + + CO -+(CH3)2CHNOZ -—€> COC3H6 -+HNO2 ( 3%) + ——€> COHNO2 -+C3H6 (16%) ——> c H I +CONO (17%) 3 7 2 ——€> other products (64%) From the reactions of Co+ with these three compounds, one would conclude that Co+ would insert into the polar C-X bond in all compounds, therefore making prediction of metal ion-CI spectra relatively straightforward. For example, using the mechanism in Scheme I, one might predict that 2-amin0pr0pane would react with Co+ to give + the products CoNH + and/or CoC3H6 . This, however, is not 3 observed. For primary amines, insertion of Co+ into the N-H, C-H, and C-C bonds dominates the chemistry [7], resulting in H2 and CH4 elimination and the formation of C3H8N+ (with the loss of the neutral CoH). -5- Although the sites of attack in amines by cobalt varies from the molecules containing other polar functional groups, the same general mechanism -— metal ion insertion/B-H atom shift/competitive ligand loss - can still be used to explain the proddcts. This general mechanism can be used to explain approximately 90% of all products in Z-propyl- amine. The failure of Co+ to insert into the C-NH2 bond has been proposed to be due to the weak Co+-NH bond that 2 would be formed. This is just one example where Co+ does not insert into the polar C-X bond. In compounds containing no polar bonds (alkanes and alkenes), the chemistry of the Co+ is dominated also by the metal ion insertion into the C-H and C-C bonds as the first mechanistic step [8,9]. From these examples, it is illustrated that the ion/molecule reaction products of species such as Co+ with organic molecules could be predicted if the initial sites of attack are known. It is apparent that Co+ prefers to attack at the polar bond in small monofunctional molecules, except in the anomalous case of primary amines. Basic studies involving molecules representative of various functional groups continue to be a major area of research for gas phase metal ion chemists. Most compounds being analyzed by mass spectrometry today are not simple monofunctional molecules, but rather complicated multifunctional molecules. To date, only three reports [4,10,11] have dealt with the ion/molecule -6- reactions of Co+ with bifunctional compounds to any extent. In the first report, (based upon some of the work in this dissertationL.the chemistry of gaseous transition metal ions with molecules containing two functional groups on adjacent carbons were studied. Results indicate that the reactions observed will depend upon the nature of the two groups. In 2—chlorobutanone, reactions of both functional groups are observed. In allylic compounds the reactions of just one group are seen. In compounds such as 2-chloroethanol one observes the reaction of one group and also a reaction where the two functional groups behave as a "new group" [4]. The second report [10] deals with two functional groups that are removed from each other, i.e., 1,4-bisubstituted butanes. Results indicate that Co+ would prefer to react at one site over another (halogen >OH), and then eliminate the second functional group, as HX, to give a Co(butadiene)+ complex [10]. Also it should be noted that the reactions of Co+ with l-butanol and l-chloro— and l-bromobutanes are more complicated than the prOpyl and ethyl alcohols and halides discussed earlier. After insertion of the Co+ ion into the C-Cl or C-Br bond a charge transfer occurs just as in 2-chloropr0pane, but elimination or retention of HCl or HBr is not observed. What is observed is the loss of H and HCl or HBr 2 yielding Co+(butadiene). In 1-butanol, the formation of -7- Co+(H20)H2(butadiene) occurs with the subsequent loss of H2 or H2 and H20. The last report deals with the chemistry of ions derived from electron impact on Fe(CO)5 and Cr(CO)6 with cyclic and linear polyethers [11]. With acyclic polyethers both bare metal ions, Fe+ and Cr+, react by insertion into a C-O bond, followed by a 8-H atom shift, and a ligand elimination process. In the cyclic polyether (12-crown-4) multiple metal-ligand interactions are observed. From the initial complex, insertion of the metal into two C-O bonds has been suggested, followed by two 8—H atom shifts to give the observed products [11]. When mass spectrometry (MS) is used for chemical analysis, molecular weight, functional group and structural information can frequently be obtained. Of these, molecular weight information, as observed by the formation of the molecular ion, Mt, is the most useful. Not only is the molecular weight obtained from the parent ion, but, if the M? ion is of significant abundance the elemental composition can be determined by isotopic abundance. Unfortunately, many compounds do not give the molecular ion under electron impact (EI) [12]. In 1966, Munson and Field [13] demonstrated the analytical applications of Chemical Ionization (CI) by using CH4 as a reagent gas and generated spectra in which + and C H + (the 5 2 5 reagent ions) in the gas phase to form the protonated the analyte molecules reacted with CH -8- molecular ion, (M+H)+, in fairly high abundance. Chemical ionization is considered a "soft" ionization method since less fragmentation is induced as compared to E1. Unfortunately, due to the small amount of fragmentation that may occur in CIMS, relatively little functional group and structural information is produced. As discussed earlier, metal centers appear to react in very Specific ways with various functional groups. Because of this specific chemistry, metal ions may be used as CI reagents. Freiser, Byrd and Burnier [3] were the first to consider metal ions as CI reagents. In an effort to evaluate the utility of Metal Ion Chemical Ionization (MICI), generation of some simple rules for spectral interpretation would be much more desirable than the need for generating a Co+-CI mass spectral library for compound identification. In an attempt to generate rules of reactivity for multi- functional compounds, we are attempting to answer three basic questions: 1) When will Co+ react with bifunctional molecules to give results typical of each functional group? 2) When will only one group dominate? 3) When will the two groups behave as a "new group"? The work to date suggests that the answers to these questions will depend not only upon the nature of the two functional groups but also upon their positions relative to each other. The second aspect of this dissertation will deal with the chemistry of various bifunctional organic -9- molecules with Co+. From the results of these experiments, various guidelines for bifunctional compounds were determined. In most of the studies to date, the reactions of the bare metal ion with the various functional groups were discussed in detail, with very little reported on the chemistry of the other ions formed from electron impact on metal carbonyls, in this case Co(CO)3NO. To be specific, the answers to the following questions were sought in the chemistry of various CoLn+ ions with monofunctional and bifunctional organic molecules- is the chemistry of these ions identical to Co+? is ligand substitution the only (or dominant) process? Or, do the new reactions based upon Co(ligand)+ insertion into a bond rather than Co+ insertion into that bond occur? In this dissertation, the discussion of the chemistry of Co+ with each compound is followed by a shorter discussion on the chemistry of the various CoLn+ ions with that compound. CHAPTER 2 EXPERIMENTAL A. Principles of Ion Cyclotron Resonance Ion cyclotron resonance spectrometry is a technique for studying gas phase ion/molecule reactions. Several articles and books have been published on ICR and Fourier Transform Mass Spectrometry (FTMS) theories and principles of operation since 1966 [14-28]. What follows in this chapter is a brief discussion of the principles behind the ion cyclotron resonance experiments. In this technique, ions are produced by electron impact as in other mass spectrometric methods but their subsequent path and means of detection differ greatly from those in a conventional mass spectrometer. For an ion of mass m, acceleration a, charge e, and a velocity in which v is the velocity vector component perpendicular to the magnetic field B, the force exerted on the ion is given by the equation: F = ma = evB = -—— (l) -10- -11- and this force will be perpendicular to both the magnetic field and velocity. Rearranging Equation (1), one can obtain Equation (2) which describes the motion of the ion in the magnetic field. E! = e8 (2) As a result the ion's net motion is a circular path (Figure l) where r is the radius of the orbit. AA‘ I’ \ x Figure 1. Motion of an ion in a magnetic field. -12- If velocity has units of cm/sec, and r is in centimeters, one arrives at m = B 3 me e ( ) where we is the angular cyclotron frequency. Rearrangement of Equation (3) gives w = 92 rad C m SEC _ eB( cycles ) V _ c an sec At a constant magnetic field and unit charge, it can be seen from the above equations that heavier ions have a lower cyclotron frequency. It is also important to note that the cyclotron frequency, we or vc, is independent of the velocity of the ions. For detection of a signal in the ICR spectrometer an amplitude limited marginal oscillator is used. The oscil- lator may be considered a constant current generator which drives a parallel LC circuit at a frequency wm.o. (Figure 2). The ICR cell (Figure 3) forms part of the capacitance of this circuit. When the cyclotron frequency equals the marginal oscillator frequency (mc== ) the ions absorb power (1.) 111.0. from the LC circuity. The output of the marginal oscillator reflects the voltage in the circuit. Since the oscillator -13- Figure 2. I ource IAnalyzet‘lcllector I I r.f frov | ' ‘ marsInnl Of‘iIIILOt”‘—_“1 I l I I LTUI Ilah.‘ ‘ II I 1op Analyzer I I Dnutle hrconintc r.f Oscillator ' trift Ilate ' Top Lource _______L_ Collector \[:3 Trapping Plates Source I I I I I I I l I Drift Flats Bottom Source FIgure 3. ICR cell. C _T- ']_CeII L LC tank circuit of marginal oscillator. Tanpinn Flater “ Al‘.“1).~"r €::::‘I11anent [rift Plate Bottom Analyzer -14- is driven by a constant current, any change in power (I*V) reflects a change voltage in the circuit. The change in voltage ( A‘V) is related to the signal intensity. As an ion absorbs power, the ion gains speed and energy. Therefore to maintain its cyclotron frequency, the ion increases its radius in proportion to its gain in velocity. For example if an ion has a cyclotron frequency of 10 orbits/sec and suddenly goes twice as fast, it must increase its orbital radius, its orbital frequency or both. Since the orbital frequency is dependent on mass only, and not energy (Equation (3)), the radius alone will increase. Figure 4. Motion of ion as it absorbs energy. One method of conducting an ICR experiment is by keeping the marginal oscillator at a constant frequency and varying the cyclotron frequency of the ions by scanning the magnetic field. For convenience, the normal marginal oscillator frequency used is 153.57 KHz; this -15- will cause a change of one mass unit per change in magnetic field of 100 gauss (.01 tesla) (assuming unit charge). v _ 1.6021 xIO'19 C><0.01 T C 6.2832 x1 u x1.66O4 xio'23 kg/w = 1.5359 x105 +2 = 153.57 kHz (4) Other marginal oscillator frequencies are used to increase the mass range (lower M.O. frequency). Decreasing the frequency has the disadvantage of decreasing the resolution and signal intensity.- Equations (1) and (2) describe the force exerted upon nonresonant ions; for resonant ions several other equations describe the force exerted and the power absorption. The force exerted is given by Equation (5): I'D _ rf F - 2 (5) where Erf is the amplitude of the marginal oscillator and the acceleration is E e rf a = 2m and the additional velocity of the ion in the oscillator field at a frequency w is defined by M.O. -16- It It erfe dt _ Erf et The instantaneous power absorbed A is the product of the force on the ion and the ion's velocity: 2 2 rfet 4m 8 A(t) = Fv The average power drawn by the ion from the rf field while in the detector for a time (0 c++o > E++F G++B -—-> C++D+H -——> 1?.+ +F+I in which the two ionic products are produced competitively from the reaction of A+ and G+ with B. Only by ejection + . . + + + of A can the contr1but1on of A to C and E be determined. D. Experimental Set—Up All experiments were performed using an ion cyclotron resonance (ICR) spectrometer of conventional design that was built at Michigan State University. The three section cell is 0.88 in. X0.88 in. X6.25 in. The source region is 2.00 in. long, the analyzer region is 3.75 in. long and the collector is 0.50 in. long. Ions are formed -20- in the source by electron impact with 70 ev electrons. The electron filament is emission regulated. The filament controller and plate voltage controller for the ICR cell were designed and constructed at MSU. The instrument is operable in the pulsed or drift ICR modes. Data in this dissertation were obtained under normal drift mode conditions by using trapping voltage modulation and phase sensitive detection. The marginal oscillator detector is based upon the design of Warnick, Anders and Sharp [29]. Ion cyclotron double resonance experiments were performed to identify precursor of ion/molecule reaction products with the use of a Wavetek Model 144 sweep generator as the secondary oscillator. The ICR cell is housed in a stainless steel vacuum system and is situated between the polecaps of a Varian 12 in. electromagnet (1.5 in. gap). The electromagnet is controlled by a Varian V-7800, 13 kW power supply and a Fieldial Mark I Magnetic Field Regulator. The instrument is pumped by a 4 in. diffusion pump with a liquid nitrogen cold trap and a Ultek 20 L/s ion pump. Samples are admitted from a triple inlet (individually pumped by a 2 in. diffusion pump and liquid nitrogen cold trap) by Varian 951-5106 precision leak valves. Approximate pressures are measured with a Veeco RGlOOO ionization gauge. Tricarbonylnitrosylcobalt(0) and 2-bromoethanol were obtained from Alfa Products. All allylic compounds -21- 2-chloroethanol, butanone, propanoic, butanoic, 2-butenoic and 2-chloropropionic acids were obtained from Chem Services, Inc. 3—Butenone was obtained from Columbia Organic Chemical Company. 3-Chlorobutanone, 2,2-dimethyl- propanoic and 2-bromoacetic acids were purchased from Aldrich Chemical Company. 3—Chloropropanoic acid was obtained from Crescent Chemical Company. Acetic acid was purchase from J.T. Baker Chemical Company. 3-Hydroxy- propanoic acid was purchased from Pfaltz and Bauer. Deuterated acetic (d4) and pr0panoic (d6) acids were obtained from MSD Isotopes. All other compounds were obtained from Fluka Chemical Corporation. In the reactions of the cobalt-containing ions with the unsubstituted carboxylic acids one product observed was CoHZO+ formed by the following reaction: Co+ + RCOOH ——> RCO-CO+ + -OH -—é> (ketene)Co+(HZO) 0+ -—€> CoH2 If the acid was not pure but contained any water the above reaction could not be distinguished from + + CO -+H20 ——€> COHZO In order to verify the purity of the unsubstituted acids, low pressure electron impact spectra were obtained on -22- the ICR and examined for any peaks corresponding to fragments of H20. None were observed for the unsubstituted acids except for 2,2-dimethylpropanoic acid. This acid was then dried and purified by vacuum sublimation. The low pressure EI mass spectra of the substituted carboxylic acids showed in most cases the presence of water. The first attempt at purifying the solid carboxylic acids was recrystallization form anhydrous ether. Samples were recrystallized at least three times from ether. When admitted to the ICR, the mass spectrum of ethyl ether was obtained. The solvent could not be removed from the sample under vacuum, so a dry technique of purification was used. 3-Bromopropanoic, 2—bromoacetic, 3-chloro— propanoic, 2-chloroacetic and 2—hydroxyacetic acids were purified by sublimation. The liquid acids, 2—mercaptoacetic, 2-mercapto— propanoic and 3-hydroxypropanoic acids also contained small amounts of water. Several different approaches were used in attempts to purify these compounds. First, pumping on the sample under vacuum (<100 mtorr) was employed in the hope that the higher vapor pressure components would be lost. This did not work. Next, distillation of 2-mercaptoacetic acid at atmospheric pressure did not accomplish any positive results. Vacuum distillations were attempted and these also were futile. More drastic means were then attempted for the purification of the mercapto-acids, P O5 and molecular 2 -23- sieves did not remove the water. Several attempts at continuous extractions were done by refluxing the acids in benzene and azetroping off the water with the benzene by using a Dean-Stark trap. After refluxing for at least three hours, the samples were then rotary evaporated or vacuum distilled to remove benzene. Both methods left traces of benzene in the sample. Similar reflux was also done using ether as the solvent. Since the mercapto acids seemed to prefer to retain the solvents used in the extraction, a final attempt at vacuum distillation was tried. The vacuumed distilled acids were analyzed by gas chromatography using a thermocouple detector and were found not to contain water. As a second check, a BI spectrum was obtained on] the department's Finnigan GC/MS, again no traces of water. When these compounds were analyzed on the ICR no traces of water were detected. 2-Hydroxypropanoic acid caused a new problem in an attempt to purify it. This acid readily reacts with a second mole of this acid to give off two moles of H20 when heated (even under vacuum). The methods used in purifying the mercapto acids were attempted here but were useless. To purify this acid, the sample was placed in a vacuum dessicator in which P205 was used. It was left undisturbed for 48 hours and then analyzed on the ICR. The EI spectrum indicated that the water was removed. -24- These purified samples and all other compounds were subjected to multiple freeze-pump-thaw cycles before use. In a typical experiment, low pressure electron impact (EI) spectra of the organic compound to be studied were recorded to determine purity. Higher pressure spectra were obtained and any ion/molecule reactions due to the organic system alone were identified and their precursors determined by ion cyclotron double resonance. Then spectra were obtained of 1:1 and 1:2 mixtures of Co(CO)3NO: organic for total pressures up to 1.5 X10-5 torr (as measured by an ionization gauge). Ion/molecule reaction products were identified and precursors of each determined by double resonance techniques. The observed reactions are facile and have a rate constant within an order of magnitude of the collision frequency. A branching ratio is the perCentage of an observed reaction is of the total for a given ion. In the Introduction a branching + 6 from 2-chlor0propane. In other words, 60% of all the ratio of 60% was given for the formation of CoC3H products of 2-chloropropane with Co+ is CoC3H6+. Branching ratios are accurate to within 10%. CHAPTER 3 THE CHEMISTRY OF COBALT-CONTAINING IONS WITH THIOLS Before discussing the chemistry of the cobalt and cobalt-containing ions with the different organic compounds, a brief discussion of the mass spectrum of Co(CO)3NO is necessary. At 70 ev, the parent ion, Co(CO)3NO+ (m/z==173), is abundant. Ions are also observed due to the loss of one, two or three carbonyls (Co(CO)2NO+ (m/z =145), CoCONO+ (m/z =117) and CoNO+ (m/z:=89» with the abundance decreasing with the loss of more CO's. Ions are also observed that do not contain a NO ligand (Co(CO)2+ (m/z =115), CoCO+ (m/z =87) and Co+ (m/z =59)). The bare metal ion is the most abundant ion in the 70 ev mass spectrum. At higher pressures of Co(CO)3NO, reactions of these ions with the neutral Co(CO)3NO are observed [30]. + CONO(CO)n -+CO(CO)3NO —-€> C02(NO)2(CO)n+2-+CO n -0-3 -—€> C02(NO)2(CO)n+l-+2CO n.=1-3 -25- -26- + CO(CO)m -+CO(CO)3NO -—€> C02(NO)(CO)m+2 +CO m-—O-2 ——€> C02(NO)(CO)m+l-+2CO Hl=0-2 In this study the reactions of these binuclear metal ions were not investigated. A. The Chemistry of Co+ with Thiols The primary site of attack by Co+ in thiols is apparently the polar C-SH bond. Reactions following Co+ insertion at this site account for a minimum of 90% of the products observed for all of the thiols studied except for 2-methyl-1-propanethiol. For all of the thiols, except 1-butanethiol and 2-butanethiol, the only products observed following a reaction sequence similar to Scheme I are CoHZS+ and Co(olefin)+ Co+-+RCH2CNZSH ——e> RCHZCHZ-Co+-+-SH -—+> + + (RCH—CH2)CO(HZS) -—€> COHZS -—€> Co(olefin)+ where retention of the olefin is the slightly favored pathway. This preference of Co+ for the olefin follows metal-ligand "affinities". Several reports in the past few years have pointed out the correlation between proton affinities and metal-ligand affinities [10,30-38]. In general, when two or more ligands compete for sites on a -27- metal, the ligand that has the lower protOn affinity is apparently the "poorer ligand" (lower bond strength to the metal ion) and is eliminated to a greater extent than the other ligands. (Relevant proton affinities are given in Appendix A.) 1. Ethanethiol In Table l are the reactions of each metal ion in the mixture Co(CO)3NO and ethanethiol along with the branching ratios. In the chemistry of Co+ with ethanol an exception to the above proton affinity/metal-ligand affinity trend occurs when H20 and CH2=CH2 compete as ligands in the (C2H4)Co+(H20) complex. In this case the metal prefers to retain the olefin which has a lower proton affinity than water. When Co+ reacts with ethanethiol, retention of the ethene is preferred rather than H25, although the latter has the higher proton affinity. In the Co+/ethanol system, retention of water constituted 13% of the reactions while in the Co+/ethanethiol system retention of H S 2 accounted for 43% of the products. This difference may be due to the higher proton affinity of the H28 which competes more effectively with the olefin than does water for a site on the metal. (Note, the percentages after each reaction are the branching ratios.) -28- Table 1. Reactions of the ions derived from Co(CO)3NO with ethanethiol (M==CH3CHZSH). + + Co + M ———4 CoC2H4 -+HZS (57%) + -——4 CoHZS +C2H4 (43%) + + CoCO + M ———) CoM +co (100%) + + Co(C0)2 + M -——) CoCOM +CO (100%) CoNO+ + M ——) No Reaction + + ' CoCONO + M ——-—) CoNOM +co (100%) + + Co(CO)2NO + M ————) CoNOM +2co (60%) ——-) CocONOM+ + co (40%) CO(CO)3NO+ + M ——) CoCONOM+ +2co (100%) -29- + + CO +CH3CHZSH ——-—> COC2H4 +st (57%) + The chemistry of Co+ with ethanethiol has been previously reported [6]. The product CoSH+, however, was not observed in our experiments. Also the branching ratios are different from those we obtained but the ordering of the reactions is the same, that is, retention of C2H4 by Co+ was favored over H23 retention. 2. Prgpanethiols In Tables 2 and 3 are the reactions of each metal ion with l-propanethiol and 2-propanethiol, respectively, along with the branching ratios. 3+, Besides the formation of Co(olefin)+ and CoH2 two other reactions are observed with l-propanethiol. Co+ +CH3(CH2)ZSH ——-> CoC3H6+ +825 (73%) —> CoHZS+ +C3H6 (18%) —> CoCH3SH+ +C2H4 ( 7%) —> CoCH cs+ +CH +H ( 2%) 2 4 2 The third reaction in this mixture involves Co+ insertion into the Cl-C2 bond, follwed by a B-H atom shift to form the Co(C2H4)(CH3SH)+ complex from which elimination of ethene occurs. In this reaction, retention of CH3SH is Table 2. -30- Reactions of the ions derived from Co(CO)3NO with l-propanethiol (M==CH3(CH2)2SH). + Co + M ———+ CoC3H ++H S Coco+ + M Co(CO)2+ + M CoNo+ + M CoCONO+ + M 11111 111 II II 1111 6 2 + CoHZS -+C3H6 + COCH3SH +C2H4 CoCH =C=S+-+CH4-+H 2 2 CoM+-+CO + COCOC3H6 -+HZS + CoC3H6 -+HZS-+CO + COCOCHBSH +C2H4 =C=S+-+CH -+H -+co COCH2 4 2 CoM+-+2CO CoCOM+-+CO + CONOHZS +C3H6 + CONOC3H6 -+H28 CoNOM+-+CO + COCONOC3H6 -+HZS + CONOC3H6 '+HZS-+CO _+ CONOCHz—C-S -+CH4-+H2‘+CO (73%) (18%) (7%) (2%) (10%) (3%) (75%) (5%) (7%) (30%) (70%) (66%) (34%) (25%) (4%) (5%) (66%) Table 2 Continues. Table 2 Continued. -3l- Co(CO)2NO+ + M -—-) ——) ———) Co(CO)3NO+ + M ———9 CoNOM+ + 2C0 CoCONOM+ + co SH++C H Co(CO)2NOCH2 2 5 CoCONOM+ + 2co (78%) (5%) (17%) (100%) -32- Table 3. Reactions of the ions derived from Co(CO)3NO with 2-propanethiol (M==(CH3)2CHSH). CO++M——- CCH++HS (59%) > 0 3 6 2 + —-) CoHZS +C3H6 (41%) + + CoCO + M -—-9 CoM -+CO (1%) —) CoC3H6+ +HZS +CO (72%) + O —-) CoCOC3H6 +HZS (6%) ———) CoCOHZS+ +C3H6 (12%) ——-) CoCH2=C=S+ +CH3CHO +112 (7%) + + Co(CO)2 + M ———+ CoCOM -+CO (60%) ———) 06M+ +2co (10%) ——) CoCOC3H6+ +823 +co (30%) + + CoNO + M ——-) CoNOHZS +C3H6 (77%) ———) CoNOC3H6+ +1123 (23%) + + CoCONO + M ———+ CoNOM -+CO (34%) ——-) COCONOC3H6+ +825 (10%) ——-) CoNOC3H6+ +st +co (10%) ———-) CoNOCH2=C=S+ +CH3CHO +H2 (46%) Table 3 Continues. Table 3 Continued. -33- Co(CO)2NO+ + M ———) ——) ——) Co(CO)3NO+-+M ———9 CoNOM+ + 2co CoCONOM+ + co SH+ +c H Co(CO)2NOCH2 2 5 CoCONOM+ + 2co (75%) (10%) (15%) (100%) -34- preferred over CH2=CH2, consistent with the greater proton affinity of metanethiol (therefore the greater metal- ligand affinity). The fourth reaction, leading to the formation of Co+-thioketene, presumably involves the insertion of Co+ into the C2-C3 bond, leading to either the intermediate (CH4)Co(C2H3SH)+, or Co(C2H3SH)+, in which CZHBSH rapidly loses H2. A mechanism is suggested in Scheme II. Transition metal ion-ketene complexes have previously been observed as reaction products [3,9,39]. The CoC2H35H+ intermediate was not observed. It is interesting to note that in some reactions involving multistep elimination processes, the intermediates are observed. For example, in the reaction of Co+ with 4-bromo-l-butanol [9] Ho ——--—l- C0(H0 )‘Mh Co(fl)+ both Co(CH2=CHCH2CH20H)+ and Co(CH2=CHCH=CH2)+ are observed as products. However, a reaction also occurs leading to -3 5.. N N 1w ZEm Iw... \I 1’ \o/ )qu w... .8/ +8 m: I- m x..- «I- _- ..... m - ..oO\I :. gm 1m 6.: @- ezoexol ..eolnxo HH mEmcum -36.. Br + C CO’ t W —__> /\/°\/\ CZHSBr HO- HO Br /” ..... Co+ OH H 3’ E)” 06’," c CH, the product Co(CH2=C=O)+. The Co(CH2=CHOH)+ intermediate is not observed, analogous to what is reported here for the formation of Co(CH2=C=S)+ from l-propanethiol. In 2-propanethiol only the products following Co+ insertion into the C-SH bond were observed. + + Co -+(CH3)2CHSH ——+ CoC3H6 -+HZS (59%) —.¢ CoHZS+ +c H 3 6 (41%) Here the metal—ligand bond strengths, as suggested by proton affinities, are consistent with the results. The ligand with the higher proton affinity is retained to a greater extent by the metal. 3. Butanethiols In Tables 4 through 7 are the reactions of each metal ion with the isomeric butanethiols along with the branching ratios. -37- Table 4. Reactions of the ions derived from Co(CO)3NO with 1-butanethiol (M =CH3(CH2)3SH) . + + Co + M ——-) CoC4H8 +HZS (17%) + ——-) Cons +C4H8 (5%) + ———-) CoC2H4 +CH3CH25H (5%) ——) c411; +COSH (11%) + ———-) CoC4H6 +H2 +H28 (38%) + ——) CoC4H6(HZS) +H2 (23%) + + Coco + M ———) CoM +CO (14%) -—) CoCOC4H8+ +825 (13%) ——) CoC4H8+ +823 +co (29%) —-) CoCOHZS+ +C4H8 (44%) + + Co(CO)2 + M ———) CoCOM +CO (76%) —-) COM+ + Zoo (16%) ——) 2C0C0C4H8+ +1125 +co (8%) CoNO+ + M ——) No Reaction CoCOIIO+ + M ——) CoNOM+ +co (100%) Co(CO)2NO+ + M —-) CoNOM+ +200 (100%) Co(CO)3NO+ + M —-) COCONOM+ +2co (100%) -38- Table 5. Reactions of the ions derived from Co(CO)3NO with 2-butanethiol (M=CH3CH2CHSHCH3) . + + Co + M ——) CoC4H8 +st (31%) + ——-—) CoHZS +C4H8 (6%) —-) CoC3H6+ +0113st (5%) ———) C4H9+ +COSH (10%) ——-) 00041161" +H2 +1125 (10%) + ——-) CoC4H6(HZS) +82 (38%) + + CoCO + M ——-) CoM +CO (10%) —-—-I CoCOC4H8+ +1125 (45%) —-—) CoC4H8+ +1425 +co (15%) ——-) CoCOH28+ +C4H8 (19%) ——) Coc3I16+ +CH3SH +co (11%) + + Co(CO)2 + M ———) . CoCOM +co (56%) ——-) CoM+ + Zoo (16%) ——) CoCOC4H8+ +828 +co (13%) + —-7 Co(CO)2C4H8 +st (15%) CoNO+ + M ———4 No Reaction + + CoCONO + M —-—-) CoNOM +CO (100%) Co(CO)2NO+ + M —-) CoNOM+ +2co (100%) Co(CO)3NO+ + M -——) COCONOM+ +2co (100%) -39- Table 6. Reactions of the ions derived from Co(CO)3NO with 2 methyl-l-propanethiol (M = (CH3) 2CHCHZSH) . Co+ + M ——) CoC4H8+ +323 (15%) + ————) COHZS +C4H8 (60%) + ————) CoCH3SH -+C3H6 (24%) + -———) CoC3H6 -+CH3SH (1%) + + CoCO + M ————) CoM -+CO (19%) + --—) CoCOC4H8 -+H28 (15%) + -——) CoC4H8 -+st-+CO (62%) + -———) CoC3H6 -+CH3SH-+CO (4%) + + Co(CO)2 + M ————) CoCOM -+CO (100%) + . CoNO + M ————) No React1on + + CoCONO + M ————) CoNOM -+CO (100%) + + Co(CO)2NO + M ————) CoNOM -+2CO (100%) Co(CO)3NO+ + M —) CoCONOM+ +2co (100%) -40- Table 7. Reactions of the ions derived from Co(CO)3NO with 2-methyl-2-propanethiol (M==(CH3)3CSH). + + Co + M —-) CoC4H8 +1128 (59%) + + + CoCO ——) CoM +CO (16%) ———) CoCOC4H8+ +HZS (30%) —-) CoC4H8+ +1125 +co (21%) ——) CoCOHZS+ +C4H8 (33%) + + Co(CO) 2 ——) CoCOM +CO (51%) —-) CoCOC4H8+ +st +Co (49%) CoNO+ ———) No Reaction + + CoCONO ——-) CONOM +Co (32%) ——-) CoNOC4H8+ +st +Co (68%) + + CO(CO)2NO -——9 CONOM -+2CO (89%) ——-9 CoCONOM+ + CO (11%) C0 (C0) 3Mo+ ——) CoCOMOM+ + 2C0 (100%) -41- In 1-butanethiol, the C-SH bond is still the primary site of insertion by Co+, but several new products are observed. First, Co+ reacts by hydrogen sulfide anion abstraction to give the butyl cation in this compound. This alkyl cation formation process has been observed in the chemistry of Co+ with numerous alkyl halides [5,6] and 2—propanol [5]. In nitroalkanes [6], RNO2 abstraction of N02- increases as the alkyl group increases in size- 0% for R=CH and R=C H 12% for R=n-C3H7, 38% for 3 2 5' R=n-C H and up to 79% for R=t—C4H9 -since, as the size 4 9 of the alkyl group increases, its ionization potential decreases, energetically facilitating alkyl cation formation. In Scheme III, the reactions of 1-butanethiol with Co+ are shown. For l-butanethiol, after the initial insertion into the C-SH bond and the B-H atom shift, structure I is obtained. From this intermediate, products observed for all thiols, CoH 3+ and Co(butene)+ are 2 formed. Co++-CHSH ———-CCH++HS (5%) n 4 9 > O 4 8 2 + ——> CoHZS +C4H8 (17%) + ——> CoC2H4 +CH3CH28H ( 5%) ——> n-C4H9+ +CoSH (11%) + ——> Co(HZS)(C4H6) +H2 (23%) + -€> Co(C4H6) 9+H23+H28 (39%) -42- Co+ can also insert into the allylic C-H bond of structure I to give structure II followed by a B-H shift to the metal to give an intermediate III Co(butadiene)(H28)(H2)+, from which elimination of H2 or H2 and H28 can occur. This insertion process has been previously suggested for the reaction of Co+ with l-butene and 2-butene [9]. Stucture III is one of the three possible pathways through which 1-butanethiol may eliminate H Using the 2. approach which Tsarbopoulos and Allison used to explain H2 elimination from l-butanol [10], the following three structures as products following H2 elimination from l-butanethiols are feasible. Armentrout and Beauchamp have reported that Co+ inserts into the C2-C3 bond of 1-butene to eliminate ethene [9]. This could also occur in 1-butanethiol. After the formation of structure I, the Co+ inserts into the C -C bond of I followed by a B-H shift to the metal 2 3 and rearrangement into structure VI. From structure VI, H28 elimination occurs to give Co(C2H4)2+, which has the -42a- 4 5.5. fl Im .IU/ .00 1\ ...:m z. a H H” In .3 rm rm :m I/ I\ Y /J .00 fl 900 IV .00 .ru\ \F .....m firm NH 1m :5 ..30/\ L .. . rm 1 =5 / x . rm .ou )1! \r\ foul) k In .3 H \\ // H a ...-53 "115-3 / a \r \ cI-U .mux. ., /rm \ _\ .3 I: 2 K (wk \ a h .mfou \ r \ :ln 3 rm . l \ a II o Lm I: II .00 Lot-.330 tw Izggs 1% ooU IIQ xm\ << /\< ..OU B Orr-0C Um >\ -43- same empirical formula as Co(Butene)+. Therefore, the first reaction in this mixture has two possible structures for the ionic product. More importantly, the elimination of H25 and C2H4 giving CoC2H4+ cannot be explained through competitive elimination just as it was for ethanethiol. In the reactions of 2-butanethiol with Co+, abstraction of SH- giving the butyl cation is observed, as well as, the formation of CoH S+ and Co(butene)+. 2 Co+ +s-c H SH —- c C H ++H s (31%) 4 9 > O 4 8 2 + -—> CoHZS +C4H8 ( 6%) + —> CoC3H6 +CH3SH ( 5%) — C H + +C SH (10%) > 4 9 o + —> Co(HZS) (C4H6) +II2 (10%) + ——e> Co(C4H6) ~+H2«+HZS (38%) The chemistry of Co+ with 2-butanethiol is also shown in Scheme III. From structure VIII, Co(2-butene)(HZS)+, insertion into the C-H bond is possible giving structure II from which H2 or H2 and H S elimination may occur as 2 in 1-butanethiol. As in explaining H2 elimination from l-butanethiol, several structures are possible for H2 elimination from 2-butanethiol. -44- .\ 'C.o+ ”SK/ The elimination of methanethiol from 2-butanethiol could be explained through isomerization of 2-butanethiol to 2-methy1-l-propanethiol (isobutyl thiol). Co+ inserts into the Cl-C2 bond of 2-butanethiol to give structure IX, which is then followed by a B-H atom shift to the metal (structure X), then rearrangement to Co+(iso-butyl thiol), structure XI. From this intermediate Co+ then inserts into the Cl-C2 bond and a B-H atom shift to give Co(C3H6)(CH38H)+. The proton affinities of the two ligands are close to each other, and as in the case of ethanethiol when the proton affinities are close enough it appears that the olefin is preferred even though it has the lower proton affinity. This may be due to a stronger interaction of Co+ with the h-electrons in the olefin then with the lone pair on the hetero atom. Isomerization of 2-substituted butyl compounds has been suggested in the chemistry of Co+ with amines [7] and of TiCl + with 2—butanone [40]. 3 -45- In the chemistry of Co+ with 2-methy1-1-propanethiol 75% of the reactions are derived from the Co(butene)(HZS)+ intermediate. Co+ +i-C4H98H —--> CoC4H8+ +st (60%) -—> CoHZS+ +C4H8 (15%) —> CoC3H6+ +CH3SH ( 1%) -———> CoCH3SH+ +C3H6 (24%) The remaining 25% of the products are due to Co+ insertion into C-C bond of 2-methyl-1-propanethiol. In 2-methyl-2-propanethiol, insertion into the C-SH bond followed by the B-H shift and then competitive ligand loss is the only process observed. The two products observed reflect the metal ligand affinities as expected. Co+ +t-C H SH —> CoC H + +H 8 (59%) 4 9 4 8 2 + —-> COHZS +C4H (41%) 8 What is surprising is that the tert-butyl cation is not formed (in 2-methyl-2-nitropr0pane the formation of C4H9+ accounted for 79% of the reactions [6]). In summary, the reactions of Co+ with thiols are dominated by insertion into the C-SH bond. In the straight chain butanethiols, the elimination of H28 and H2 -46- forming a Co+(butadiene) complex also occurs. Insertion of Co+ into a C-C bond is observed to be a minor pathway in thiols. B. The Chemistry of Co(CO)X+ and Co(CO)yNO+ (x =1,2; y==0-3) Ions with Thiols. Ligand Effects. As mentioned earlier, in the electron impact mass spectrum of Co(CO)3NO not only is the bare cobalt ion formed, but other ions containing CO and NO ligands. These ions also react with neutral organic compounds in a way similar to Co+. Looking at the chemistry of CoCO+ and CoCONO+ with alkyl halides and alcohols we see the similarity with Co+ [5]: CoCO+ +i-C3H7C1 -—> CoCOC3H6+ +HC1 (16%) -—> CoC3H6+ +HC1 +co (67%) ——> CoCOHCl+ +C3H6 (17%) CoCONO+ +i-C3H7Cl ——> CoNOC3H6+ +HC1 +co (53%) —> CoMO(C3H7C1)+ +CO (47%) CoCO+ +i-C3H7OH ——-> Cc>Coc3H6+ +HZO (21%) ——> CoC3H6+ +1120 +CO (21%) ———> Co(C3H7OH)+ +co (58%) CoCONO+ +i-C3H7OH -—> CoNOC3H6+ +H20 +Co ( 5%) —> CoNO(C3H7OH)+ +CO (95%) -47- The last reaction in CoCONO+ with 2-chloropropane and in CoCO+ and CoCONO+ with 2-propanol represent ligand substitution reactions. The reactions observed for CZ-C4 thiols are typical for metal-containing ions, MLn+, in their chemistry with monofunctional organic compounds in the following ways: (1) The bare metal ion reacts to form the largest variety of products [6]. (2) The organic rearrangement processes observed for CoCO+ are similar to those observed for Co+ [4-6,10,11,41]. For example, Co+ and CoCO+ show similar reactions with l—propanethiol (M=CH3(CH2)ZSH): + + Co +M —> CoC3H6 +1125 (73%) + ——€> CoHZS +C3H6 (18%) + ———> CoCH3SH +c2H4 ( 7%) _ _ + —-> CoCHZ-C—S +CH4 +H2 ( 2%) + + CoCO -+M ——€> CoM -+CO (10%) —>CCOCH++HS (3%) O 3 6 2 + ——€> CoC3H6 -+HZS-+CO (75%) + ——€> CoCOCHBSH «+C2H4 ( 5%) —> CoCH2=C=S+ +CH4 +112 +co ( 7%) (3) In contrast with the carbonyl ligand, NO ligand deactivates the metal center [4-6,10,11,41]. (Note; only with the propanethiols did CoNO+ react.) (4) AS the -43- number of ligands present on the metal ion increases what appears to be ligand substitution reactions predominate [6]. In the case of Co(CO)2NO+ reacting with ethanethiol Co(CO)2NO+-+EtSH —e> Co(CO)2NO(EtSH)+* —e> Co(CO)NOEtSH+-+CO there are no empty orbitals available on the activated complex which can be used to rearrange the ethanethiol. Therefore one can assume that the ethanethiol is still intact in the product ion, i.e., a true ligand substitution. In the case of CoCO+, is the loss of COIoccurring from CoCO(C2HSSH)+ or from CoCO(C2H4)(HZS)+? The proton affinities of the different ligands indicate that CO should be lost from either case. (The proton affinity of CZHSSH was not available but from the trend observed for H S, CH SH, H20 and CH OH indicate that the PA(C2HSSH) 2 3 3 should be greater than PA(C2HSOH).) There is no conclusive evidence for just a substitution reaction occurring or if a rearrangement also occurs. For simplicity we will assume that whenever a neutral loss of 28 u occurs, CO only is lost, and the organic molecule is intact on the product ion. One should remember, however, the probability of the molecule remaining intact is reduced for reactions of CoCO+, Co(C0)2+ and CoCONO+ due to the availability of empty orbitals on the metal. In 2-propanethiol,the reactions CoCO+ are similar to + . . . . Co . There 18 the observance of a new react1on in this mixture based upon insertion of CoCO+ into the C-C bond: -49- Coco+ + (CH3) 2CHSH —-—> Co(CH2=C=S)+ +C2H6O This reaction could occur by metal insertion into the C-C bond, followed by a B-H atom shift and the elimination of CH4 to give the intermediate Co(C2H3SH)+, which then undergoes insertion of Co+ into a C-H bond and B-H atom shift to yield the Co(thioketene)+ (similar to scheme VI). However, this reaction is not seen for Co+ but only for CoCO— and CoCONO+. This implies that the carbonyl actively participates in the chemistry via CoCO+ insertion into the C-C bond. O I -H n CoCO++ >- SH —> CH3 -C-Co*-CIZH-CH, éfir’ CH3 -C-Co’---- [k 5" H SH %H2 3 H 'CHSCHO ..... f 91' (___. __,., L ... fi """ Co Shwt l/JL\ ‘ Co ”\\ HS C0: H SH As seen from the above mechanism, the neutrals lost are CH3CHO, and H2 not CH4, H2 and CO as in l-prOpanethiol. . . . . + S1nce thls reactlon occurs w1th CoCONO+ and not CoNO further evidence for CoCO+ insertion is provided. -50- The insertion of CoCO+ into the C-C bond is not unprecedented. One example of CoCO+ insertion has been reported by Radecki and Allison [7]. They have shown that CoCO+ inserts into the C-C bond of ethyl amine to give CoCH NH + as the product ion [7]. Coco+ +CH CH NH 3 2 3 2 2 + + ——€> CH3CO-CO '-CH2NH2 -—€> (CHz—C—O)CO (CH3NH2) ———> CoCH3NH2+ . In the chemistry of Co(CO)2NO+ with the two propanethiols there is a reaction corresponding to the loss of m/z =29 from the [Co(CO)2NO(M)+]* complex. Most likely this product also involves a CoCO+ insertion. In this case insertion into the S-H bond where the CO prefers to bond to the hydrogen: Co(CO)2NO+-+C3H7SH ——> C3}! S-Co(NO)CO+ -COH ——> C3H7S -CoCONO+ +COH. 7 The chemistry of CoNO+ with organic compounds in general, unique. Either this metal ion does not react at all 95 it produces reactions that are contradictory to known trends. The "deactivation" may be due to the way the NO ligand bonds to Co+. NO may be a one or three electron donor, and if it behaves as a three-electron donor in CoNO+, the proper orbitals needed for rearrangements are not present. In the chemistry of CoNO+ with thiols, the only reactions are observed in the two propanethiols. Both give the same products: CoNO+-+C3H7SH ——€> CoNOHZS+ + C3H6 ——€> CoNOC3H6+.+HZS. From the proton affinity/metal— ligand affinities trends thus far observed, one would -51- expect the olefin to be retained more by the metal but it is the H28 which is retained more. There are two possible explanations for this phenomenon. First, the ionization potential of the CO ligand is 14.0 eV compared to 7.87 eV for Co, therefore CoCO+ has the charge retained by the metal, that is, Co+-CO. However, the ionization potential of the NO is only 9.25 eV, so CoNO+ may almost look more like Co—NO+ than Co+-NO. Therefore the relative ligand affinities may not be the same because HZS-Co is formed not HZS-Co+. (Presently there are not rules for relative ligand affinities for Co.) The second possible explanation is that instead of Co+ insertion into the R-SH bond, CoNO+ insertion into the bond occurs. This is analogous to CoCO+ insertion processes discussed above. This would then yield product ions of the structure: I (2"? n N Co’-N-SH '1 II I CH: c H l 2’ \ on, CH, CH, Metal-NO insertions occurring in condensed phases have been discussed in a review by McCleverty [42]. CHAPTER 4 THE CHEMISTRY OF COBALT-CONTAINING IONS WITH BUTANONE AND 3-BUTENONE A. The Chemistry of Co+ with Butanone and 3-Butenone In the past six years three reports of the gas phase chemistry of transition metal ions with ketones have appeared [39,40,44]. Allison and Ridge [40] have reported the reactions of ions derived from electron impact on TiCl4 with ketones. Ti+ and TiCl+were observed to reduce aldehydes and small ketones (CS or smaller). H (x=0,1) ——e> Ti(C1)XO+-+C4 8 Ti(C1)X+-+CH COCHZCH 3 3 TiC12+ is nonreactive in this mixture, while TiCl3+ shows extensive reactions with butanone [40]. Freiser et al. [39] reported the gas phase chemistry of Fe+ with aldehydes and ketones. In a,a'-d5-butanone, 70% of the reactions involve Fe+ insertion into either C-CO bond followed by a B-H atom shift or an alkyl shift. The remaining 30% of the reactions involve Fe+~induced dehydrogenation of butanone. -52- -53- Fe+ +CD3COCD2CH3 ——> FeCO+ +C3H3D5 (10%) ——e> FeCDZCH2+-+CD3CHO ( 7%) —> FeCD3CHO+ +CD2CH2 (38%) ——> Fe(CO)CD2CH2+ +CD3H (15%) -—€> FeC4D4H20+-+HD (30%) In light of newly discovered mechanisms for H2 elimination, HD elimination may occur either through Fe+ insertion into a C-D bond followed by a B-H atom shift and the elimination of HD yielding an Fe(3-butenone)+ complex, (analogous to a metal(butadiene)+ complex), gr, insertion of Fe+ into the CO-CDZCH bond followed by a 3 1,4 elimination of HD to give the product: 0. O T + II: N - Ie tCDSC-CDZCHa ———>CO,c-Fe+-COZCH3-—H9+ 002-4090 Fe" 02C =CH2 Fe+ also reacts similarly with 3-butenone. + _ + 9 Fe +CH3COCH—CH2 ———> FeCZHZ +CH3CHO (28.) ——> FeC3H6+ +co (36%) + ———> Fe(CO)C2H2 +CH4 (36%) -54- All of the products in these two Fe+-ketone mixtures follow the proton affinity/metal-ligand affinity trends. Beauchamp et a1. [44] have reported similar reactions of Co+ with butanone, in an atomic ion beam experiment, with a kinetic energy of the reactant ion of ~0.5 eV. Co+ +CD3COCD2CH3 ———> Coco+ +C3H3D5 ( 8%) —> 'CoCOZCHZJ' +CD3CHO (12%) —> CoCD3CHO+ +CD2CH2 (61%) —> CoCO(CH2CD2)+ +CD3H (19%) ——-> CoCD3COCD=CH2+ +HD trace 1. Butanone In Table 8 are the reactions of each metal ion with butanone as well as the branching ratios. The initial interaction of Co+ with butanone is most likely at the oxygen atom, to give an activated species I‘ 7* -55- Table 8. -Reactions of the ions of Co(CO)3NO with butanone (M==CH3COCH2CH3). Co+ + M ——) Co(CO)C2H4+ +CH4 CoCH3CH=C=o+ +CH4 + ———4 COCH3CHO +C2H4 ———) CoC2H4+ + CH3CHO + CoCO +C3H8 Coco+ + M CoM+ +Co II + COCH3CHO -+C2H4-+CO Co(CO)2+ + M CoM+ +2co CoCOM+ + CO II + . CoNO + M No Reaction I CoCONO+ + M CoNOM+ +CO + COCONOCH3CO +C2H5 II + Co(CO) 2No + M CoNOM+ +2CO + COCONOCH3CO -+C2HS-+CO II + Co(CO) 3190 + M CoCONOM+ + 2C0 I (3%) (76%) (21%) (90%) (10%) (16%) (84%) (80%) (20%) (88%) (12%) (100%) -55- which can follow several reaction pathways. The metal ion appears to insert into either C-CO bond in this compound (Scheme IV). The mechanism suggested in Scheme IV is modeled after that proposed by Freiser et a1. [39]. + + CO +CH3COCH2CH3 -—€> COCO(C2H4) +CH4 ( 3%) + ——€> COCHBCHO +C2H4 (76%) + ——€> COC H -+CH CH0 2 4 3 1 (21%) COCO '+C3H8 When insertion occurs between the CZHS—CO bond, a B-H shift leads to the complex Co(C2H4)(CH3CHO)+. From this structure, the ligands undergo competitive elimination to yield the observed products. The proton affinity of the aldehyde is greater than the olefin, therefore the aldehyde should be retained more by the metal ion, as is observed. When Co+ inserts into the CH -CO bond, an alkyl shift 3 has been suggested to give structure XIV. A B-H atom shift followed by the elimination of the poorer ligand, CH4, gives CoC3H4O+, which would be in the form Co(CO)C2H4+. Elimination of methane can also occur if, after insertion into the CH3-CO bond, a B-H shift occurs followed by elimination of CH4 giving Co(CH3CH=C=O)+. From structure XIV it is possible to eliminate C3H8 to give CoCO+ (m/z =87). The product CoC2H4+ has the same nominal mass to charge ratio as CoCO+. -57- +88 "I ~oeo Lieu-888 «QNUI 00I nIo I Io; TIO- “.5- / \ w _ . canto aIoII.h Io , "I68 .oIonIoao :2 n O I In . I .ew Io .oIonIoH Hf"? IUI~ 00 +0000 _ n I I0 -I u /_\ I o n I0 _ .o. T.- tEm I a . I o- e __.-.oo-w-nIo own-tolnIo I0 0 =2 0 EN tam It .IJ ”Io ”NH-Mm It a “NEH » cam / \ 22m 4 a A I o-o-eoIIo T .8 I I 386qu .. + :56 - .56 _. 0 00 0 - I NH oEocom -58- 2. 3-Butenone Table 9 are the reactions of each metal ion with 3-butenone along with their branching ratios. 3-Butenone reacts with Co+ similarly to the chemistry observed for butanone: + + - Co +CH3COCH-CH2 ———> CoCO(C2H2) +CH4 (45%) 4. ——> CoCH3CHO +C2H2 (12%) + ———> CoCZHZ +CH3CHO (25%) ——> CoCH3CH=CH2+ +Co (18%) Co+ also eliminates CO from 3-butenone giving CoC3H6+ which could be either Co(prOpene)+ or CH3-Co+-C2H3. Since the proton affinity of propene is greater than that of CO, either form would be possible. When the analogous pair of ligands are formed, in the case of butanone, CO and propane, the ligand with the lower proton affinity is also lost giving rise to Co(CO)+ (m/z =87). B. The Reactions of CoLn+ with Butanone and Butenone The reactions observed for these two ketones are . . . . + . . . typical for metal conta1n1ng lons, MLn , in their chemlstry with monofunctional compounds (although butenone is a bifunctional compound, the observed chemistry is that of the ketone group): (1) Co+ reacts to form the largest -59- Table 9. Reactions of the ions of Co(CO)3NO with 3- butenone (M==CH3COCH=CH2). Co+ + M —-) Co(CO)C2H2+ +CH4 (45%) ——+ CoCH3CHO+ +C2H2 (12%) —-) CoC2H2+ +CH3CHO (26%) —-) CoC3H6+ +CO (18%) + + CoCO + M ——) CoM +Co (87%) ——) CoC3H6+ + 2co (3%) ——) CoCOC2H2+ + CHBCHO/ (1%) + CoCOC2H2 +CH4 +CO ———) CoCOCH3co+ +C2H3 (7%) —-) CoCOCII3CHo+ +C2H2 (3%) + + Co(CO)2 + M ———a CoM -+2CO (64%) ——-) CoCOM+ + co (3%) ———) CoCOCH3C0+ +C2H3 +co (33%) CoNO+ + M -———) No Reaction + + CoCONO + M ————) CoNOM I+CO (100%) Co(CO)2No+ + M ———) CoNOM+ +2CO (100%) + + Co(CO)3NO + M.-—-) CoCONOM -+2CO (88%) ——-) CoNOM+ + 3C0 (12%) -50- variety of products and (2) as the number of ligands present on the metal ion increases, ligand substitution reactions dominate. The reactions of Co(CO)2+ with butanone show two ligand substitution reactions. This ion complexes with butanone to give [Co(CO)2CH3COCH CH3+]*. If this complex 2 then rearranges itself, via reactions similar to Co+, to Co(CO)2(CH3CHO)(C2H4)+ or Co(CO)2(CH4)(CH3CH=C=O)+, the ligand with the lowest proton affinity/metal-ligand affinity should be lost. In the first the CO's have the lowest PA therefore either one or two CO's are lost. In the latter complex CH4 has the the lowest proton affinity. Therefore one would expect this ligand to be lost. No product corresponding to the loss of CH4 has been observed in the chemistry of Co(CO)2+. Since the CO has the lowest proton affinity of the remaining possible ligands, it is not possible to determine if the product ion contains the intact organic molecule or if it has undergone any rearrangements. As in the case of thiols, for simplicity, one should consider the products losing one or two CO's primarily as ligand substitution. The reactions of the bare Co+ ion with butanone require three or four empty orbitals on the metal (see Scheme IV). In CoCO+ only three orbitals are available. Therefore, the elimination of methane via pathway I, which requires four empty orbitals, cannot occur. In order to -61- release the excess internal energy of the CoCO+—butanone complex either elimination of CO (ligand substitution reaction) or insertion into the weakest bonds of the ketone occurs. The former process accounts for 90% of the reactions of CoCO+. Insertion into the CO-CZH5 bond (D(CH3CO‘-C2H4) =78.9 kcal/mole) is favored over other C-C bonds, and only the metal center inserts into the C-C bond. That is, the CO remains a spectator in this reaction. After insertion a B-H atom shift and competitive ligand loss occurs to give CoCH3CHO+. Co(CO)2+ undergoes ligand substitution reactions with butanone only (loss of one or two CO's). As expected, CoNO+ does not react. The NO ligand as previously discussed deactivates the metal center and tends to react with few compounds. CoCONO+ and Co(CO)2NO+ react to give the same two products, CoNOM+ and CoCONOCH3CO+. CoNOM+ is the result of displacement of one or two CO respectively from CoCONO+ and Co(CO)2NO+, while the loss of CZHS is via Co+ insertion -62- into the weaker C-CO bond. After insertion no orbitals are available for a B-H atom shift. Therefore a bond cleavage occurs. This suggests (D(Co+(NO)(CO)-CH CO‘>D(Co+(NO)(CO) 3 -C2HS). Co(CO)3NO+ reacts with butanone via ligand substitu- tion with loss of two carbonyls. CoCO+ shows a richer chemistry with butenone than it did with butanone. Ligand substitution still accounts for 87% of the total reactions but more types of rearrangements and bond cleavage processes occur. The reactions + - + * COCO +CH3COCH-CH2 -—€> COCOCZH2 -+CH3CHO + ——€> COCOCH3CHO +C2H2 ( 3%) + -—€> COCOCH3CO -+C2H3° ( 7%) involve Co+ insertion into the Co-C2H3 bond followed either by a B-H atom shift/competitive elimination process 95 by bond cleavage. CZHZ and CHBCHO compete better than C2H4 and CH3CHO as noted by the elimination of CH3CHO. The other reactions observed between CoCO+ and butenone involve the loss of the CO ligand first to give + . . . Co -butenone complex then 1nsertlon into the CH3-CO bond, followed by an alkyl shift and elimination of methane or the elimination of CO. * The product CoCOC2H2+ is formed by two different pathways; the amount from each could not be determined. -63- ++CH +co * Coco+ + CH COCH=CH 2 4 3 2 -—€> CoCO(C2H —> CoC3H6+ +CO +Co ( 3%) Co(CO)2+ reacts with butenone either by ligand substitution or by insertion/bond cleavage. CoNO+ is nonreactive while Co(CO)XNO+ ions displace one or more CO's. In the chemistry of Co(CO)NO+andCo(CO)2NO+ with butanone there was insertion into the Co-CZH5 bond followed by bond cleavage with the loss of CZHS' The analogous reaction (loss of C2H3) is not observed in 3-butenone primarily because the C0-C2H3 bond is approxi- mately 9 kcal/mole stronger than the CO-CZHS bond in butanone. A second possibility for the lack of C2H3 elimination from 3-butenone is that the metal-containing ion may complex strongly with the conjugated system and may prefer to lose CO rather than insert into a C-C bond. Why is CH =CHCO lost in the chemistry of Co(CO)2+ 2 with 3-butenone and not CH3CH2CO in the chemistry of the same ion with butanone? The CH3-CO bond is slightly weaker in the enone, this may make this bond more accessible to attack. * The product CoCOC2H2+ is formed by two different pathways; the amount from each could not be determined. CHAPTER 5 THE CHEMISTRY OF COBALT CONTAINING IONS WITH UNSUBSTITUTED CARBOXYLIC ACIDS A. The Chemistry of Co+ with Unsubstituted Carboxylic Acids In general, the chemistry of Co+ with carboxylic acids resembles its chemistry with ketones. That is, the initial sites of attack appear to be bonds on either side of the CO group, which is then followed by either a 8-H atom shift or an alkyl shift to the metal. 1. Formic Acid In Table 10 are the reactions of each metal ion with formic acid along with the branching ratios. Formic acid, HCOOH, reacts with Co+ to lose CO. This decarbonylation Co+ +HCOOH ——> Cono+ +co (100%) . . + . . . reaction involves the Co insertion into the C-OH bond followed by an d—H shift to the metal. Then the weaker ligand is eliminated from the Co(CO)H20+ complex. In the -64- -65- Table 10. Reactions of the ions of Co(CO)3NO with formic acid (M==HCOOH). Co+ CoCO+ + Co(CO)2 CoNO+ COCONO+ Co(CO)2No+ Co(CO)3NO+ + M ———4 ——) —-) II II CoH20+-+CO CoM+-+CO CoH20+-+2CO + CoCOM -+Co No Reaction CoNOM+-+CO CONOM+-+2CO CoCONOM+-+CO CoCONOM+-+2CO Co(CO)2NOM+-+CO (100%) (87%) (13%) (100%) (100%) (78%) (22%) (50%) (50%) -66- ion/molecule reactions of metal ions with organic molecules after the initial insertion of the metal into a bond, a B-H atom shift is usually preferred over other processes (alkyl shifts, a-H shifts, bond cleavages, etc.). However other processes can occur. For example, Co+ reacts with methane [8] via insertion into a C-H bond followed by an a-H atom shift with the elimination of H2 yielding CoCH2+. In ketones,alkyl shifts are frequently observed [39-41]. So in the case of formic acid the a-H shift was not unexpected. Formation of CO and H20 from formic acid by Co+ probably does not involve metal insertion into the C-H bond, since an OH shift would have to follow. No evidence for a hydroxyl group shifts has been reported to date. Therefore, one can conclude that insertion into the C-H bond does not occur for formic acid although this is the weakest bond in the molecule. This is due to multiple interactions of the Co+ with the electron rich oxygen atoms as shown in Scheme V. 2. Acetic Acid In Table 11 are the reactions of each metal ion with this acid along with the branching ratios. The chemistry of Co+ with acetic acid is richer than with formic acid. Besides the decarbonylation reaction, several new reactions are observed (see scheme VI). Following metal insertion into the C-OH bond an alkyl -57- .-o~I6o I I0... H mEmcom -57- Lori-8 T__ :0 J05 1.3 oo -:zm +I‘Il Ibo I0 +3 H mEmcom -68- Table 11. Reactions of the ions of Co(CO)3NO with acetic acid (M =CH3COOH) . Co + M —-) CoCH3OH+ +H20 (13%) ——-) CoHZO+ +CH2=C=O (58%) ———) CoCH2=c=o+ +H20 (21%) + —-) CoCH202 +CH2 (8%) + + CoCO + M ——-) CoM +CO (72%) —-) CoCOH20+ +CH2=C=O (6%) ——-) Co(CO) (HCOOH)+ +CH2 (11%) ————) CoCII2=C=o+ +H20 +co (6%) ——) CoCOCII2=C=o+ +HZO (5%) + + Co(CO)2 + M ——) CoM +2CO (5%) ——-—) CoCOM+ +co (95%) + + CoNO + M ———) CoNOHZO +CH2=C=O (100%) + + CoCONO + M ———) CoNOM +co (92%) ———) CoNOCH2=C=O+ +H20 +co (8%) Co(CO)2NO+ + M —-) CoNOM+ +2co (78%) ——) CoCONOM+ +co (22%) + + Co(CO)3NO + M —-———) CoCONOM +2CO (75%) ———) Co(CO)2NOM+ +CO (25%) -69- I - N “tr-m n 450001-00 IMIII. IOI0I.00n I0 I56 IO-0-.o0I I0 (LI 3. i- a ~ 0 I00 A , + o.ou~Io- I0/ \I .00. :Em Snouanloo ouowfo 4 of- IA- n I0 IO. IO n o - :Em » m .. n / n +AIO I00 0 AIIOIUIII IOI+0M III-NxI-nlvl IOIoOUIH-NI I0 4| .0-I I0 1."). WI I0++00 O0 0 +0020 0 H mEmcom -70- shift can occur to give CoCH3OH+, or a B-H shift can occur giving the Co(HZO)(ketene)+ intermediate. From this intermediate Co+ can retain either ligand, in this case preferring to retain the water. Also Co+ can insert into the C-C bond of the ketene to eliminate CH : 2 Co+ +CH3COOH ——> CoCH30H+ +CO (13%) —> CoH20+ +CH2CO (58%) —> CoCHZCO+ +H20 (21%) --—> CoHCOOH+ +CH ( 8%) 2: Co+ can also insert into the CH3-CO bond, followed by an a-H shift to give intermediate and then eliminate CH2: to give Co(HCOOH)+. While loss of CH2: to be an unlikely process, it is a minor product. Also, may appear CH2: elimination in such reactions is not unprecedented. Freiser, Cassady, McElvany and Allison [6] reported the elimination of CH2: from small nitroalkanes by Co+ to be a minor pathway (<3%). Also, Babinec and Allison reported the elimination of CH : accounted for 4% of the reactions 2 of Fe+ with l-aminopropane [45]. 3. Propanoic Acid In Table 12 are the reactions of each metal ion with propanoic acid along with the branching ratios. -71- Table 12. Reactions of the ions of Co(CO)3NO with propanoic acid (M==CH3CH2COOH). + + Co + M ——) Coco +C2H4+HZO (25%) ——) Cono+ +CH3CH=C=O (40%) ——) CoCH3CH=C=o+ +H20 (6%) —-) CoHCOOH+ +C2H4 (19%) + Co(CO)HZO +C2H4 -—-—) CoC2H4++HCOOH (11%) + CoC2H4 +1120 +CO + + CoCO + M ——) CoM +co (80%) ——-I CoCH3CHZOH+ +2co , (6%) + CoC2H4(HZO) +2CO ——) CoCOHZO+ +CH3CH=C=O (5%) CoHCOOH+ +c2H4 +co ——-—) CoH20+ +CH3CH=C=O +co (9%) + + Co(CO)2 + M ———-) CoCOM +CO (88%) ———-I CoM+ +2co (5%) ——) CoCOHZO+ + CH3CH=C=O +Co (7%) CoNO+ + M ——) CoNOH20+ +CH3CH=C=O (100%) + + CoCONO + M ———) CoNOM +Co (85%) ——-) CoNOH20+ +CH3CH=C=O +Co (15%) Table 12 Continues. Table 12 Continued. -72- Co(CO)2NO+ + M Co(CO)3NO+ + M CoNOM+-+2CO CoNOHZO+-+CH3CH=C=O + CoCONOM -+2CO Co(CO)2NOM+-+CO (60%) (40%) (89%) (11%) -73- Many of the reactions of propanoic acid (see scheme VII) with Co+ are similar to those observed with acetic acid. One noticeable exception is that Co(CH3CHZOH)+, i.e. loss of CO is not observed. What is observed, however, is a series of products CO(CO)(H20)+, 2H4+, CoH20+ and CoCO+. (Note: in order to distinguish CoC CoC2H4+ from CoCO+ (m/z==87) deuterated propanoic acid (d6) was used. All hydrogen containing products are given as hydrogen atoms and not deuterium.) This suggests that metal insertion/alkyl shift mechanism occurs to give ethanol as a ligand, but then the ethanol undergoes the reactions typical of alcohols (i.e. the formation of H20 and ethene). Co+ +CH3CH2COOH —> CoHZO+ +C3H4O (40%) ——> CoC3H4O+ +H20 ( 6%) —> CoCH202+ +C2H4 (19%) ——> CoC2H4+ +CH202 (11%) —> Coco+ +C2H6O (25%) Co+ also inserts into the C-CO bond in propanoic acid, a B-H atom shift leads to the Co(ethene)(HCOOH)+ complex. Competitive elimination of the ligands occurs. The reactions above are given for the ionic and neutral products regardless of the stoichiometry. All of the products formed in this mixture occur by at least two -74- .0.Ioo .-o.o.Io-.ro-oo ouoqufo. 0.x- / .0560 JIMIII. .oo\ (hem-MI I 0.00- In . e o u 0 "rum-10 o-I 0:0030 IIINWUHI .-o~I-00-eo IIufl-«olL :0 v a 0 III: I 0.- .I 0 0 4MWfl4de. 00 . 0 I :2...- Io ~10 ._~Iu m :0 - O 600 U AMWNJHHMM 00-000 .0000 :Em Irofofo- O I-I000I-eo .-.I.0-eo I U I 10001. r0.o-.oo-.ro-.ro ... i - a n a n n n u n :0 I0 ...6 Io-.3-w- r0- IoIIIo---w- Io- I0 IIIIo-.0- Io- 5:3 - .00 O o a «Eu-5m -75- independent pathways, therefore, the proposed structures for the product ions are given in scheme VII. The branching ratios calculated represent the extent of a given product being formed regardless of the pathway. 4. Butanoic Acid In Table 13 are the reactions of each metal ion with this acid along with the branching ratios. Butanoic acid exhibits reactions with Co+ (see scheme VIII) similar to those observed with prOpanoic acid. In this mixture, a product ion of m/z =ll9 is observed. Co+ +CH3(CH2)2COOH —> CoH20+ +CH3CH2CHCO (34%) -——> CoCH3CH2CHCO+ +H20 (31%) ——> CoHCOOH+ +C3H6 (13%) ——€> CoC3H6+ +HCOOH (<1%) —> CoC3H80+ +CO (22%) CoC2H4OZ+ +c2H4 This corresponds to an ion with the formula CoC3H8O+ or CoC2H402+. This discussion will assume both are present since both are reasonable based on previously discussed mechanisms. Regarding the ion of formula CoC3H80+, this ion may be Co(l-prOpanol)+ due to insertion into the CO-OH bond followed by an alkyl shift and loss of CO, or it may -75- Table 13. Reactions of the ions of Co(CO)3NO with butanoic acid (M==CH3CH2CH2COOH). + + Co + M —) Cono +CH3CH2CH=C=O (34%) -—-) CoCH3CH2CH=C=O+ +H20 (31%) ———) CoHCOOH+ +C3H6/ (13%) + Co(CO)H20 +C3H6 —-) CoC3H6+ +HCOOH/ (<1%) + . CoC3H6 +co +H20 ——) CoCHBCHZCHZOH+ +CO/ (22%) Co(C3H6)H;O +CO/ CoCH3COOH +C2H4 + + CoCO + M —) CoM +CO (27%) —) CoCOHZO+ +CH3CH2CH=C=O ( 9%) CoHCOOH+ +C3H6 +co + . --—) Cono +CH3CH2CH-C-O (45%) —) CoCH3COOH+ +C2H4 +CO/ (8%) CoCH3CH2CHZOH+ + 2C0 + CoC3H6(H20) +2co -——--) CoCOHCOOH+ +C3H6 (5%) —) CoCOCH3CH2COOH+ +CH2 (6%) + + Co(CO)2 + M —) CoCOM +Co (40%) —) CoM+ +2CO (5%) + —) Co(CO)2H20 +CH3CH2CH—C—O (55%) CoNO+ + M —) ConIOCH‘3CH2CH=C=o+ +H20 (100%) Table 13 Continues. Table 13 Continued. -77- CoCONO+ + M ——) —-) ——) Co(CO)2NO+ + M ———) Co(CO)3NO+ + M —-—) —-) CoNOM+ + co CoNOCH3CH2CH=C=O+ + H20 + co 4. CONOHZO -+CH3CH2CH-C-O CoNOM+ + 2C0 CoCONOM+ + 200 Co (CO) ZNOM+ + CO (35%) (2%) (63%) (100%) (64%) (36%) -78- I A a 6. saw / o0 +1000 1000 *IvIfl-MIWMI 1000 101+ 0 ..... = IAIIIIIIIIIIIIQ O\0<+ < LIOOUI-ou Q - I I 0 own/\ILEMI I0 0 00\/\ o I I | I +-/\-0 00% 0.. . I-m .. . 0 o N n 0u0uI0 Iu I0- « .0 I60 Iodoxz . a A ...-...... nu «A o 0....0NI0I010 ..o 0 I0 I0 I0. 0 of. :0 Lear-80.60)), g- N 0 III). +0 I “wad/\I O 000 0% /§ you O- U .ONII fi IO— :0 :F-m - :Em + 00 N 0 00.. 00-: IIIIIIOUI 00 I10. 00 LIIOx LIIIIIO/\/\ . .25: 8- ._ ..-. . ...... . 0.2 . «< .1 «10 I0 :0 . O .00 .O u . HEM 0605040. H) -79- have the structure Co(propene)(HZO)+. Since products corresponding to C0H20+, CoCOHZO+ and CoC3H6+ are also observed, this ion most likely has the latter structure. This is also consistent with what is observed in the propanoic acid reactions. For the formula CoC2H4OZ+, this product can be formed by Co+ insertion into the C2-C3 bond followed by a B-H shift and then competitive elimination from the intermediate (C2H4)Co+(CH3COOH) yielding CoCH3COOH+. It has been observed in other studies [10,46] that as the alkyl chain length increases more alkane like chemistry occurs, i.e. insertion into the C-C bonds occur with increasing frequency in addition to insertion into C-CO and Co-OH bonds. This may be what occurs in this compound. In order to resolve this ambiguity, future experiments could be performed using deuterated butanoic acid (d7), or collision induced dissociation (CID) experiments could be used to determine the product ion structure. 5. 3-Butenoic Acid _ In Table 14 are the reactions of each metal ion with this acid along with the branching ratios. The presence of the C-C double bond conjugated with the C=O bond system severely reduced the chemistry of Co+ with acids. In trans-Z-butenoic acid, + CO (+CH3CHCHCOOH -—€> COCH3CCH+-+HCOOH (100%) —80- Table 14. Reactions of the ions of Co(CO)3NO with 2-butenoic acid (M==CH3CH=CHCOOH). + _ + Co + M ——-—) CoCH3C=CH +HCOOH (100%) + + CoCO + M -—-) CoM +co (100%) + + Co(CO)2 + M ——-) CoM +2CO (32%) ———9 CoCOM+ +CO (68%) CoNO+ + M -—-—) No Reaction CoCONO+ + M ———) CoNOM+ +co (100%) Co(CO)2No+ + M -———) CoNOM+ +2co (100%) Co(CO)3NO+ + M ——) CoCONOM+ +2CO (90%) —-) Co(CO)2NOM+ +co (10%) -31- the only product observed is CoC3H4+. For acids such as formic acid the Co+ interacts with the CO and the OH at the same time. (Du-COI II: This interaction weakens the Co—OH bond, which is then most susceptible to Co+ insertion. Because of this structure XVI is preferred over structure XVII. (I OH CO+ 0 /C\ / / \II R Co" R C-OH m 21:11 But in trans-Z-butenoic acid the Co+ has two possibilities: MOH MOI-I m m From the observed reaction, the latter structure is preferred. Earlier in our discussion of the proton affinity/metal ligand affinity we noted that Co+ preferred to retain the olefin over H20. This serves as further . . . . . + . ev1dence that the latter structure is the initial Co -ac1d -82- complex. From this complex, Co+ can easily insert into the CO-C bond followed by a B-H atom shift and the elimination of formic acid. 6. 2,2-Dimethy1propanoic Acid In Table 15 are the reactions of each metal ion in this mixture along with the branching ratios. 2,2-Dimethylpropanoic acid does not react with Co+. Note that if insertion of the Co+ ion into the C-OH bond occurred there are no hydrogens in the position to shift onto the metal. An alkyl shift is also unlikely since this process decreases as the chain length and branching increases. Therefore a kinetic factor may account for the lack of formation of CoCO+ and Co(tert-butanol)+/ Co(HZO)(2-methylpropene)+. Also Co+ insertion into the C-COOH bond may not occur due to the steric hinderance of the bulky tert-butyl group. 7. Ligand Effects The reactions of CoLn+ with carboxylic acids, as also observed for thiols and butanone, are typical for monofunctional compounds also. CoCO+ reacts for the most part like Co+ with all of the unsubstituted acids studied. The only new reactions occur in butanoic acid. Table 15. 2,2-dimethylpropanoic acid -83- Reactions of the ions of Co(CO)3NO with (Mw=(CH3)3CCOOH). Co+ + M ———+ + CoCO + M ——++ Co(CO)2+ + M ——-) CoNO+ + M —-) CoCONo+ + M ———) Co(CO)2NO+ + M -———) Co (CO) 3NO+ +M ———) ——> No Reaction CoM CoCOM+ + CO CoM+ + zco No Reaction CoNOM+ + CO CoNOM+ + 2co CONOM+ + 3co CoNOCOC3H5+ + co + COOH (100%) (94%) (6%) (100%) (100%) (90%) (10%) -34- + + CoCO -+CH3(CH2)2COOH -—+> CoCOCHBCHZCOOH -+CH2: 6% ——e> CoCH3CH2COOH+-+CH2:-+CO ——e> CoCOHCOOH+-+C 5% 3H6 Initially one would assume that CoCO+ inserts into the C3-C4 bond followed by a B-H atom shift to the metal to give Co(CH CH COOH)(CH2CO)+, but CH : is lost either 3 2 2 alone or with the loss of CO. If CO partakes in the reaction it should always be eliminated. In this case where CoLn+ induces a new reaction but not because of ligand participation. What may induce this reaction to occur is that when CoCO+ complexes with the acid a higher energy species is formed than with Co+ only, therefore higher energetic reactions may occur [41,46]. A horizontal trend (a trend observed as alkyl group increases in size) is observed in the chemistry of CoNO+. In acetic and propanoic acids, CoNO+ reacts via insertion of Co+ into the C-OH bond/B-H atom shift mechanism to give CoNO(ketene)(HZO)+ which then loses the ketene and retains the H20. When CoNO+ reacts with butanoic acid (via the same mechanism) water is eliminated. After the initial reaction to give CoNO(H20)(ethylketene)+ the metal center, Co+ (of CoNO+), inserts into the C H ---CH=C=O bond, 2 5 followed by a 8-H atom shift to give CoNO(HZO)(C2H4) (CH2=C=O)+ which then loses the poorest ligand, H20, to give CoNO(C2H4)(CH2=C=O)+. The retention of water in the -85- smaller acids parallels what was observed in the thiols and CoNO+, that is, the ligand with the smaller proton affinity is retained. As discussed in the previous chapters, whenever a reaction occurs where a carbonyl is the only elimination product, the structure of the product ion has to be determined. Is the product in the form of CoM+, where M is the intact molecule, or is in the form Co(A)(B)+, where A and B represent ligand structures based upon rearrangement of the acid. In the carboxylic acids, all possible rearrangement fragments have proton affinities (therefore, metal—ligand affinities) higher than CO. Therefore, in each case, the CO should be lost. In ions such as Co(CO)X+ (x==l,2) and CoCONO+ when CO is lost there is a good probability that the product is a mixture of CoM+ and Co(A)(B)+, while in the Co(CO) NO+ the parent 2,3 molecule most likely is intact because of lack of empty orbitals on the metal. CHAPTER 6 THE CHEMISTRY OF COBALT CONTAINING IONS WITH ALLYL-X COMPOUNDS The five compounds whose chemistries are reported in this chapter contain neighboring functional groups where one group is a double bond. Armentrout and Beauchamp have reported the chemistry of Co+ with alkenes [9]. At thermal energies, no exothermic reactions occur between Co+ and prOpene. 1-Butene reacts predominantly by H2 elimination (97%) with some ethene elimination (3%). =CHCH=CH + + H + CO -+CH —CHCH CH 2 2 2 2 2 3 ‘——€> COCH =CH + +CH CH ——€> CoCH2 2 2 2 Elimination of H2 occurs with butenes much more extensively than other olefins, presumably because of the stable butadiene complex which is formed as a product. One object of this work is to determine whether these allyl-X compounds give products typical of olefins and/or X-containing species. -86- -37- A. The Chemistry of the Cobalt Ion with Allyl Compounds In Tables 16-20 are the reactions for each metal ion with the different allyl compounds. Also given are the branching ratios. n-Propylamine reacts with Co+ as follows [7]: + — -—€> COCH3NH +CH2-CH2 Co+ +CH CH CH NH 2 3 2 2 2 + -—€> COC3H7N -+H2 ——> CoCH2=CHCH=NH+ + 2H2 Amines are atypical of polar saturated organic compounds (such as alcohols) in their gas phase chemistry with transition metal ions in that metal insertion into the polar C-N bond apparently does not occur [7]. Amines actually react analogously to alkanes-—i.e. via H2 (insertion into C-H bond) and alkene elimination (insertion into C-C bond). Similar products are observed with allyl amine. Both allyl amine and n-propylamine react to form CHSN+ and CoC3H5N+. In the case of allyl amine, one . . . + . does see strong ev1dence for insertion of Co into the C-N bond: -88- Table 16. Reactions of the ions of Co(CO)3NO with allyl amine (M==CH2=CHCH2NH2). + + Co + M ————) C0C3H4 +NH3 (14%) + ————) C0C3H5 +N4 (7%) ———) CoCH2=CHCH=NH2+ +H2 (43%) + -——-) COCHBNH2 '+C2H2 (36%) + + CoCO + M -———) CoM «tCO (37%) + -———9 COC3H4 ‘+NH3-+CO (8%) + -—-) C0C3H5 -+NH2-+CO (4%) + —) CoCOC3H5 +NII2 . (1%) —-) CoCH2=CHCH=NH+ +H2 +CO (28%) + -———) COCHBNH2 -+C2H2-+CO (15%) + -——-) COCOCHBNH2 '+C2H2 (7%) + + Co(CO)2 + M ————) COCOM -+CO (59%) + -——) CO(CO)2NH3 3+C3H4 (12%) + -———7 CO(CO)2C3H5 '+NH2 (29%) CoNo+ + M —) No Reaction + + CoCONO + M ————) CoNOM (+CO (100%) Table 16 Continues. Table 16 Continued. -39- Co(CO)2NO+ + M Co(CO)3No+ + M CoNOM+-+2CO + CONOCH3NH2 + C2H2 + C0 + COCONONH3 -+C3H4-+CO Co(CO)2NONH3+-+C3H4 CoNOM+-+3CO + COCONONH3 -+C3H4-+2CO Co(CO)2NONH3+-+C3H4«+Co (74%) (17%) (3%) (6%) (86%) (3%) (11%) -9o- Table 17. Reactions of the ions of Co(CO)3NO with l-chloro-Z-prOpene (M==CH2=CHCH2C1). + + Co + M ————) CoC3H4 3+HC1 (100%) + + CoCO + M ————) CoM -+CO (74%) —) CoC3H4+ +HCl +CO (26%) + + Co(CO)2 + M -———) CoM -+2CO (100%) CoNO+ + M -———) No Reaction + + COCONO + M —) CONOM +CO (100%) + + Co(CO)2NO + M ————) CoNOM -+2CO (100%) Co(COI3NO+ + M —) CoCONOM+ +2co (100%) -91- Table 18. Reactions of the ions of Co(CO)3NO with l-bromo-Z-propene (M = CH2=CHCH2Br) . + + Co + M -———) CoC3H4 -+HBr (73%) ———) CoC3H5+ +Br (10%) —-——-) CoBr+ +C3H5 (17%) + + COCO + M -———) CoM -+CO (51%) —) C0C3H4+ +HBr +CO (43%) —) CoC3H5+ +Br +CO (6%) + ' + Co(CO)2 + M -——-) CoM -+2CO (65%) —) CoCOM+ +co (35%) + + CoNO + M ——-) CoNOC3I-IS +Br (100%) + + COCONO + M -———) CONOM d+CO (23%) + -——) CoCONOC3H3 -+HzBr (77%) + + Co(CO)2NO + M ---—) CoNOM +2CO (80%) —) CoCONOM+ +CO (20%) + + Co(CO)3NO + M -——> CoNOM +3C0 (14%) —) CoCONOM+ + 2CO (53%) —) Co(c0)2NOM+ +co (33%) -92- Table 19. Reactions of the ions of Co(CO)3NO with l-hydroxy-Z-prOpene (M = CH2=CHCHZOH) . + + Co + M ————) CoC3H4 -+H20 (100%) + + CoCO + M -———) CoM -+00 (93%) + . -———) Coc3H4 -+H20-+CO (7%) + + Co(CO)2 + M ————) CoM -t2CO (4%) -———) CoCOM+-+CO (96%) + . CoNO + M -———) No Reaction + + CoCONO + M -———) CoNOM -+CO (100%) + + Co(CO)2NO + M ————) CoNOM -t2CO (11%) ————) CoCONOM+-+CO (89%) + + Co(CO)3NO -+M CoNOM -+3CO (90%) ———+ -——9 CoNOCOC3H +.+co-+COOH (10%) 5 -‘_-‘... " _.—.. ._ .-—__ -93- Table 20. Reactions of the ions of Co(CO)3NO with allyl ether (M = (CH2=CHCH2) 2O) . C++M——-)CCH++OCH (100%) ° ° 3 5 3 5 + + CoCO + M —) CoM +CO (12%) + -——) CoCOC3HS +OC3H5 (65%) —) CoCOCHZCHCHO+ +C3H6 (23%) + + Co(CO)2 + M -———) CoM +2CO (78%) —) CoCOCHZCHCHO+ +C3H6 + CO (22%) CoNO+ + M -—) No Reaction CoCONO+ + M -————) No Reaction + + Co(CO)2NO + M ——> CoNOM +2CO (60%) + ——-—> CoNOC3H5 +C3HSO+2CO (16%) + ’1 —-—) CoCONOC3H5 +C3HSO+CO (17.) + —) Co(CO)2NOC3H5 +C3HSO (7%) Co(CO)3NO+ + M —-) CoNOM+ +3CO (90%) + ——-) Co(CO)2NOC3H5 +C3HSO+CO (10%) -94- H NH Co+ + <5"\\//’ 2 ————+> <;>\\\//'C0+- NH2 B-H shift <..Co-NH2 9' - - .co -NH2 ‘ CH2 H -NH2 -NH3 69 kcal/mole. The CoC3H5+ formation reaction also occurs with allyl bromide, which implies D(Co+-C3H5) >79 kcal/mole. Since the following reaction also occurs Ii++CO+NH' CoCO+«+CH =CHCH NH -—e> Coc3 5 2 2 2 2 it follows that D(Co+-C3HS) >{79 kcal/mole-+D(Co+-CO)}. It has been suggested that the Co+-NH2 bond is very weak [7], which explains why CoNHz+ is not an observed reaction product. Thus, for the CoCO+ ions, reactions with allyl amines are typical of amines. In addition, the fortuitous selection of the allyl group makes metal -95- insertion into this C—N bond thermodynamically accessible, and the products which follow such insertion are observed. The chemistry of Co+ is straightforward with allyl halides and alcohol. Apparently Co+ eliminates HX from allyl chloride, bromide and alcohol by the mechanism previously discussed [47]. CH2 c++/\/x—9 c+x—> C|---C+XH CCH++HX O /' 4¢>\0v o - II 0 - -€> O 3 4 CH2 X==CI,Br,CM The energy required for the process C3H5X —-> C3H£1 +HX is fairly small and similar for these compounds, ranging from 19.9 kcal/mole for allyl alcohol to 25.1 for allyl bromide [43]. In the case of the bromide we see direct evidence + . . for Co insertion: + Co+ + //\/ 8r ——>/\/Co-Br ——-> CoBr+ /\E.:> HBr elimination This may be evidence for D(Co+-Br) ---n 0. WI LEHE===30—CHEi + I/ COC3H4 + HX Evidence for the insertion, H shift mechanism has been reported for similar compounds [47]. As further evidence, the mass spectra of related compounds provide very useful information.‘ Consider the halogenated compound-allyl bromide and allyl chloride. Both have fragment ions in their spectra which indicate HX elimination can occur for the compound alone. CH2=CHCH x BI > c H + +HX 2 70 eV 3 4 BI j> Hx+3+c 70 eV H 3 4 -97- These ions are very minor in the mass spectra [48]. The mass spectra of metal-allyl halide compounds have also been reported. Iron complexes of this type, synthesized by the reaction [49] CH =CHCH X-I-Fe(CO)5 -€> (C3H5)Fe(CO)3X (X==C1,Br) 2 2 contain significant amounts of the elimination product [49], H > FeC H + +HX+3CO C "'7 0‘ —eV 3 4 3H5Fe(CO)3X Thus, with a metal present (and already inserted into the allyl-X bond), HX elimination becomes much more facile- supporting this mechanism. Allyl ether could be considered a trifunctional molecule or a bifunctional compound where the functional groups are a double bond and an -OR group. The presence of hydrogen atoms on the -OR functional group allows for additional mechanisms to occur. The metal center apparently inserts into the C-0 bond of allyl ether; as implied by the CoC3HS+ product. CO+ + MOM ___> o.:-;O—O .\ I- -98— This is not unexpected since allyl ether contains the weakest allyl-X bond of the compounds studied, D(C -OC3H5) =66 kcal/mole [43]. 335 B. The Chemistry of CoLn+ Ions with Allyl Compounds. Ligand Effects. An interesting set of reactions are observed for the Co(CO)nNO+ ions with allyl amine. First, the reactivity of the Co(CO)nNO+ ions is the reverse of what is usually observed with monofunctional molecules. The lesser coordinated metal centers containing NO are less reactive. CoNO+ and CoCONO+ do not react, while the Co(CO) NO+ do. 2,3 Co+ rearranges allyl amine into C38 and NH and 4 3’ eliminates NHB‘ Co(CO)2NO+ does the same, but retains the NH3. Since NH3 retention requires the presence of CO's, and C3H4 can be eliminated without concurrent loss of the CO (in fact, at least one CO is always retained) CO may be involved in the reaction as follows: -99- 0 0C C\+ /C0 \+ 9 /‘Co + NH2\/\ \————-> NO-Clzo-C-NH 0N >tf‘gb CH2 B-H shift No-Coi-c-NH2 / H + CH2=C=CH2 Here is proposed a COCO+ insertion into the C-N bond followed by a B-H atom shift. As the H atom shifts the cobalt remains tetra-coordinate. The allene, which cannot interact with the metal, is eliminated - leaving the elements of ammonia (as -NH2 and -H) remaining in the complex. Allyl chloride, bromide and alcohol also participate in ligand substitution reactions - displacing one or more CO's as seen in Tables 17-19. In the case of allyl alcohol (Table 19), again one sees the case of an increasing "reactivity" of the metal center as the number of ligands on the metal increases. -100- (By "reactivity", one is referring to processes other than simple ligand substitution.) For the Co(CO)nNO+ ions, no reaction is observed for n =0, only ligand substitution is observed for n==l and 2, while substitution and OH loss is observed for n =3. Again this may evidence for insertion of CoCO into the C-OH bond. If a group of atoms such as CoCO inserts into Rx, at least two possibilities exist: + {+ 0 CoCOLx + RX —~——9 R-(Eo-C-X (I) In the case of saturated compounds [7], the results have been explained by assuming that the alkyl group goes to the carbonyl, case (II) above. With allyl compounds, however, the strong allyl-metal interaction appears to direct the X-group to the carbonyl, case (1) above. Thus, the chemistry of Co(CO)3NO+ with allyl alcohol is as follows: + + ' Co(CO)3NO +WOH T9 NOCO C3HSOH + 3c0 - O N0\+ /CO 40 + 7/co 0C —Co—-C—OH -—> \ Co +CO -101- As the CO and COOH are lost, the allyl group can change from a one electron donor to a three electron donor, as empty orbitals become available on the metal. Following metal insertion, a B-H atom shift is common as a second mechanistic step, which in the other allyl compounds gave CoC3H4+ as a product. In the case of allyl ether, the insertion product has two types of B-H atom. A shift of a B-H atom from the -OC3HS group allows the complex to form a ligand similar to butadiene (i.e., propenal) which can efficiently donate four electrons to the electron deficient metal. This is seen in the reaction of COCO+ with allyl ether. CH CoCO+ + \/\0 -co 2 \ /\// mar-tier) CHK 0 \CH CH / 2\C/ to" \CH/ / H H H CH2 CH2 l J, I + / 0C3H5 h ----Co I CH H -102- Note in Table 20, that the increasing reactivity of the Co(CO)nNO+ ions with bifunctional molecules as n increases is also observed with allyl ether. CHAPTER 7 THE CHEMISTRY OF COBALT-CONTAINING IONS WITH 2-CHLORO- AND 2-BROMOETHANOL In Table 21 are the reactions for each metal ion with 2-chloroethanol along with the structure of the products. Also given are the branching ratios. The chemistry of gas phase transition metal ions with alkyl halides has been reported [5,47]. For a number of alkyl halides, the recurring mechanism is a metal insertion, B-H atom shift sequence, from which all product ions can be explained [47]. For example, consider the reactions of Co+ with ethyl iodide [47]: -C H + + 2 S + Co -+C2H5I insertion:> CZHS Co I —————+> CoI lB-H shift V CH 2 -HI + ‘1 ---cO+---IH‘ ”':> COC2H4 CH2 -c2H4 + - > CoHI The reaction of Co+ with ClCH CH Br would, then, be 2 2 expected to yield HCl and/or Br elimination products to parallel metal—induced HI elimination from ethyl iodide. -lO3- Table 21. 2-chloroethanol ~104- Reactions of the ions of Co(CO)3NO with (M =CH ClCI-I OH). 2 2 Como+ + M Cocouo+ + M CoCHZCHOH+ + HCl 4.. COClOH +CH2-CH2 CoM+ + co CoCOClOH+ + c2214 CoCHZCHOH+ + HCl + co CoCOCHZCHOH+ + HCl CoH20+ + CHZCHCl + co CoCOH20+ +CH CHCl 2 + 2OHCH2 ‘+Cl-+CO CoClOH+ +c2H4 + co CoCH CoM+-+2CO + CoCOM -+CO CoCOCHZCHOH+ + HCl + co CoCOH20+ + CHZCHCl + co Cococnzoacnz“ + c1 + co No Reaction CoNOM+ + co CoNOCHZCHOH+ + HCl + co CoNoazo+ + CH2C1CHCO CoNOOH+ + CH2C1CHCO (44%) (56%) (10%) (52%) (7%) (22%) (3%) (4%) (2%) (27%) (37%) (9%) (22%) (5%) (35%) (26%) (25%) (14%) Table 21 Continues. Table 21 Continued. -lOS- Co(CO)2NO+ + M Co(CO)3NO+ + M CoNOM+-+2CO CoCONOM+-+CO CoNOCHZCHOH+-+HCl-+2CO CoCONOCHZCHOH+-+HCl-+CO CoNOH20+-+CH2C1CHOl-+CO O+-+CH C1CHCO 2 2 CoNOOH+-+CH2C1CHCO-+CO CoCONOH CoCONOOH+-+CH2C1CHCO CoCONOM+-+2CO CoCONOOH+-+CH2C1CH co-+co 2 CoCONOH20+-+CH2C1CHCO-+CO CoCONOCHZCHOH+-+HCl-+2CO (54%) (5%) (9%) (2%) (8%) (9%) (10%) (3%) (48%) (20%) (15%) (17%) -lO6- Apparently Co+ interacts very strongly with both halogens- and a cyclic intermediate is more stable than the metal- insertion type intermediate. Co+ also forms products such as CoClOH+, with chloroethanol, indicative of a five membered ring intermediate. Co+ also appears to insert into the C-Cl bond. ,£9 + Cl OH ,' ‘\ + Co + \ / —- Cl ‘OH —.-CoCIOH + PH CH _. CH 2 2 CH2 ...-....C H2 1 Insertion 4.. Cl—Co—Cb-Itz--CH2 ‘fl-H 5m" ('H CH * Q 2 -HCI * 2 ‘ *’ CO----\\ «‘———- HClooo Co----\\ + COHCI 4' CZHJOH . CH X CH // ( )H OH OH Following the B-H atom shift, the complex will lose HCl but not CZHBOH' This is an indication that the interaction of C2H3OH with Co+ is much stronger than if this species was simply ethene. Both the double bond and a lone pair on the oxygen can bond to the metal as shown. This also follows the proton affinity/metal affinities trends discussed in previous sections. In terms of the analytical utility of metal ions as CI reagents in mass spectrometry, these results are intriguing. In the work done to date, specific mechanisms -1o7- are observed for specific functional groups. Alkyl chlorides react to give HCl elimination products. Alcohols react to give H20 elimination products by an analogous mechanism [47]. The work of Freiser [39] indicates that when two functional groups are far removed in the molecule, products representative of both groups can be observed. Here investigation of neighboring functional groups were performed. In the case of l-chloro-Z-bromoethane the two groups react as a "new" group. Elimination of HX is not observed due to the strong halogen-metal interactions which restricts mechanistic access to HX elimination products. In the case of 2-chloroethanol, insertion and ring formation apparently compete as reaction intermediates. Thus in the case of adjacent functional groups it cannot be assumed that products indicative of each group will be observed in metal ion CI. Another fascinating aspect of the chemistry of gas phase metallic species with bifunctional molecules is typified by the reactions in Table 21. Most of the current interest in this field today is the chemistry of the bare metal ions with organic molecules. Electron impact on metal-containing compounds generate metal centers with various numbers of ligands attached. In the case of metal carbonyls, as the number of CO's on the metal increases, the likelihood of any reaction except for ligand substitution decreases [5]. The previously reported reactions of the ions derived from Co(CO)3NO with -108- ethyl iodide typify this behavior. Co+ eliminates HI from ethyl iodide, as does CoCO+, but only ligand substitu- tion is observed when 2 CO's are bound to the metal [5]. H I ——> CoCOC H 1* +co + Co(CO)2 +C2 5 2 5 The same is observed for the Co(CO)nNO+ ions. For n.=0, the elimination reaction is observed, but for n==l,2,3 only displacement of l or 2 CO's by ethyl iodide is observed [5]. This is not the case for the compounds studied here. All of the metal containing ions derived from Co(CO)3NO exhibit a rich chemistry with bifunctional molecules. Presumably, in cases where two groups can bond to the metal, much more energy is released in the intermediate complex than in the case of monofunctional molecules. In general, the chemistry observed in these mixtures is much richer and more varied than mixtures involving monofunctional molecules. A number of interesting changes occur in the chemistry of a metal ion as ligands are added to the metal. Co+ alone reacts with 2-chloroethanol by oxidative addition to form CoClOH—, and by HCl elimination. CoCO+ appears to react by metal insertion into the C-Cl or the C-OH bonds and also by a cyclic intermediate. -109- Coco+ +M ——> CoM+ +co (10%) CoCOClOH+ + C2H4 -—> CoCHZCHOH+ +HCl +co (52%) ——> CoCOCHZCHOH+ + Hc1 ( 7%) —> CoH20+ +CH2CHC1 +co (22%) —> CoCOH20+ +CH2CHC1 ( 3%) -—> CoCHZOHCH2+ +c1 +co ( 4%) —> CoC10H+ +c2H4 +co ( 2%) There are also a number of processes in which the ligands may become more than just Spectators in a reaction, but may be ultimately involved in the overall process. To give one of a number of this type of reactions which can be found in Table 21, consider the following: CoNO+ does not react with Z—chloroethanol. CoNo+ +C1CH2CH20H —> no reaction With one CO present, H20 elimination is observed. 0+ +c H oc1 CoCONO+ +C1CH CH OH ——> CoNOH2 3 3 2 2 If the 2-chloroethanol rearranged to H20 and vinyl chloride on the metal, H20 elimination and vinyl chloride retention would be expected due to the possibility of multiple -llO- ligand-metal interactions with vinyl chloride. This is not observed. Since the reaction is only observed with concurrent CO elimination, this may be evidence for, not Co+ insertion, but CoCO+ insertion into a polar bond, i.e., C0 0" co NO'EOI/ + C1-CH -CH -0” a >\ +./’ $1 . a . 2 2 /Co \ CH *9 8~H shift e1 im'na- HO CH’I' 2 tion of CO and / insertion 2 C2H3c] . 0 o ”\+ 0 N o C0_E B’H \ l 4. n + / «\ CH Shift 7 O/CO "'E > CoNOHZO + 53H3C10 H0 H’//' H L CH ““v"‘/ CH C] .. _ r." /CH2 2 CO L2H3t C) 1. 2-Bromoethanol In Table 22 is a summary of the reactions for each of the metal containing ions with this compound as well as the branching ratios. The chemistry of Co+ with 2-bromoethanol is richer than with 2-chloroethanol. The analogous reactions for the Co+ ion are present in 2-bromoethanol but to a lesser extent due to other reactions being observed. A five member intermediate is again proposed for the elimination of C2H4. However the elimination of BrOH also occurs. From the identical branching ratios, these two ligands -lll- Table 22. Reactions of the ions of Co(CO)3NO with 2-bromoethanol (M==CHzBrCHZOH). ‘ Co+ + M —) CoCHZCHOH+ +HBr (3%) ——> CoBrOH+ +CH2=CH2 (15%) ———> CoCHZCH20H+ +Br (4%) —) CoC2H4+ +BrOH (15%) —) CHZCHZBr+ +CoOH (25%) ——-) CoOH+ +CH2CHZBr (8%) ——-—) CoCHZCHZBr+ +0H (13%) —) CoCHZBr+ +CH20H (7%) —) Co(M-H)+ +H (5%) Coco+ + M —) CoM+ +co (5%) CoCOBrOH+ +C2H4 —) CoCHZCHOH+ +HBr +co (34%) ——> CoH20+ +CH2CHBr +co (12%) —) CoCOH20+ +CH2CHBr (6%) —) CoCHZOHCH2+ +Br +co (6%) CoCOOH+ + CHZCHZBr ——> CoBrOH+ +C2H4 +co (2%) ———-) CoCOC2H4+ +BrOH (9%) —> CHZBrCH2+ +CoCOOH (14%) —) CoCHZBrCH2+ +co +OH (6%) —-——) CoCHZHr+ +CHZOH +co (3%) —) Co(M-H)++H+CO (3%) Table 22 Continues. Table 22 Continued. -112- Co(CO)2+ + M CoNO+ + M + CoCONO + M Co(c0)2No+ + M Co(c0)3N0+ + M CoCOM+-+ co CoCOHZO+-+CH2CHBr-+CO No Reaction No Reaction CoNOM+-+2CO CoNOHCOOH+-+CHZBrCH=C=O CoNOM+-+3CO CoCONOM+-+2CO CoNOHCOOH+-+CH BrCH=C=O~+CO 2 CoCONOHCOOH+-+CH BrCHCO 2 (78%) (22%) (82%) (8%) (l5%) (70%) (5%) (10%) I 'I i I! II ___———_—. ._.-—-._ ~113- are retained by Co+ to the same extent. This may be due to similar metal-ligand affinities. As was observed in allyl bromide, we see direct . + . . ev1dence for Co insertion: HO Co’-Br H0 """ an Br HO r ...—9 Cb*+‘\__)?‘ a \\_b/ Small amounts of Br atom elimination were observed (4%). This again is evidence for D(Co+-Br) CoOH+ +CHZBrCH2+ (28%) ——> CoOH+ +CH2BrCH2 ( 8%) ——> CoCHZBrCH2+ +OH (13%) CoCO+ reacts just as Co+ does with 2-bromoethanol. In most cases, the CO is also lost in the reaction. Just as it was in 2-chloroethanol, CoCO+ forms CoCOH20+ and + . . . . . . CoHZO in this mixture while Co+ does not. This is a new -ll4- reaction that occurs for these ions and is not due to the ligand participation reaction since the CO is both lost and retained by the metal ion. The chemistry of Co(CO)2NO+ and Co(CO)3NO+ are dominated by ligand substitution reactions (92% and 85%, respectively). The remaining chemistry of these ions involves insertion of CoCO+ into the C-OH bond with a CO inserting onto both fractions of 2-bromoethanol, followed by a B-H atom shift and the elimination CHZBrCH=C=O from the intermediate (CHZBrCHCO)Co+(CO)nNO(HCOOH) (n =0 for Co(CO)2NO+ and n =0,1 for Co(CO)3NO+). R “I” 9 o "0 OH H0 3' Br c—Co*-c-0I-I ,9 H u é* ’ _. O-C Co(co Now ——-> I -,—-> c I° \ 2.3 U U (CO) Shift H (CO) H (IZH 9+2 Br Co(CO)” NO(HCOOH)* Since these reactions do not occur for CoNO+ and CoCONO+, further evidence for two CO's insertion is provided. CHAPTER 8 THE CHEMISTRY OF COBALT-CONTAINING IONS WITH SUBSTITUTED BUTANONES As stated earlier, most compounds which are subject to analysis today by mass spectrometry are multifunctional in nature. Using chemical ionization, it is useful to have knowledge of any preference a reagent ion has for different functional groups in order to evaluate a CI spectrum of an unknown. In the continuing efforts to investigate whether metal ions can be utilized as CI reagents, bifunctional ketones and carboxylic acids were studied. In this discussion, using what has been established as reaction pathways for the various functional groups involved, we have been able to predict products based upon the assumption that products representative of each functional group will be formed. When a metal insertion/B-H shift leads to the intermediate of the type Co+(A)(B), where further decomposition of, e.g., B, have been observed, these reactions were also predicted. The predicted reactions with the branching ratios for those observed are given in Table 23. -115— {cob n. NM -116- m N. IO-x $00. «.3 33.3... .332“. 5.32:. .o .23 I o .53.: ....:.oxoo...o..l.8ox.xox.ao..831...) n 05.5 . x: .93 8 T 205.513.3821. J d xx. x o :88 on 9 xx .8 . «:68 ”II.— 7% 202.832: (IR/K 7832: o m 220252... .0 226.com anaomnam 38605 on on 513.218 I \/\ + 0:3 HIM—A . “MAXI. ...:m o 4x )1 . :.q x m. m .x. A .00 _ o \JW J .00 x +\./.\ .x8) L/R ..I \ffi ..8 o o 3200 35.85... .x Bu 0222.5 2982.: x.x.o . .ozonzooo ill m .... oxofu . .foou inflorooazofo. 8. .5. ..xfdioo U m x m .5. .Axfeoxooioo :fJxooioifo. fx . .oo o 5.x 5.x 966 3:983... 23.3. ..on. 3832a Bozooom 3:030 .\. II I: I) 38:25 35:33 -n 5.3 .& at 3.285 «8.32m mm 39: -ll7- The discussion of the chemistry of CoLn+ ions will follow the approach used with other mixtures. Ligand substitution reactions will be noted as well as similar chemistry of the CoLn+ to Co+ but the emphasis will be on new reactions involving ligand participation. A. The Chemistry of Co+ with 3-Substituted Butanones In Tables 24-26 are the reactions of each metal ion with 3-X-butanone (X =Cl,Br,OH), respectively, along with the branching ratios. In 3-chlorobutanone, 93% of the reactions were predicted (see Table 23). The only observed reaction that was not predicted was the elimination of CH3C1. Apparently, following metal insertion into the C-Cl bond, a B-H or B-methyl shift can occur leading to HCl and CH3Cl elimination. 3 CH=C=O+-+CH Co+-+CH3COCHC1CH3 -—+> CoCH c1 ( 7%) 3 3 Assuming that the last set of reactions in Table 27 are due from further decomposition of the intermediate Co(HCl)(CH3COCH=CH2)+, 86% of the reactions of 3- chlorobutanone are due to initial attack at the chlorine by Co+. Two of the possible products in the substituted butanones (products from intermediate C) are suggested by -118- Table 24. Reactions of the ions of Co(CO)3NO with 3-chlorobutanone (M==CH3COCHC1CH3). + + Co + M > CoCO(C2H3Cl) +CH4 (9%) CoCHBCCl=C=O+-+CH4 + ————) CoCH3CHO -+C2H3Cl (5%) ————> CoCH3cocH=CH2+-+Hc1 (35%) -———> CoCO+-+HC1-+C3H6 (9%) -———) CoC3H6+-+HCl-+CO (18%) --——9 CoCH3CH=C=O+i+CH3Cl (7%) -———) CoCH3COCHCH3+-+Cl (6%) ————9 CoCOC2H2+-+CH4-+HC1 (6%) ————9 CoC2H2+-+HCl-+CH3CHO (5%) + + COCO + M ————) CoM -+co (3%) -———+ CoCOCH3COCH=CH2+-+HC1 (10%) -——+ CoCH3COCH=CH2+-+HC1-+CO (66%) -——9 CoCH3COCHCH3+-+Cl-+CO (5%) ———4 CoCOCH3CHo+-+CZH3CI / (7%) + CoCOCH3CHO -+C2H2-+HC1 ———4 CoCH3CH=C=O+-+CH3Cl-+CO (9%) + + Co(CO)2 + M ———+ CoM -+2co (52%) -———a CoCOM+-+CO (31%) -——+ CoCH COCH=CH ++Hc1+2co (17%) CoNo+ + M I 3 2 No Reaction Table 24 Continues. Table 24 Continued. -ll9- CoCONo+ + M Co(C0)2No+ + M Co(c0)3No+ + M CoNOM+-+CO CoNOCH3COCH=CH2+-+HC1-+CO + CoNOM -+2co CoNOCH3COCH=CH2+-+HC1«+2CO + CoNOM -+3CO CoCONOM+-+2CO (42%) (58%) (97%) (3%) (51%) (49%) -120- Table 25. Reactions of the ions of Co(CO)3NO with 3-bromobutanone (M==CH3COCHBrCH3). + + Co + M ——> CoCH3COCH=CH2 +HBr (55%) —) CoC3H6+ +HBr +co (30%) —) CoCH3COCHCH3+ +Br (15%) Coco+ + M —) CoM+ +co * —) CoC2H3Br+ +CH3CHO (8%) + CoC2H3Br -+CH4I+CO —) CoCOCH3COCH=CH2+ +HBr (7%) ——> CoCHBCOCH=CH2+ +HBr +co (74%) -—> CoCH3COCHCH3+ +Br +co (5%) + ---—9 CoCH3CHO -+C2H38r (6%) Co(CO)2+ + M —) CoM+ +2co * —) CoCOM+ +co * ——> CoCOCH3COCH=CH2+ +HBr +co (26%) —) Cococ2H3Hr+ +CH3CHO +co (74%) + CoCOC2H3Br +cH4 +zco CoNo+ + M —> CoNOCH3COCHCH3+ +Br (100%) Cocouo+ + M -—-) ConM-HBr)+ +HBr +co (100%) + + Co(CO)2NO + M —) CoNOM +2co * ———> CoCONOM+ +co * Table 25 Continues. ~121- Table 25 Continued. Co(CO)3NO+ + M ———> CoNOM+ +3co * —) CoCONOM+ + 2co * ——> Co(CO)2NOM+ +co * ———> CoCONO(M-HBr)+ +HBr +co (27%) —-—) Co(CO) 2NO(M-HBr)+ +HBr +co (13%) ——-) CoNOCZHBr+ +CH3CHO + 3C0 (60%) * Peaks at m/z corresponding to these ions were of too low an intensity to perform double resonance. -122- Table 26. Reactions of the ions of Co(CO)3NO with 3-hydroxybutanone (M==CH3COCHOHCH3). + + Co + M ——-) CoCOC2H3OH +CH4 (3%) CoCH3c0H=c=o+ + CH 4 —-) CoCHBCHO+ +C2H3OH (12%) CoCH2=CHOH+ +CH3CHO ——) CoCH3COCH=CH2+ + H20 (32%) ——-) Coco+ +H20 +C3H6 (11%) ——) CoC3H6+ +co +H20 (13%) ——-) CoCH3CH=c=o+ + CHBOH (2%) ——-) CoCH3COCHCH3+ +0H (4%) ——-) CoH20+ +CH3COCH=CH2 (3%) ——-> CoCHOHCH3+ +CH3CO (4%) —-—) CoCOCHOHCH3+ + CH3 (3%) —-) CoCH3coc0HCH3+ +H (13%) + + CoCO + M -——4 CoM «+CO (17%) ———> CoH20+ + CH3COCH=CH2 +co (5%) —-) CoCOCH3COCH=CH2+ + H20 (2%) —-) CoCH2=CHOH+ +CH3CHO +c0/ (4%) + CoCH3CHO -+C2H3OH«+CO —) c:o(c0)c2H30H+ +CH4 +c0/ (9%) Co(CO)CH3CHO+ +C2H3OH / Co(c0)c2H30H+ +CH3CHO / CoCH3c0H=c=o+ +cH4 +co —) CoCOCHOHCH3+ +CH3CO (10%) ———> CoCOCH3COCH=CH2+ 4» H20 +co (34%) Table 26 Continues. Table 26 Continued. -123- Co(C0)2+ + M ——> '—-4 ‘—-—9 ‘———9 ——> ————) CoNO+ + M ————) + CoCONO + M ————) -———) CoCH3COCHCH3+-+OH-+CO CoCO(CO)CH2CHOH+-+CH4 CoM+-+2CO Coc0M+.+co COCOH20'+CH3COCH=CH2-+CO CoCOCHZCHOH+-+CH4-+2CO / CoCOCH3CHO+-+CH2CHOH-+CO/ CoCOCHZCHOH -+CH3CHO-+CO/ + CoCH3COH=C=O+-+CH4-+2CO CoCO(CO)CH2CHOH+-+CH Co(CO)2CH3CHO+-+CH2CHOH+ / Co(CO)2CH2CHOH+-+CH3CHO / Co(CO)CH3COH=C=O+-+CH -+co 4-+CO/ 4 ++CH -+c0/ 3 3 + 3 +CH3CO CO(CO)COCHOHCH CO(CO)2CHOHCH No Reaction CoNOM+-+CO CoCHOHCH3+-+CH3CO-+CO CoNO(CO)C2H3OH+-+CH 4 +CO/ CoCONOCH3CHO+-+CH2CHOH / CoCONOC2H3OH+-+CH3CHO / CoNOCH3COH=C=O+-+CH4-+CO CoCONOCHOHCH3+-+CH3CO (3%) (16%) (1%) (22%) (10%) (6%) (44%) (18%) (18%) (71%) (6%) Table 26 Continues Table 26 Continued. -124- Co(CO)2No+ + M Co(CO)3NO+ + M CoNOM+ + 2C0 CoNOH20+ + CH3COCH=CH2 + 2C0 CoCONOH20+ + CH3COCH=CH2 + C0 C0 (C0) 2NOHzo+ +M-H20 + co CoNO (CO) c2H3OH+ + CH CoCONOC2H3OH+ + CH3 CoCONOCH3CHO+ + CZH3 CoNOCH3C0H=c=o+ + CH 4 CoCONOCHOHCH3+ + CH 3co + co 4-+2CO/ CHO-+CO/ 0H + CO/ +2CO CoNOM+ + 3C0 CoCONOH20+ +M-H20 + 2C0 Co(CO)2NOH 0+ +M—H o +co 2 2 CoCONO(CO)C2H3OH-+CH4I+2CO / Co (c0) ZNOCHBCHO+ + C2H3OH + C0/ C0 (c0) 2NOC2H30H+ + CH3CHO + CO/ CoHOCH3C0H=C=o+ + CH 4 + 2co Co (CO) ZNOCHOHCH3+ + CH3CO + co _' .. (43%) (3%) (2%) (1%) (38%) (13%) (58%) (1%) (2%) (23%) Qomh n. NM 425- N. 10.x $00. On no 9 bmix 00> .030 ... 4.8 2.262.. 2262a 8:83. .o .23 i o fo 3:. ....:.oxoo3o4l6uxfoxio..832:+1 o 05.5 . x: 93 8 4| ..ozofoszrooixz. :1 o x:.zo..oo8U a. x58 . ..zJS . ( 28783:.)1444 :83: o o 2202525 .0 29331 826233 02235 on 5: .3438 U m aim a o ‘J\ 99X! U A)“ v. UAXT: IIQ o x w .x0 o A .8 . o \.J.\ A oou x +4 mxou) r Al F‘ :8 o o 3200 .2355“. .x 3“. 3862,". 39831 .33 . 65.58 All .m. .... oxofo . .fooo ixnraoixozofo. 8. .5. ..foto U 44. x m .5. ..foxoozo 31.983315. \rfi . .8 o 5.x 965 85:25... 28.3. to... coasted 393.com 33:25 622.326 -n 8.1.” .3 Mm 62262.. «corona km 39: ..V‘ -126' the reactions of Co+ with alkyl halides and alcohols (reactions 1 and 2). After the initial metal insertion process a bond cleavage occurs. We do not observe Cox as a product, presumably due to a unique allyl-like ligand which is formed. This indicates that the observed products are the most stable. 0 Co" The strength of the R-X bond in the three compounds in Table 27 decreases in the order OH >Cl >Br (D(R-OH) = 89.3 kcal/mole, D(R-Cl) =74.2 kcal/mole and D(R-Br) =59.9 kcal/mole.) Thus the reaction + Co +RX —> CoR++X becomes increasingly exothermic in the order X =OH,Cl,Br. Note that the branching ratios reflect this order, with the process becoming more prominent as the reaction becomes more exothermic. Reactions of Co+ with 3-bromobutanone are dominated by the bromine atom. This may be due in part to the fact that, for the compounds in Table 27, the weakest C-X bond -127- is X=Br. Also, the C-Br bond is the weakest bond.in 3-bromobutanone, while in 3-chlorobutanone the C-Cl bond strength was about the same as the two C-CO bonds (70-75 kcal/mol). In 3-bromobutanone, the C-Br bond is much weaker and appears to be the preferred site of insertion. This is indicated by the lack of any reactions typical of the ketone group. In 3—hydroxybutanone, 78% of the reactions were predictable. The loss of methanol (2%) could be considered as analogous to the loss of CH3C1 from 3-chlorobutanone. The remainder of the products of Co+ with 3-hydroxy- butanone corresponds to the loss of neutral radicals: Co+ +CH3COCH(OH)CH3 ——> CoCH(0H)CH3+ +CH3C0- ( 4%) —> CoCOCH(OH)CH3+ +CH3° ( 3%) ——> CoCH3C0C(0H)CH3+ +H: (13%) All three products appear to be Co+ insertion followed by a bond cleavage. The loss of a radical would be a relatively high energy process, usually much lower energy pathways follow the insertion process (elimination of stable neutrals) as observed in butanone. The question arises, when would the reaction M++AB —> AM+B -—> A +M+B -128- be thermodynamically favored? This would be the case if one or both of the products were unusually stable. If the neutral radical lost was unusally stable these reactions would also be observed for butanone, but they are not. Therefore, the reactions must become prominent due to the nature of the ionic product. M+B may be unusually stable because the metal-ligand bond strength (D(M+-B)) is high, due to multiple interactions of B with the metal. Note, these reactions only occur with 3-X-butanone where X=OH. One would expect the M+-OH interactions to be stronger than the M+-Cl,Br. This seems feasible since the fragment of the molecule containing the hydroxy group is always retained. Therefore, it is suggested that the loss of CH3CO gives the ionic product containing both an M-C o—bond and a further M-O interaction: H I 9 C9 ‘CfiCH’ O \H Presumably, the process is observed because D(Co+-CH(OH)CH3 >D(Co+-C2H5). To place a lower limit on D(Co+-CH(OH)CH3), assuming that the Co+-C bond strength is approximately the same as the Co+--CH3 bond (D(Co+-CH3) =61 :4 kcal/mole)[8]. Therefore the additional -129- (Co+-O) interaction must be greater than or equal to 11.3 kcal/mole. . If a 8-0 can interact with the metal, a similar interaction is proposed in the ionic product of the reaction leading to the loss of CH3. 0 \\ C / \CHa C.9*-— CH 0 \H Finally, H-loss could occur via the formation of a five member ring product with the loss of the hydroxyl H. ....CO* / ‘\/ CH3 CH35 -l30- It is interesting to note that when intermediate 2, Co(HX)(Butenone)+, dissociates, it loses HX when X-Cl and Br, but when X=OH there is competitive ligand loss which is consistent with what would be predicted from proton affinities, i.e., D°(Co+-HC1) Co(CO) 2+ +C3H6 +HC1 —> CoCH CH=Co+ +CH3Ci+ +co 3 Both pathways are observed in the chemistry of Co+ and therefore both are likely to occur with CoCO+. As expected the chemistry of Co(CO)2+ is reduced due to the lack of empty orbitals on the metal. Co(CO)2+ readily loses two CO's (52%) to give sufficient empty orbitals so HCl elimination can occur to give Co+- butenone complex. CoNO+, as observed in many other cases, is nonreactive but CoCONO+ and Co(CO)2NO+ react to eliminate CO and HCl -131- giving a product of the above structure with an additional NO ligand on the metal. Reactions at the C-Br bond also dominate the chemistry of CoCO+ with 3-bromobutanone (86%). As discussed in the allyl compounds when the carbonyl ligand actively participates in the insertion process two possibilities exist: + 1+ 0 COCOLX + RX *9 R‘FO-C-X (I) In allyl alcohol the carbonyl prefers the alkyl group (case II). In 3-bromobutanone both pathways occur. 0 O o o o o ‘ I - O COCO + N , ACd-C-(IZH-CH, _B _H__) A ll __) Co'mC": | ‘ C&’ C CH3 H3 C04- 0 344 %}_<0 «($3 -—9 C-CflVCH-CH -—————> / \ )L ( 3 sum ._ * (L l Br -132- The heat of reaction for the neutral organic molecule to undergo the above reaction with CO has been calculated to be 28.35 kcal/mole. This is of the same magnitude of most other ion/molecule reactions. Most likely the initial complexation of the CoCO+ with the substituted ketone provides this energy. Also the feasibility of obtaining a Co+-(butadiene-like) complex plays a role in this reaction. The former reaction is also the predominant reaction for Co(CO)2+ (74%). The remaining reactions are the elimination of CO and HBr. In 3-bromobutanone, direct evidence for Co+ insertion into C-Br bond is observed. .Co -Br Ova; c N—nN—eN—e Br + Br Br .-Co’ -BrH '37“ N—o/VHIB, Shift + . . . . . . CoBr is not observed which is also indicative of a weak + . . . Co -Br bond. CoCONO+ reacts to eliminate HBr with the loss of a CO. Co(CO)2NO+ doesn't give this elimination -l33- but Co(CO)3NO+ eliminated one or two CO's and a HBr. This indicates that a CO ligand actively participates in the chemistry via CoCO+ insertion into the C-Br bond where the CO prefers the alkyl group. After insertion, a CO from the complex is lost. This provides an empty orbital so a B-H atom shift may occur resulting in the elimination of HBr. Co(CO)3NO++ )OH/ —-> JOH/ .—CO—> jg/ Br OC\CI:'O $‘O +_ + Co NO OC-Co -NO /’| l O 3"“ -HBr —-> —-—) -. Shift C '.:Co"(CO)(NO) l“ 0 Due to low resolution and signal intensity on higher mass ions, confirmation of ligand substitution reactions was not obtainable in this mixture. Ions appearing at m/z corresponding to the ligand substitution processes were observed however. The reactions of CoCO+ and Co(CO)2+ with 3- hydroxybutanone resemble those of Co+ with the number of reactions reduced for the Co(CO)2+ ion. Reactions of Co(CO)XNO+ ions are typically dominated by ligand -134- substitution processes, but the rearrangement processes which are present are dominated by the ketone functionality due to the weaker C-CO bonds rather than at the C-OH bond. CHAPTER 9 THE CHEMISTRY OF COBALT-CONTAINING IONS WITH SUBSTITUTED CARBOXYLIC ACIDS A. Reaction of Co+ with Substituted Carboxylic Acids These results will be presented for a given skeletal chain with various second functional group (X =Cl,Br,OH, SH). 1. The Chemistry of Co+ with Substituted Acetic Acid Table 28 contains the predicted reactions and the branching ratios for the observed reactions for chloro, bromo, hydroxy and mercapto acetic acids. All predicted products are those expected for acids. No products for the second functional group are suggested in Table 28 since, if Co+ inserts into the C-X bond, there is no B-H available. The only products that would be predicted would correspond to X-abstraction, however in no case was this observed. The reactions observed are dominated by the chemistry of the COOH group, more Specifically insertion into the CO-OH bond. -135- -136- CoCH2=C=O+-+HOSH (22%) -—+> CHZSCOOH+-+C0H (11%) These reactions involve Co+ forming a five member ring between the S and either the hydroxyl group (structure XX) or the carbonyl oxygen (structure XXII). The first cyclic intermediate (structure XX is similar to the cyclic intermediate prOposed for the reaction of Co+ with 2-chloroethanol [4]. From this intermediate cleavage of the C-SH and CO-OH bond occurs to give an intermediate, structure XXI, which then loses HOSH. There is a possibility that this product exists as two radicals, OH and SH. In doing so, a higher energetic pathway is involved. (Note, HSOH, hydrogen thioperoxide, has been observed previously by Smardzewski and Lin in an argon matrix isolation studies [50].) OH O'-T;C..=-'CH2 O=-7.C.;"=CH2 HQ- .sH e 'ci'o+ ——-a-"C'o+ + HOSH .+ / \ Co HO SH ICXI IICEE -138- Eleven percent of the products correspond to the loss of CoH, i.e., formation of (M-l)+. In the chemistry of Co+ with thiols and carboxylic acids, the formation of CoH was not observed in any case. Therefore, in order to observe this reaction product for this acid there must be interaction of Co+ with both groups which causes this reaction to occur. The proton affinity of thiols are greater than the corresponding alcohol. Therefore the presence of the SH group breaks the dominating effect of the Co+-acid group interaction (structure XVIII). This leads to the formation of the cyclic intermediate XXII. OH - OH OH m< /‘\< m< HS... 30" 5 ..O -:C°H+ 3"“ O (364’ '\V:O¥, \ m H From this intermediate metal insertion into the S-H bond can occur followed by a charge transfer. This type of reaction is not observed in the 2-hydroxyacetic acid and may be related to the fact that the O-H bond is 15 kcal /mole stronger than the S-H bond in the thiol acid. It is interesting to note the similarity between this reaction and the elimination of H- from 3-hydroxybutanone. If it goes through the intermediates prOposed for the hydroxyl -139- ketone and thiol acid, cleavage of the Co+-H bond is observed for the ketone. However, in the thiol acid, a charge transfer and loss of CoH is observed. This is consistent with the fact that sulfur containing compounds have, in general, lower ionization potentials than the oxygen containing analogs. Therefore, loss of CoH is not unexpected for R-SH while not observed for R-OH. 2. The Chemistry of Co+ with 2-Substituted Propanoic Acids In two of the three compounds studied (X=Cl,SH) all of the reaction products were predicted (see Table 29) based on what is known of the chemistry of Co+ with alkyl halides, thiols, and prOpanoic acid. Several points to be noticed: first, elimination of HX was observed to a good degree. Insertion by Co+ into the C-X bond gave a species from which a B-H shift can readily occur to give Co(HX) (propenoic acid)+ which loses the HX to give product where the Co+ interacts with the resulting bonds which is similar to a Co+-butadiene complex. 99’ 0 OH Second, retention of water is favored over retention of the substituted ketene. Similar results, i.e. retention -l40- 300. o\.. O e\eOO. mm 0? In M _N m V I m. x ID ..x 5.x 62:38 .x. “9.362.“. 3268.“. 3.803. 3.23 xnr..rooo-ru....ro.8 a a N C 0 N X o r ofxroo U 2:08.18» 18.83:: _ .... ... 3088.8 . x8 108.8. .8 A) rooomrunro . .8 965 68:23... -x 3m 8.8.35 880?... ro8r..x.r~o8 m xfno . .8818 5082.8 :38. 4.] x 8. ..ourzxfdao III 8~rz878§ro~roi +1 0868 -.ro . .oar8 DI x x x _ _ _ of 28883188 :0. .8-mronro AL 58-8-an . .8 955 93. Bu 88.85 mcozoooa 364 8688i- x-m 5:5 . u .i 39808. 628.35 mm 203 -l4l- of H20 over ketenes as ligands, are observed in the unsubstituted acids. Third, insertion into both sides of the carbonyl group is observed to give products typical of the acid. From the type of reactions occurring and the branching ratios, it has been observed that Co+ prefers to attack at the acid functional group more than at C1 or SH. In 2-bromopropanoic acid, none of the predicted reactions of Co+ were observed. In fact, no reactions of Co+ with 2-bromopropanoic acid were observed. At present, there is no explanation for this lack of reactivity. 3. Reactions of Co+ with 3-Substituted Propanoic Acids Table 30 shows that the ability to predict the products of the reactions of Co+ with this series of compounds varies from 47% to 100%. In 3-hydroxypropanoic acid, all reactions were predicted but the exact pathway and product structures could not be determined. Two products, CoH 0+ 2 +), can be formed either through and CoC3H402+, (Co(M-HZ) reactions at the OH group, that is alcohol-like reactions, or through reactions typical of acids by attack at the -COOH group. Attack at either group would yield products of the same formula therefore the branching ratios are given for the total product regardless of the structure. In 3-bromopropanoic acid, 29% of the reaction products are due to primary rearrangement of the intact molecule, -l42- e\eOO. o\ohm $5. 0. mm mm m 02.: NN v amnm. 02.: N O. .0 n :0.x 5.x .0... 02.030 .x. .5000 0. :8... 20 :0...) .0 500 633500 230.20 03. 0.; 0250. 00 >06 ”0.1.000 000 6.13 680020 03. 009; .0 020.00... 30000.... 00:00.". .0 .20... :000: . x: . ..: .08 8.x: . .x:0.:0o0 :0:0.:0 ...x:.8o0 208:0. .:0..o0.x:. m 20.02520. .0... 020.095. 80:00.0. 00:003. 2.000033 x u I x:...:08:0. :08 HH— 0 . .ou0 ... :8:0. .:0. .xzau .:08:0..:0..o0.x:. fl :08.:0.:0 .l :08.:0 .:0..8 0:05 33:00:“. x .0... 020.00.... 30:003. 0.: . ..o "0.:0x .:0.8 U x x _ . 0.0.:0x.:0 .o.:o0 ..0.:..o0 85.5.5.1 :0- .00 -w.:0.:0 0 n u 0 x : 0..:08: 0U ... ... :08: . .x .:.0o0 280:-..00; .:.0. AI :o-0-.o0-.:0.:0 :08.:0.:0 . .00 o 0:05 0.04 .0“. 02200;. 000.808.. 00.04 0.0000000-.. -n 5.3 .mm Mm 000.800”. 020.020 on 0.00» -l43- 68% of the products result from subsequent rearrangement of the Co(HBr)(propenoic acid)+ intermediate. From this intermediate insertion into the Co-OH occurs followed by a B-alkyl shift and rearrangement to give either Co(HBr) (c0)+ (6%) or Co(cnzcaoa)+ (62%). The remaining 3% of the reactions in this mixture involves insertion into the CO-OH bond by the Co+ followed by cleavage of the Co+-OH bond. + Co ~+CHZBrCH2COOH ——e> CoCHzBrCH2C0°+-+OH ( 3%) The reactions observed for Co+ and 3-chloropropanoic acid are due primarily to reactions of each functional group directly or from an intermediate similar to what was discussed for bromopropanoic acid. The following reactions were not predicted for this mixture: Co+-+CH2C1CH2COOH ——e> CoOHCl-+C3H4O+ (27%) ——+> CoHOCl-+H-+C3H3O+ (21%) ——+> CoCH2=C=O+-+CH20HC1 ( 5%) The first reaction can occur when Co+ interacts with the Cl and the OH groups simultaneously giving a six member ring intermediate. -l44- CL... .-OH followed by bond cleavage/charge transfer process to yield CoOHCl as the neutral lost in this reaction. One possible structure for the product ion is CH2=CH-CHO+, the acrolein molecular ion. In the electron impact mass spectrum of acrolein a large (M-H)+ ion is observed [48]. Therefore if the product from the loss of Cl and OH resemble acrolein one would expect to see a loss of H to give the C3H30+ product. Also, Co+ reacts with 3-chloropropanoic acid to eliminate chloromethanol. Propanoic acid does not eliminate the corresponding alcohol, methanol. After insertion of Co+ in the CO-OH bond, instead of B-H atom shift a B-CHZCl shift may occur. The presence of the bond (D(CH -CH chlorine atom weakens the CZ-C COOH) 3 3 2 z 83 kcal/mole, D(CH2Cl-CH2COOH):z77.9 kcal/mole) thereby, making the C-C bond more accessible for Co+ insertion, which would then shift to the metal, resulting in the elimination of the substituted alcohol. -l45- 0 CH 8 = C.J,’CI:H2CH2 COOH“ CHz-CH g'C-J _ A. . B-CH2CI 2 .st 0 CH2:C=O CHIC! OH B. The Chemistry of CoLn+ with Substituted Carboxylic Acids. Ligand Effects. 1. The Chloro-Acids In Tables 31-33 are the reactions of each metal containing ion with the chloroacids. Also in these tables are the branching ratios. In an earlier section, the chemistry of Co+ with 2-chloropr0panoic acid was discussed in terms of the predictability of the reactions. As seen in Table 32 the reactions of CoCO+ with this acid follow those of Co+ with attack at the C-Cl bond being favored more than attack at the acid group. The chemistry of Co(CO)2+ is dominated by ligand substitution (82%). The remaining 18% of the reactions of Co(CO)2+ is attacked by the metal- containing ion at the bonds on both sides of the carbonyl of the acid group. Reactions of the Co(CO)nNO+ ion are also what one may expect from earlier discussions, that is, CoNO+ is nonreactive and ligand substitution predominates the Table 31. -l46- The Reactions of cobalt containing ions with 2-chloroacetic acid (M==CH 2 ClCOOH). Co+ + M ————> CoHZO+-+CHC1=C=O CoCO+ + M Co(c0)2+ CoNO+ CoCONO+ Co(c0)2No+ Co(c0)3No+ +M ———=; HUM CoCHCl=C=O+-+H o CoM+-+CO 2 CoH o+-+cuc1=c=o-+co 2 CoCOH o+-+cnc1=c=o CoCHCl=C=O+-+HZO-+CO Co(M-HCl)+-+HC1-+CO CoCH ClCH+-+2CO 2 CoM+-+2CO CoCOM+-+CO CoCOH20+-+CHC1=C=O-+CO CoNOH o+-+cnc1=c=o CoNOM+-+CO CoNOH O+-+CHC1=C=O-+CO CoCONOM+-+CO CoNOM+-+2CO CoCONOH20+-+CHC1CO-+CO ) CoNOCOM+-+2CO ————> CoCONOH20+-+CHC1CO-+CO (52%) (48%) (24%) (48%) (5%) (8%) (2%) (13%) (20%) (38%) (42%) (100%) (11%) (89%) (6%) (42%) (52%) (67%) (33%) -147- Table 32. The reactions of cobalt containing ions with 2-chloropropanoic acid (M==CH3CHC1COOH). + + Co + M -———> CoHZO ~+CH3CC1=C=O (48%) -———> CoHCOOH+-+C2H3Cl (4%) ————9 Co(CH2=CHCOOH)+-+HCl (48%) + + CoCO + M ————) CoCOHZO -+cn3cc1=c=o (5%) -———¢ Coconzo+-+cn3cc1=c=o/ (48%) CoHCOOH+-+C2H3Cl-+CO ——> Co(M-ch+ +HCl +co (62%) -———> CoCO(M-HC1)+-+HC1 (9%) ————) Coconcoon+i+czn3c1 (13%) + + Co(CO)2 + M ————) CoM -+2co (32%) ——> CoCOM+ +co (50%) ————) CoCOH20+-+CH3CC1CO-+CO/ (14%) CoHCOOH+-+C2H3Cl‘+2CO -———> Co(CO)2HCOOH+-+C2H3Cl (4%) + . CoNO + M ————) No Reaction + + CoCONO + M ————> CoNOM -+co (35%) -———9 CoNOH20+-+CH3CC1CO'+CO (59%) -———9 CoNOCOHCOOH+-+C2H3Cl (6%) Table 32 Continues. Table 32 Continued. -l47a- Co(CO)2N0+ + M Co(c0)3No+ + M —-a CoNOM+-+2CO CoCONOH20+-+CH3CC1CO-+CO CoNOCOHCOOH+-+C2H3Cl zClCOOH+-+CH2CO-+CO Co(CO)2NOHCOOH+-+C CoNOCH 2H3Cl CoNOM+-+3CO CoNOCOM+-+2CO CoCONOH20+-+CHC1CO-+CO Co(CO)2NOHCOOH+-+C2H c1-+co 3 CoCONOHCOOH+-+C H c1-+2co 2 3 (62%) (16%) (7%) (13%) (2%) -l48- Table 33. The reactions of cobalt containing ions with 3-chloropropanoic acid (M==CH2ClCH2COOH). Co+ + M ——-) CoCH2=CHCOOH+ +HCl (4%) —-) CoCZHZ+ +HCl +HCOOH (10%) —-+ CoC2H3Cl+ +HCOOH (5%) ——-9 (232140+ +CoOHCl (27%) ——-9 c3H3o+ +COOHCl +H (21%) -—-> CoCHZCHOH+ +HC1 +co (28%) ——-) CoCH2=C=O+ +CHZOHCl (5%) Coco+ + M ———> CoM+ +co (19%) ——-> CoHZo+ +CH2C1CHCO +co (42%) ——> Co(M-HCl)+ +HCl +co (33%) —> CoCOM-MCI)+ +HCl (4%) —) CoC2H3Cl+ +HCOOH +co (1%) ———> CoCOHCOOHf +C2H3Cl (1%) Co(CO)2+ + M ——> CoM+ +2co (78%) —> CoCOM+ +co (11%) ——> Coc0(M-Hc1)+ +HCl +co (10%) CoNO+ + M ) No Reaction CoCONO+ + M > CoNO(M-HC1)+ +HCl +co (100%) Table 33 Continues. Table 33 Continued. -l49- Co(CO)2NO+ + M Co(CO)3NO+ + M CoCONOM+ + co CoNO(M-HC1)+ +HCl + 2co CoCONOH o+ +CH 2 2ClCH=C=O CoNOCOM+ + 2co CoCONOH20+ + CHClCO + co CoNOHCOOH+ + C2H3Cl + 3co CoCONO(M-HC1) + + HCl + 2co (76%) (16%) (8%) (62%) (24%) (14%) ————-1 -150- Table 34. The Reactions of cobalt containing ions with 2-bromoacetic acid (M =CHZBrCOOH) . + + Co + M ————9 CoCHBr=C=O -+H20 (100%) + 4- Coco + M —) CoM +CO (74%) -——> CoCHBr=C=O+ + H20 +co (26%) + + ' Co(CO)2 + M -—-9 CoCHBr=C=O «+H20-+2CO (100%) + . CoNO + M -—) No Reaction ‘ + + CoCONO + M —) CoNOM +CO (100%) + + Co(CO)2NO + M ———> CoNOM +2co (100%) + + Co(CO)3NO + M -———) COCONOM -+2CO (100%) -— .—.—-——-— 3| -lSl- chemistry of the other ions. There is one new reaction which involves insertion of a CoCO+ into a C-C bond, followed by a B-H atom shift then ligand loss process to give CoNO(CH2ClCOOH)+. C' HOOClZ (I) a H H00? 7 COCONO‘i- Arc -—_)$H-c|:°+-C-CH‘_S-fifi-_) C'ZH-Clio’-------O+CH2-C-O OH Cl NO Cl NO h'C“-‘( “- ElHClCCOr‘ All the reactions involving the loss of the cobalt atom that was observed for the reactions of Co+ with 3-chloropropanoic acid are not observed once a ligand is placed on the metal. This is because CoCO has a higher ionization potential than that of Co (IP(Co) =7.87 eV and IP(CoCO)==l6.S eV) [43]. Other than the ligand substitution reactions, the chemistry of the CoLn+ ions are dominated by attack at the chlorine atom. What is interesting to note is that attack at the C-OH bond (to give Co(water)+ type products) is observed in these ions while absent in Co+ 2. The Bromo-Acids In Tables 34—36 are the reactions of each metal- containing ion with the bromoacids. Also in these tables are the branching ratios. -152- Table 35. The reactions of cobalt containing ions with 2-bromopropanoic acid (M = CH3CHBrCOOH) . Co + M -———) No Reaction + + CoCO + M —) CoCOHZO +CH3CBr=C=O (22%) —) CoH20+ +CH3CBr=C=O +co (28%) ——) Co(M-Hrar)+ +HBr +co (41%) -——) CoCO(M-HBr)+ +33: (9%) + + Co(CO)2 + M —) CoCOM +co (22%) -——) CoCOHzO+ +CH3CBr=C=O + co (78%) + . CoNO + M ————) No Reaction + + CoCONO + M —) CoNOM +co (14%) —) CoNO(M-HBr)+ +HBr +co (55%) —-—> CoNOHZO+ + CHBCBrCO + co (31%) + + Co(CO)2NO + M ————9 CoNOM -+2CO (51%) -———> CoCONOH20+ +CH3CBrCO +co (27%) ——> CoNOH20+ + CH3CBrCO + 2co (22%) + + Co(CO)3NO + M ————) CoNOM -+3CO (78%) ——> Co(CO) 2N0H20+ +CH3CBrCO +co (22%) .— -lS3- Table 36. The reactions of cobalt containing ions with 3-brom0propanoic acid (M==CH2BrCH2COOH). + + Co + M -———) Co(M-HBr) -+HBr (27%) —-—) CoCHZCHOH+ +313: +co (62%) -———) Co(M-OH)+ +OH (3%) —-) Co(CO)HBr+ +CH2CHOH (6%) ——) COH20+ +CH2BrCH=C=O (2%) + + CoCO + M ——) CoM +co (25%) —) Com-mar)+ +33: +co / (36%) CoCO (CH2=CHOH) + + HBr + co —) CoCO(M-HBr)+ +HBr (4%) —) CoCHZCHOH+ +HBr + zco (17%) —> CoCOHBr+ +CH2CHOH +co (2%) —) Gonzo+ +CHzBrCHCO +co (5%) —) CoCOH20+ +CHZBrCHCO (11%) + + Co(CO)2 + M —) CoM +2co (35%) ———> CoCO(M-HBr)+ +HBr +co / (1%) Co(CO) ZCHZCHOH+ +HBr +co —) Co(M-HBr)+ +HBr +2co (31%) CoCO(C2H30H+ +HBr + zco —) CoCHZCHOH+ +HBr + 3co (17%) -———) CoCO(HBr)+ +CH2CHOH + 2co (2%) —-) cocoazo+ +CH2BrCHCO +co (14%) Table 36 Continues. Table 36 Continued. -154- CoNO+ + M 4. CoCONO + M Co(C0)2No+ + M Co(c0)3No+ + M --—) CoNO(M-HBr)+-+HBr + CoNO(M-HBr) -+HBr + CoNOM «+2CO CoNO(M-HBr)+-+HBr-+3CO CoCONO(M—HBr)+-+HBr-+2CO CoNO(M-HBr)+-+HBr-+3CO CoCONO(M-HBr)+-+HBr-+2CO (100%) (100%) (96%) (3%) (1%) (67%) (33%) -155- Co+ reacted with bromoacetic via insertion into the C-OH bond with the loss of water after a B-H atom shift. This reaction also occurs with CoCO+ and Co(CO)2+ with the loss of all CO's from the metal giving the Co(CHBr=C=O)+ ion as the sole product. Co(CO)nNO+ ions react only via ligand substitution. The reactions of CoCO+ and Co(CO)2+ with 2-bromo- propanoic acid is richer than Co+. As pointed out previously Co+ does not react with this acid. The reactions of CoCO+ are divided equally between attack at the CO-OH bond and the C-Br bond. The reactions of CoCO+ and Co(CO)2+ are similar to Co+ with 3-bromopropanoic acid. Attack at the C—Br bond to eliminate HBr and a number of carbonyls is dominant reaction pathway. In the chemistry of Co+ with this acid it was noted that the elimination of both HBr and CO occurred (via successive reactions). This also occurs for both of these ions. CoCO+ reacts with 3-brom0propanoic acid to eliminate Egg CO's and HBr to form Co(CH2=CHOH)+. The product corresponding to the loss of one CO and HBr can have two possible structures: Co(Z-propenoic acid)+ or CoCO(CH2=CHOH)+. Similar product structures are also possible for the reactions of Co(CO)2+. Co(CO)nNO+ ions react to eliminate HBr from this acid. CoNO+ also reacts to eliminate HBr. This can serve as evidence that the carbonyls do not participate in the reactions. In order for the elimination of HBr to occur -156- from Co(CO)2NO+ and Co(CO)3NO+ ions, the elimination of at least one CO must occur to give a sufficient number of empty orbitals on the metal to allow insertion and B-H atom shift. Elimination of CO's remove much of the excess inter- nal energy of an activated complex. Therefore the most likely rearrangement to CO and vinyl alcohol does not occur. 3. The Hydroxy-Acids In Tables 37 and 38 are the reactions of each metal- containing ion with the hydroxyacids. Also in these tables are the branching ratios. The Co+, CoCO+ and Co(CO)2+ all react with 2- hydroxyacetic acid to give CoCHOH=C=O+. In CoCO+, 98% of the reaction is due to ligand substitution and the remainder leads to the formation of CoCHOH=C=O+, while in Co(CO)2+ rearrangement to eliminate 2CO's and H20 is the only reaction pathway. Co(CO)nNO+ ions react by ligand substitution reactions with loss of up to 2 CO's. In this compound, two new reactions are observed for CoCO+ that were not observed for Co+: elimination of acetic acid and the elimination of CH2:. The former reaction involves insertion of CoCO+ into the CH OH-CH COOH bond, where the CO is adjacent to the CHZOH 2 2 . . +- group, then a B-H atom shift to give a (CHOH=C=)Co (CH3COOH) complex from which elimination of the acid occurs. -157- Table 37. The reactions of cobalt containing ions with 2-hydroxyacetic acid (M =CHZOHCOOH) . + + Co M —-) CoCHOH=C=O +H20 (100%) + + COCO M -—-—) COM +CO (98%) -—-—) CoCHOH=C=O+-+HZO-+CO (2%) Co(c0)2+ M -———) CoCHOH=C=O+-+HZO4+2CO (100%) + . CoNO M —) No Reaction + + COCONO M —) CONOM +CO (100%) + + CO(CO)2NO M -——-) CONOM +2CO ' (100%) + + Co(CO)3NO M -——) CoCONOM +2CO (100%) ... ._—__.. .._____ ...... ...-...... .— —_____————_ -m' ...—...-..“ -158- Table 38. The reactions of cobalt containing ions with 3-hydroxypropanoic acid (M==CH20HCH2COOH). + + Co + M ———-) Col-120 +CI-12=CHOOH / (25%) CoHZO+ +CH20HCH=C=O ——) Cocuz=cncoon+ +H20 / (14%) CoCHZOHCH=C=O+ +H20 —-) CoCHZCHOH+ +HCOOH (51%) ——-—-) CoHCOOH+ +C2H3OH (10%) + + CoCO + M ————> CoM -+CO (9%) ——-) CQHZO+ +CH2CHCOOH +co/ (62%) CoHZO+ +c1120chco +co ~—) CoCOHZO+ +CH2CHCOOH/ (3%) Coconzo+ +CHZOHCHC0 ——-) CoCHZCHCOOH+ +320 +c0/ (19%) Cocnzoncnco+ + H20 + co -——) CoCOCH20+ + CHBCOOH (5%) ———) Co'cocnzoacoou" +CH2 (2%) + + Co(CO)2 + M -———) COM -+2CO (16%) ——-> CoCOHZO+ +CH2CHCOOH +c0/ (81%)- CoCOH20+ + CH 2oncnco + co 4. ———) Co(CO) ZCHZ-C(OH) 2 +CH20 (3%) CoNO+ + M.————) No Reaction Table 38 Continues. Table 38 Continued. -159- Cocouo+ + M Co(CO)2NO+ + M Co(c0)3No+ + M CoNOM+-+CO CoNOCH CHCOOH+-+H O-+CO/ 2 + 2 CONOCHZOHCHCO l+H20-+CO CoNOH20+-+CH2CHCOOH-+CO/ CoNOHZO+-+CHZOHCHCO-+CO CoNOM+-+2CO CoNOCHZCHCOOH+-+HZO-+2CO/ CoNOCHZOHCHCO+-+H20-+2CO CoNOHZO+-+CH2CHCOOH-+2CO/ CoNOH 20+-+CH20HCHCO-+2CO CoCONOHZO+-+CH2CHCOOH-+2CO/ CoCONOH + 20 -+CH20HCHCO-+2CO Co(CO)2NOH Co(CO)2NOH O+‘+CH2CHCOOH-+CO/ 2 20+-+CH20HCHCO-+CO -.——.._- .——__.—_——.- ————~ (8%) (15%) (77%) (20%) (16%) (64%) (42%) (58%) ...——————. -—.- -160- ° 0 CH2 -g -CO’ -CHZCOOH E. H 3......c + -CH C .+? I snm n ,° 2 00“ ——> Cannfi + CH3COOH OH 9" H CH 0“ OH The elimination of CH2: occurs via CoCO+ insertion into the CHz-COOH bond followed by a B-CHZOH shift to the metal. As observed in the 3-chloropropanoic acid, the CHZX-CHZCOOH bond is weakened, making the likelihood of the CHZOH shift feasible. ° ‘3 a-cu 0H 9° ‘3 -cn = 9° CHz-CHz-C'Z-Co’-COH WCchi-COH—Acohcom OH CH20H CH20H As was observed in the unsubstituted acids, the elimination of CH2: is a minor product. This reaction is also observed in Co(CO)2+. The Co(CO)nNO+ ions also react to give Co(CO)n'NO(H2)+ or Co(CO)n'NO(M-HZO)+ products. As discussed in the . + . chemistry of Co , attack at either OH group can occur -161- readily making an absolute structure assignment difficult. Most likely both structures are present. 4. The Mercapto-Acids In Tables 39 and 40 are the reactions of each metal containing ion with the mercaptoacids. Also in these tables are the branching ratios. With one small exception, CoCO+ reacts with 2- mercaptoacetic acid the same as Co+ does. Co+ was observed to react with this acid to lose HOSH; CoCO+ loses either just the 8 atom or the S atom and a CO as SCO. 'The mechanism for this reaction involves the formation of the five membered ring intermediate that was proposed for the CoH elimination, except the C-5 bond is broken in the reaction. After the formation of the ring intermediate and H atom shift to give CoCO(S)(CH2=C(OH)2)+, competitive ligand loss occurs via the elimination of S or SCO. When Co+ reacted by the same intermediate, it loses CoH. Co(CO)2+ and CoCONO+ react to give similar products as above. In all CoLn+ ions, except for CoNO+, reactions at the CO-OH bond occurs with the retention of H O by the 2 metal containing ion. . . . + . . . CoCO+ again reacts Similarly to Co With this aCid. + . CoCO reacts at both functional groups and also reacts via the cyclic intermediate. The product ion (m/z =l33) formed by the latter pathway could also be CoCO(HCOOH)+ -l62- Table 39- The reactions of cobalt containing ions with 2-mercaptoacetic acid (M==CHZSHCOOH). + + Co + M ———> Cono +CHSH=C=O (35%) —) CoCHSH=C=O+ +H20 (32%) —> (M-H)+ +COH (11%) —) CoCH2=C=O+ +110er (22%) + + CoCH + M —) CoM +co (25%) -——) Cono+ +CHSHCO +co (12%) —> CoCHSHCO+ +H20 +co (4%) —) CoCH2=C(OH)2+ +s +co (18%) —'—> CoCOCH2_=C(OH)2+ +5 (34%) —) (M-H)+ +COCOH (7%) Co(CO)2+ + M -——) CoCOH20+ +CHSHCO +co (25%) —) CoCOCH3COOH +s +co (72%) —) CoCH3COOH +5 + 2co ' (3%) + . CoNO + M -———9 No Reaction CoCONO+ + M —) CoNOHZO+ +CHSH=C=O +co (10%) —) CoNOCH3COOH +s +co (90%) Table 39 Continues. Table 39 Continued. -l63- Co(CO)2NO+ + M Co(c0)3No+ + M CoNOM+-+2CO CoCONOH20+-+CHSHCO-+CO CoNOH20+-+CHSHCO-+2CO CoNOM+-+3CO CoNOH20+-+CHSHCO-+3CO CoCONOHZO+-+CHSHCO-+2CO (24%) (63%) (13%) (12%) (22%) (66%) -164- Table 40. The reactions of cobalt containing ions with 2—mercapt0pr0panoic acid (M =CH CHSHCOOH) . + + Co + M —) CoHZO +CH3CSH=C=O (21%) ———-) CoHCOOH+ +CH2CHSH (41%) —) Co(M-HZS)+ +st (38%) CoCO+ + M ——-—) CoM+ +co (5%) —) Co(M-HZS)+ +st +co (29%) ———-) CoCO(HCOOH)+ +C2H3SH (20%) CoCH3CH=C(OH) 2* +5 +co —) CoCOCH3CH=C(OH)2+ +3 (23%) —) CoCOHZO+ +CH3CSHCO (15%) —-—) CoH20+ +CH3CSHCO +co (8%) + + Co(CO)2 + M —) CoM +2co (14%) ——). CoCOM+ +co (24%) ———> CoCOHZO+ + CHBCSHCO +co (38%) —) CoCOCH3CH2COOH +3 + co (24%) + + CoNO + M —) CoNOCH3CSH=C=O +H20 (100%) + + CoCONO + M —) CoNOM +CO (47%) —) CoNOCH3CH2COOH +s +co (53%) + + Co(CO)2NO + M —-> CoNOM +2co (64%) -—-> CoNOCH (36%) 3CH2COOH +5 + 2C0 Table 4 O Continues . -l65- Table 40 Continued. Co(c0)3No+ + M ) CoNOM+ +3co (45%) ——> CoCONOM+ + zco (55%) -l66- formed by insertion of the metal into the C-COOH bond and then a B-H atom shift process. CHAPTER 10 SUMMARY AND CONCLUSIONS The chemistry of alkyl thiols with Co+ is similar to that observed with alkyl halides and alcohols. Primarily, Co+ inserts into the C-SH bond followed by a B-H atom shift to give a (HZS)Co+(olefin) intermediate. This intermediate then undergoes competitive ligand loss according to the metal-ligand affinities. In several thiols insertion of Co+ into C—C bonds followed by B-H atom shift may occur to yield smaller alkenes and thiols. In l-butanethiol and 2-butanethiol, insertion by Co+ into C-H bonds with the elimination of H2 or H2 and H28 to give a Co(butadiene)+ complex may also occur. The reactions of CoCO+ and CoCONO+ resemble the chemistry of Co+ with thiols with one exception. In 2-propanethiol, CoCO+ and CoCONO+ react via CoCO+ insertion into a C-C bond and eventually form Co(thioketene)+. Co+ alone does not react with 2— propanethiol to give the thioketene. The reactions of the other CoLn+ ions with the thiols are dominated by ligand substitution reactions. -l67- -l68- The chemistry of Co+ with butanone is dominated by_ insertion of the metal ion into either the CH3-CO or Csz-CO bond. From the appropriate intermediate either a B-H atom shift can occur or an alkyl shift onto the metal center occurs. The CoLn+ ions reacts primarily via ligand substitution. "Decomposition" reactions of these ions with butanone follow metal ion insertion into the CZHS-CO bond, which is the weakest bond in this compound. Co+ and CoLn+ ions react analogously with 3-butenone as they did with butanone. The chemistry of Co+ with unsubstituted carboxylic acids involves initial metal ion insertion into either the C-CO bond or the CO-OH bond with the latter being the more favored site. After insertion into the C-CO bond a B-H atom shift occurs followed by elimination of a ligand. When insertion into the CO-OH bond occurs either a B-H atom shift occurs or an alkyl shift occurs to yield a Co(CO)(alcohol)+ complex which then eliminates a ligand 2; undergoes reactions typical of the alcohol to give CoCO(HZO)(olefin)+ which then loses various ligands. Co(CO)X+ and Co(CO)yNO+ ions react with carboxylic acids similarly to the reactions of these ions with monofunctional compounds. CoNO+ shows a variation of its chemistry when the alkyl chain increases in length. In acetic and propanoic acids, after insertion into the Co-OH bond and a B-H shift to give CoNO(H20)(ketene)+, -l69- the water is retained by the metal center. In the reaction of CoNO+ with butanoic acid elimination of H20 occurs. After the formation of the CoNOH20(ketene)+ complex there lies the possibility for the metal center to insert into a C-C bond of the ketene. Allylic compounds were the first bifunctional‘ molecules to be studied. The reactions of Co+ with these compounds were typical of the X-functional group with the exception of allyl amine. Normally, Co+ will not react with amines via insertion into the C-N bond to give either Co(olefin)+ or CoNH3+. This is due to the weak Co+-NH2 bond formed. However in allyl amine, the weak Co+-NH2 bond is overshadowed by the very strong Co+-C3H5 interaction which then allows a B-H atom to shift with the elimination of NH3. + ions react differently than one The various CoLn would expect. As the number of ligands increases, new reactions due to ligand participation occur. For example, Co(CO)2NO+ typically reacts via ligand substitution, but with allyl amine, retention of NH3 (as H and NH2) and loss of allene is observed. With substituted ethanol, Co+ shows a preference for the halogen functional group over the OH. Also formation of a five-membered ring intermediate also occurs, which eliminates C2H4. For 2-bromoethanol, the elimination of HOBr is also observed as well as numerous reactions involving radical formation. -170- The CoLn+ ions also show several cases of ligand participation in the reaction especially in the chemistry of Co(CO)2'3NO+ with 2-bromoethanol. The ability to predict the reactions of Co+ with bifunctional compounds can vary from 0% to 100% accuracy depending upon the compound. In each case more reactions were predicted than actually observed. This is because the reactions were predicted based upon the known chemistry of each functional group and not on metal-ligand affinities. In cases where 100% of the reactions were not predicted new reactions involving some type of interaction between the metal and both groups explained the observed reactions. The CoLn+ ions for the most part show similar chemistry to the bare metal ion, in a few cases where ligand participation is a factor. The following trend for site of attack by Co+ can be suggested: -COOH > -Br > -C1 > -SH > -OH > C=O > H This trend was derived from the observed reactions and branching ratios for Co+ with the substituted ketones and carboxylic acids. In Table 27 are the reactions of Co+ with the substituted butanones. Examining the branching ratios for reactions due to the ketone functional group and the X group, it is observed that the X-group contributes -l7l- more to the total reaction pathway. Also from this table, the Br functional group is preferred more than the chloride and hydroxy groups. Also the Cl is preferred over the OH. Placement of the -SH functional group was based upon the amount of H28 formed from 2-mercapt0propanoic acid compared to the amount of H20 formed from 3-hydroxy- propanoic acid. This may not be a valid method of comparisonchuato obvious structural differences, but using these results in conjunction with the fact that the C-SH bond is weaker than the C-OH bond one can assume that the C-SH bond would be more reactive. The acid functional group was placed as being the most reactive site by Co+. There could be arguments against this placement. Looking at Tables 28-30 there is a possibility of placing this group below the halides and alcohol from the branching ratios due to cases where the X-group appears to be more reactive. But in more cases the Co+ prefers to react at the COOH group. The hydrogen group was placed at the bottom of the list because very few reactions were observed due to Co+ insertion into C-H, O-H and S-H bonds. In 1974, Michnowicz and Munson [51] observed a similar trend in the methane CIMS study of steroidal ketones, with one exception -COOH >C=O >-Br=¥-Cl >-OH >H -172- The C=O functional group is more reactive than the halides and alcohol. No thiol substituted steroids were studied. Munson's study reflects the difference in the basicity of the various functional groups. This study not only depends upon the ability of the metal to complex with the functional group but also the metal's ability to insert into various bonds in the molecule. APPENDICES APPENDIX A -l73- APPENDIX A Table 41- Pertinent Proton Affinities. Compound PA (kcal/mole) Reference H2 101 52 CO 139 53 CH4 128 54 CZHZ 152 55 C2H4 160 55 C3H6 179 56, 57 n-C4H8 183 58 i-C4H8 193 56, 57 HCl 141 $3 54 HBr 140 :3 54 CH3C1 160 59 CH3Br 163 59 NH3 205 60 H20 165 :3 61 CHBOH _ 182 61 CZHSOH 190.3 62 H25 170 $3 61 CH3SH 185 61 HCOOH 178 63 CH3COOH 188 64 CHBCHZCOOH 189.5 65 CHZClCOOH 182.4 65 CH3CHO 185 61 CHZCO 201 :2 66 APPENDIX B APPENDIX B WATER GAS SHIFT REACTION In this appendix is described a short project attempted in Fall 1981 -— the investigation of the water gas shift reaction using ICR. A. Introduction Whenever a carboneous deposit, such as coal or oil, is found, a mixture of C0, C02, H20 and H2, known as water gas or synthesis gas, is also found. This water gas is believed to be a precursor of these carboneous materials, which are formed through the reactions of hydrogen or water with carbon monoxide according to the following: 3H -+CO ——e> CH -+H 0 2 4 2 and ' nH2-+mCO -—€> H20-+alkanes, alkenes, alcohols, etc. -174- -l75- The second reaction is the Fischer-Tropsch reaction. Water gas, as it occurs naturally, contains between one and two moles of H2 per mole of CO. In order for production of organic compounds to occur, the H2 content must be increased. This is effected through the water gas shift reaction (WGSR): CO-+HZO === CO -+H A major goal of catalysis research is to develop catalysts that would increase the hydrocarbon or H2 yield from the last two reactions. This work encompasses the pre-World War II experiments by Fischer and others [67] in the search for heterogeneous catalysts to produce hydrocarbons, as well as more recent work by Darenbourg and others [68-75] using metal carbonyls as homogeneous catalysts for the water gas shift reaction. Not only yields of hydrocarbons and hydrogen were investigated, but also the mechanisms involved. Among the mechanisms suggested, the intermediates most often prOposed include metal carbides, formaldehyde-metal complexes (M(CH20)), and n2- CO bonding complex [75]. The proposed intermedi- ates are specific for each metal and experimental methods used. One of the more recent and exotic approaches to study CO reduction has been through temperature programmed decomposition experiments (TPDE) [73]. The most efficient -l76- homogeneous catalysts have been polynuclear carbonyls. This may be because these multinuclear centers are required to activate the carbon monoxide substrate. An alternative concept is that mononuclear carbonyls (which are stable under the reaction conditions of high tempera- ture and pressure) are catalytically active. In TPDE, metal carbonyls are thermally decomposed on alumina or silica to give subcarbonyl species, i.e. Ni(CO)4 -—€> Ni(CO)3 ——€> Ni(CO)2 -—€> NiCO ——€> Ni Among compounds tested were Mn2(CO) Ru3(CO)12, 10' Cr(CO)6, Fe(CO)S and Ni(CO)4. The activities of these compounds were measured as a function of the moles of CH4 produced per mole of metal. The results indicate that polynuclear carbonyls are more reactive than the mononuclear. As different investigators develOp new techniques and use different and exotic compounds for the WGSR, new mechanisms arise. Some examples follow. Wong and Labinger [74], using (nS-C5H5)NbH3 and Cr(CO)6 in the presence of H2, observed selective formation of C2 hydrocarbons, mainly ethane. By using D2, and intermediates of methylene, :CH2 (:CDZ), is shown to be formed. ~177- Re(NO)(CO)2+ and BH4- proposed Re(NO)(CO)(CH3). Casey [76] using CSHS a mechanism using a -CH3 complex, CSH5 Other work on other compounds suggest intermediates such as a n-CO conditioning mode. B. Project Description Most of the work done on the WGSR has been done on a pressure range from 0.9 atm up to 1000 atm, and a wide range of temperatures as well (loo-400°C). Under these conditions it is difficult to isolate intermediates and only final products are observed. For example, the TPDE subcarbonyl species Ni(CO)3, Ni(CO)2 and NiCO are adsorbed on alumina and cannot be isolated separately. Also one may not be sure if one species or another is more reactive. With ICR some of these problems can be solved. For example, if we take Ni(CO)4 at low pressure, the following positive ions are formed: Ni(CO)4+, NI(CO)3+, Ni(CO)2+, NiCO+, Ni+ The reactions of Ni(CO)4+ will parallel experiments with mononuclear carbonyls, while the other ions will parallel TPDE studies of subcarbonyl species. The intended work center around the reactions -l78- M(CO)X+-+HZO-+CO ——€> products M(CO)X+-+H2-+CO ——€> products If the reaction + M(CO)2+-+H ——€> MCOH2 -—€> products 2 it may be evidence for a formaldehyde intermediate. To test this, using formaldehyde we can see if M(CO)2+ reacts in the same way. C. Results and Discussion The majority of work done on the water gas shift reaction involved the ions derived from Co(CO)3NO. None of the ions reacted with H2 at any pressure or ratio of metal:H2. This is due to the poor nature of H2 as a ligand (PA(H2):=lOl kcal/mole). If a complex were to form between Co(CO)X+ or Co(CO)XNO+ with H2, the poorest ligand would be lost. Most likely there would not be any chance for the H-H bond to cleave enabling the H atoms to bond to the metal or CO. In investigating the chemistry of these ions with the other component of the WGSR, (H20), ligand substitution reactions only occurred. -l79- Coco+ +H20 ——> CoHZO+ Co(CO) + +H 0 ——> CoCOH o+ ——> Co(H 0) + 2 2 2 2 2 + + CoCONO +H20 ——> CoNOHZO + + + Co(CO) NO +H 0 —> CoCONOH o -—> CoNO(H 0) 2 2 2 2 2 + ——> CoNOHZO +2co + + Co(CO)3NO +H20 —> Co(CO)2NOH20 +co —> Co(CO)NOH20+ +2co ——> CoNO(H20)2+ As can be seen here the components of the WGSR are placed on the metal, that is, CO and H20 are together on the metal ion. The exact form (CO/H20, COZ/H2 or HCOOH) could not be determined. In an attempt to elucidate the product ion structure, I would have liked to have done further ligand displacement studies. If a third gas, with a higher metal-ligand affinity than CO and H20, was admitted to the ICR, displacement of CO, H20 or both should occur. By varying the proton affinity of this third gas sample, the determination of the product ion structure might have been possible. This ligand displacement study was not performed because at that time the ICR was equipped with only two inlet lines. Similar reactions with H20 and the ions of Fe(CO)5 were observed. The ligand displacement reactions were identical to those observed by Foster and Beauchamp [77]. -180- Fe(CO)X+ +H20 —— Fe(CO)X__l(HZO)+ +co x = 1-4 + — Fe(CO)X_2(H2O) 2 +2co x =2-4 + - ——— Fe(CO)X_3(H20)3 -+3CO x-—4 Again ligand displacement studies were desired but could not be performed. Our initial goal of this project was to see if any small hydrocarbons of aldehydes or alcohols could be formed under the conditions in the ICR. In all the studies performed no evidence for these reactions was observed and it was decided to shelf this study for one that looked more promising. APPENDIX C APPENDIX C USING THE ZX 81 TO CONTROL TRAPPED ICR EXPERIMENTS Experiments in this dissertation were performed on an ICR in the drift mode. Drift mode is the experimental set-up in which the ions start moving from the source region of the ICR cell to the analyzer region as soon as they are formed. This set-up allows for a few collisions before the ions are detected. An alternative method of using the ICR is in the trapped mode. The ions are held in the source region for a finite time and then allowed to drift to the analyzer for detection. Since the ions are held in the source for longer periods, more collisions occur which result in greater conversion of the reactants to product ions [78]. The ICR spectrometer in use at Michigan State University is capable of being used in either mode. In the drift mode, one trapping plate in the source has a square wave signal on it to modulate the experiment. For example, if one is analyzing positive ions, one of the two source trapping plates would always be at a positive potential, the other trapping plate would be at that -181- -183- which can be Opened and closed using POKE statements. To close a relay, the user POKEs in a number corresponding to the binary equivalent of the relay to be closed into memory location 32766. For example, to close the rightmost relay one would enter POKE 32766.1; to close the next relay enter POKE 32766,2. If the user wants to close more than one relay, he/she simply adds the two numbers. For example, POKE 32766,255 will close all the relays. Each relay has a LED associated with it to indicate if it is closed (LED on). The relays are rated at 0.75 amps max switch and 2 amps max carry. They are rated to carry a maximum of 200 VDC. A card edge connector is used to connect the ICR voltages to the relays. The proqram contains a BASIC routine which allows the user to enter the pulsing sequence that he/she desires. In an early version of this proqram, the user has to enter which relays to be closed and a parameter to determine the pulse width. This required the user to look up in a table the appropriate factor. The shortest usable pulse was 1.52 msec, the longest pulse that could be used in one step was 400 msec. If pulses longer than 400 msec were desired, the user entered a multiple request for that relay. As pulses became longer, limitations due to resolution of the programming arose. By this is meant, if a pulse of 202 msec was desired that user -184- could not POKE in the appropriate value directly, a number of 178 save a pulse width of 200 msec and 179 save a pulse width of 205 msec, therefore the user had to make at least two relay requests to get a pulse width of 202 msec (number 127 and 128 gives pulse widths of 100 and 102 msec, respectively). Therefore, before using this program the user had to calculate each relay and for how long it was to be on. A modification of the BASIC program needed to be done. A second proqram made an assumption that the minimal pulse width was 5 msec and that longer pulses were multiples of this. This allowed for 200 individual on-off transitions to occur. (The earlier version allowed 255 transitions to occur.) The BASIC routine prompted the user when each relay was to be closed and opened in the one second time span, multiple closings and openings are permitted. The computer then calculates the sequence of which relays are closed at each 5 msec interval and stores them in memory. After calculating the sequence the computer then inquires if any errors were made in the entry of the sequence. If there were, the memory is cleared and the sequence is re-entered. If no errors were made the BASIC routine yield control to the machine language user routine. The machine language routine has three parts to it. The first is portion loads the starting memory location of the pulse sequence into register HL, the contents of this -l85- location is then loaded into register A and stored in memory location 40BB. The A register is decremented by l and compared to register B (set equal to zero). If the result. is equal to zero the last item in the sequence has been reached and the machine language routine jumps to START which loads the starting meory location into HL. If the last item in the sequence was not reached, the second portion of the routine is encountered. This portion loads the RELAY REGISTER with the relay to be closed. This is the machine language equivalent of a POKE statement. The next item in the list, the pulse width parameter, is then stored in registers B and C. A nested loop then decreases each register causing the selected relay(s) to be closed for 5 msec (if the second BASIC program was used). After the nested loop are executed the machine code enters the user routine ending subroutine. This subroutine calls a ROM subroutine that looks to see if any key was struck [79]. If a key was struck, the program returns to the BASIC program and prompts the user if the same or a new sequence is to be used or if the user wishes to stop. If no key was struck, the machine language routine then jumps to 4089b and loads the next item from the list and continues as before. One item to note is the pulse width is not just determined by the nested 100p but also by the number of operations done between the POKE statement. Also the -186- second BASIC program allows 200 5 msec pulse width intervals in one second; if the minimal pulse width is to be longer a change in the pulse width parameter in the BASIC routine has to be performed.(Figure 5). These programs were tested using a 5 volt power supply as the load on the relays and the output was monitored on an oscilloscope. an actual experiment to date. address (hexadecimal) 4082 4085 4086 4089 408C 408E 408F 4092 4093 4095 4096 4097 409A 409B 409D 409E 409F 40A0 40A1 40A3 40A4 40A5 40A7 40AB 40AB 40AC 40AD 40AB 4OAF 40B0 40B2 4084 MC Label 213075 START 73 323340 _ 3ABB$0 LOOP 0600 3D 323340 38 2330 23 73 323373 23 3300 46 43 2S on 15 39 20FC 05 38 2037 35 CEBB02 44 4D 51 31 14 3300 CA8940 c9 END These programs were never used in 0P LD HL,LIST LD A,(HL) LD COUNT, A LD A,COUNT LD B,0 DEC A LD COUNT, A CP B JR Z, START INC HL LD A, (HL) LD RR,A INC HL LD A,0 LD B,(HL) LD C,(HL) DEC C CP C JR NZ 1$ DEC B CP B JR NZ 2$ PUSH HL CALL KSCAN LD B,H LD C,L LD D,C POP HL INC D LD A,0 JP Z, LOOP RET -187- The BASIC controlling program for the modified pulsing sequence is given below. 1 2 1000 1010 1020 1030 1040 1050 1060 1065 1070 1080 1090 1100 1110 1120 1130 1140 1150 1160 1170 1180 1190 1200 1202 1204 1206 1210 1220 1230 1240 1250 1260 1270 1275 1280 1290 1300 1310 1320 1340 1350 1360 1370 "RELAYS" MACHINE LANGUAGE ROUTINE- R(201) N(8) 3(3) FOR I=1 TO 200 LET R(I)=0 NEXT I FOR R=1 To 8 CLS PRINT OR N)" INPUT AS IF A$(1)="N" THEN GOTOT 1270 PRINT "INPUT AT WHAT TIME DOES RELAY ":R;"COME ON AND GOES OFF"; INPUT N(R) INPUT F(R) PRINT N(R) ,F(R) LET J=N(R)/5 LET K=N(R)/5 13 R=1 THEN IF R=2 THEN IF R=3 THEN 13 R=4 THEN IF R=5 THEN IF R=6 THEN IF R=7 THEN LET 13 R=8 THEN LET FOR L=J+1 TO K LET R(L)=R(L)+RF NEXT L PRINT "DOES RELAY ";R;" COME ON AGAIN IN THE SAME SECOND?(Y OR N)" INPUT A$ IF A$(l)="Y" NEXT R CLS LET 1=30000 POKE 1,201 FOR M=1 TO 200 LET I=I+l POKE I,R(M) LET 1=I+1 POKE 1,27 NEXT M CLS REM REM DIM DIM DIM NON PRINTABLE "DOES RELAY ":R:"COME ON IN THIS EXPERIMENT?(Y LET LET LET LET LET LET II II II II II II II II HQWHthH wam 8888832188 THEN GOTO 1100 1380 1390 1400 1410 1420 1430 1440 1450 1460 1470 1480 -188- PRINT " TO STOP PROGRAM NOW INPUT ""8"". TO STOP PROGRAM ANY OTHER TIME HIT ANY KEY. INPUT ANY LETTER." INPUT A$ IF A$(l)="S" THEN STOP FOR D=1 TO 200 NEXT D RAND USR 16514 PRINT " WELCOME BACK. DO YOU WANT THE SAME SEQUENCE, NEW SEQUENCE OR QUIT? (INPUT S OR N OR Q)" INPUT A$ IF A$(l)="S" THEN GOTO 1410 IF A$(1)="N" THEN GOTO 1030 STOP 4189- C: ...: .C U . O 0 [K C +4 . q, Q) Q 3 Q o F—‘ r—--‘ ...: C Q) O. O E L. 3 +— L: __J __ L‘} j£ LI :2: Figure 5. Timing Diagram for Pulsed ICR. TIME LIST OF REFERENCES 10. 11. 12. LIST OF REFERENCES Articles representative of different research labs in the field. (a) J. Wronka, D.P. Ridge, J. Am. Chem. Soc. 106, 67 (1984); (b) R.W. Jones, R.H. Staley, J. A—m—. Chem. Soc. 104, 2296 (1982); (c) D.B. Jacobson, B.S. FrEIEer, Organometallics 3, 512 (1984); (d) L.F. Halle, P.B. Armentrout, J.L. Beauchamp, Organomet. 1, 963 (1982). For a detailed survey of the gas phase chemistry of metal containing ions with organic molecules see: S.K. Huang, Ph.D. Dissertation, Michigan State University, 1983 and B.D. Radecki, M.S. Thesis, Michigan State University, 1982. R.C. Burnier, G.D. Byrd, B.S. Freiser, Anal. Chem. 22, 1641 (1980). M. Lombarski, J. Allison, Int. J. Spectrom. Ion Physics 49, 281 (1983). J. Allison, D.P. Ridge, J. Am. Chem. Soc. 101, 4998 (1979). C.J. Cassady, B.S. Freiser, S.W. McElvany, J. Allison, submitted to J. Am. Chem. Soc. B.D. Radecki, J. Allison, J. Am. Chem. Soc. 106, 946 (1984). P.B. Armentrout, J.L. Beauchamp, J. Am. Chem. Soc. 101, 784 (1979). P.B. Armentrout, J.L. Beauchamp, J. Am. Chem. Soc. 103, 6624 (1981). A. TsarbOpoulos, J. Allison, Organometallics 3, 86 (1984). S.K. Huang, J. Allison, Organometallics 2, 883 (1983). F.W. McLafferty, "Interpretation of Mass Spectra" 3rd ed., University Science Books, Mill Valley, CA, 1980, Chapter 3. -190- 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. ~191- M.S.B. Munson, F. Field, J. Am. Chem. Soc. 22, 262 (1966). L.R. Anders, J.L. Beauchamp, R.C. Dunbar, and J.D. Baldeschwieler, J. Chem. Phys. 22, 1062 (1966). J.D. Baldeschwieler, Science 159, 263 (1968). G.C. Goode, R.M. O'Malley, A.J. Ferrer-Correia, K.R. Jennings, Nature 227, 1093 (1970). G.C. Goode, R.M. O'Malley, A.J. Ferrer-Correia, K.R. Jennings, Chem. Brit. 2, 12 (1971). J.D. Baldeschwieler, S.S. Woodgate, Acc. Chem. Res. 2, 114 (1971). G.A. Gray, Adv. Chem. Phys. 22, 141 (1971). G.C. Goode, R.M. O'Malley, A.J. Ferrer-Correia, K.R. Jennings, Adv. Mass Spectrom. 2, 195 (1971). M.L. Gross, C.L. Wilkins, Anal. Chem. 22, 65A (1971). J.L. Beauchamp, Ann. Rev. Phys. Chem. 22, 527 (1971). M.B. Camisarow, A.G. Marshall, Chem. Phys. Lett. 22, 282 (1974). M.B. Camisarow, A.G. Marshall, Chem. Phys. Lett. 22, 489 (1974). M.T. Bowers, D.D. Elleman, J.L. Beauchamp, J. Phys. Chem. 22, 3599 (1968). J.L. Beauchamp, L.R. Anders, J.D. Baldeschwieler, J. Am. Chem. Soc. 22, 4569 (1967). J.C. Dunbar, J. Phys. Chem. 12, 469 (1969). T.A. Lehman, M.M. Bursey, "Ion Cyclotron Resonance Spectrometry", John Wiley and Sons, New York, 1976. Warnick, Anders, Sharp, Rev. Sci. Instr. 22, 929 (1974). G.H. Weddle, J. Allison, D.P. Ridge, J. Am. Chem. Soc. 22, 105 (1977). R.R. Corderman and J.L. Beauchamp, J. Am. Chem. Soc. 98, 3998 (1976). 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. -192- J.L. Beauchamp, A.E. Stevens, and R.R. Corderman, Pure Appl. Chem. 22, 967 (1979). R.W. Jones and R.H. Staley, Int. J. Mass Spectrom. Ion Processes 22, 35 (1981). M.M. Kappes and R.H. Staley, J. Am. Chem. Soc. 104, 1813 (1982). M.M. Kappes and R.H. Staley, J. Am. Chem. Soc. 104, 1819 (1982). R.W. Jones and R.H. Staley, J. Phys. Chem. 22, 1387 (1982). J.S. Uppal and R.H. Staley, J. Am. Chem. Soc. 104, 1235 (1982). J.S. Uppal and R.H. Staley, J. Am. Chem. Soc. 104, 1238 (1982). R.C. Burnier, G.D. Byrd, and B.S. Freiser, J. Am. Chem. Soc. 103, 4360 (1981). J. Allison and D.P. Ridge, J. Am. Chem. Soc. 100, 183 (1978). J. Allison, S.K. Huang, M. Lombarski, S.W. McElvany, B.D. Radeck, A. Tsarbopoulos, presented at the Thirty-First Annual Conference on Mass Spectrometry and Allied Topics, Boston, MA, 1983. J.A. McCleverty, J. Mol. Catal. 22, 309 (1981). A11 thermochemical data, unless noted otherwise, were obtained from: (a) J.L. Franklin, J.G. Dillard, H.M. Rosenstock, J.T. Herron, and K. Draxl, NSRDS-NB526, Nat. Stand. Ref. Data Ser., Nat. Bureau Standards 22, (1969); (b) S.W. Benson, "Thermochemical Kinetics", Second Edition, J. Wiley and Sons, New York, 1976. L.F. Ha11e, W.E. Crow, P.B. Armentrout and J.L. Beauchamp, unpublished reSults. S. Babinec and J. Allison, submitted to J. Am. Chem. Soc. A. Tsarbopoulos, B.D. Radecki and J. Allison, presented at the Thirty-Second Conference on Mass Spectrometry and Allied Topics, San Antonio, TX, 1984. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. -l93- J. Allison, D.P. Ridge, J. Am. Chem. Soc. 22, 7445 (1976). E. Stenhagen, S. Abrahamsson, and F.W. McLafferty, "Atlas of Mass Spectral Data", Interscience, New York, 1969. A.N. Nekrasov, Yu.S. Nekrasov, N.P. Avakyan, I.T. Kritskava, J. Organomet. Chem. _2, 375 (1971). R.R. Smardzewski and M.C. Lin, J. Chem. Phys. 22, 3197 (1977). J. Michnowicz and M.S.B. Munson, Org. Mass Spectrom. 2, 49 (1974). A.J. Duben and J.P. Lowe, J. Chem. Phys. 22, 4270 (1971). D.K. Bohme, in "Interactions Between Ions and Molecules", P. Ausloos, ed., Plenum Press, New York, 1975, p. 489. M.A. Haney and J.L. Franklin, J. Phys. Chem. 12, 4328 (1969). J.L. Beauchamp, in "Interactions Between Ions and Molecules", P. Ausloos, ed., Plenum Press, New York, 1975, p. 413. F.P. Lossing and G.P. Semeluk, Can. J. Chem. 22, 955 (1970). w. Tsang, J. Phys. Chem. _7_§_, 143 (1972). J.L. Beauchamp and M.C. Caserio, J. Am. Chem. Soc. 22, 2638 (1972). J.L. Beauchamp, D. Holtz, S.D. Woodgate and S.L. Patt, J. Am. Chem. Soc. 22, 2798 (1972). D.H. Aue, H.M. Webb, M.T. Bowers, J. Am. Chem. Soc. 22, 311 (1976). M.A. Haney and J.L. Franklin, J. Chem. Phys. 22, 2028 (1969). D.H. Aue and M.T. Bowers, in "Gas Phase Ion Chemistry", M.T. Bowers, ed., Academic Press, New York, Vol. 2, 1971, p. 7. S.L. Chong and J.L. Franklin, J. Am. Chem. Soc. 22, 6347, 6630 (1972). 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. -194- R. Yamdagni and P. Kebarle, J. Am. Chem. Soc. 22, 1320 (1976). J.P. Briggs, R. Yamdagni, and P. Kebarle, J. Am. Chem. Soc. 22, 5128 (1972). J. Long and M.S.B. Munson, J. Am. Chem. Soc. 22, 2427 (1973). H.H. Storch, N. Golumbic, R.B. Anderson, "The Fischer-Tropsch and Related Synthesis", Wiley and Sons, Inc., New York, 1951. P.C. Ford, R.G. Rinker, R.M. Laine, C. Ungermann, V. Landis, and S.A. Moya, Advances in Chemistry 22, 81 (19 ). C.C. Frazier, R. Hanes, A.D. King, Jr., R.B. King, Advances in Chemistry 173, 94 (19 ). D.J. Darensbourg, M.Y. Darensbourg, R.R. Burch, Jr., J.A. Froelick, and M.J. Incorvia, Advances in Chemistry 173, 106 (19 ). R. Pettit, K. Cann, T. Cole, C.H. Maudlin, W. Slegeir, Advances in Chemistry 173, 121 (19 ). C.P. Casey, S.M. Newmann, Advances in Chemistry 173, 131 (19 ). A. Brenner, D.A. Hoovl, J. Am. Chem. Soc. 102, 2484 (1980) and J. Catalysis 22, 216 (1980). K.S. Wong, J.A. Labinger, J. Am. Chem. Soc. 102, 3652 (1980). E.L. Meutterties, J. Am. Chem. Soc. 22, 1296 (1976). J.M.S. Henis, "Ion-Molecule Reactions", Vol. 2, Plenum Press, New York, 1972, Chapter 9. C.P. Casey, J. Am. Chem. Soc. 102, 1927 (1980). a) R.T. McIver, Jr. and R.C. Dunbar, Int. J. Mass Spectrom. Ion Phys. 2, 471 (1977); (b) R.T. McIver, Jr., Rev. Sci. Instrum. 22, 111 (1978). T. Baker, "Mastering Machine Codes on Your ZX 81", Reston Publishing Co., Inc., Reston, VA, 1981.