1V1531_] RETURNING MATERIALS: Place in book drop to LJBRARJES remove this checkout from ”- your record. FINES win be charged if book is returned after the date stamped be10w. ‘Ifl m m 0! A (”BLISIOI an m I'm-B8017!!! I“ mm W By Brian Allen Bckenrode A THESIS Sub-itted to Michigan State University in partial fulfill-eat of the require-ent- for the degree of mm 0? 3cm Depertlent of Ch-ietry 1% DISIGADWO’ANIJJSIGMM “MIND I“ mm W By Brian Allen Bckenrode By e-ploying ion source pulsing with tine resolved detection through a magnetic sector a new technique celled tins-resolved ion Imentt- spectruetry (TRDB) has been developed. This technique has the capability of providing EM infornstion. Comared to sore conventional Elm techniques, TRDB offers several advantages. However, it is also plagued with several liaitations. One of the lisitations is the lack of an operationally adequate collision cell. A collision cell has been designed and inleIented for TRDB. The cell is differentially M and is nodular in construction. The results indicate an inrovuent over the collision cells previously used for TRDB. A sexin- CAD efficiency for helitn is approxi-ately 5.58 at 2.0 x 10“ torr (corrected Penning). This cell, with its peripheral pulping addition, also exhibits superior target gas exhaustion when W to preview cells. In a qualitative Brian Allen Bckenrode sensitivity test with phsnetole, the cell has proved to be superior to those previously used, but not as good as that achieved with the "solecular hes-P design. This thesis probes the CAD process, and investigates collision cell designs. The static—dynasic nature of the new cell is discussed and evidence of its perforsence is presented. Dedicated to the senory of sy sister Joan First and fore-est, I would like to thank sy advisers Jack flatson and Chris Rnke. Both have taught se to attack problels intensely and fro. sany different perspectives. Their guidance and support have assisted sy develop-ant greatly. The now Dr. John T. Stults deserves a great deal of thanks for his patience, concern and ability to answer the snltitude of questions I’ve had for his! I would also like to thank the seibers of the Hatson and Bake groups as well as the Mass Spectrosetry Facility selbers for their help and friendship. Special thanks are in order for Russ Geyer of the sachine shop for an ekcellent construction of the collision cell and Bruce Newcose for design ideas. I would also like to thank the other seabers of sy colaittee, John Allison, Jack Holland and Willies Reusch. They have provided valuable discussions and criticise. Finally, I wish to thank my wife, JoAnne, sy fasily and her fasily for their love and encouragesent throughout this period of sy develop-ant. TIBLB OF CONTENTS LIST’O' TIILIS LIST'OI Flflflflli CHAPTER I. INTRODUCTIOI’ .............................. . ......... Thesis Goals .... ........................... . ......OOOOOOOOO MEANS and the TRDMS Technique . .................. References ..... ............ ............. CHAPTER II. THE DIRTIIOF CAD AND ITS POIIR ..................... The Discovery of CAD .......... ............................... The Isportance of Collisional Activation Data ................ References .... ....... ... ................... . ............... .. CHAPTER III. CAD THIGH! ........................................ The Basic CAD Process ................ ...... .................. The Mechaniss ...................... ...... ...... . Revised Reaction Schese ................ . ......... . ....... . Energy Considerations ................... . ..... . ........... Kinetic Energy Loss ......... ..... ...................... Kinetic Energy Release ..... ............................ CAD Events ... .......................... ,..2..... Before the Collision Cell ................................. During Passage Through the Cell .. Subsequent Fragsentation .................................. Relationship of CAD to El .......... ' ......... . ....... ........ Collision Cross Section ................................. ..... References .................................................. . CHAPTER IV. INSTRUMENTATION AID CELL DESIGIB . 1.118 mm Imtmt sososoeesseseeoe 0000000000000 Hardware ........ .......................................... Software ........................... . ...................... Scan Options ...... . ....................................... Conductance Considerations ............................ ...... Knudsen Nusber .............................. . Flow Conductance and Ispedance .............. . Viscous Flow ................. . ............................ Molecular Flow ........ ...... OOOOOOOOOOOOO The New Cell Conductance Analysis . ............ . ........... Collision Cell Designs ACo-on Cell Design.... ..... .................. . ....... A Short Collision Cell ........... Theplmtin‘cell ......OOOOOOIOOO The Molecular Deal Cell (Dynasic) ......................... A Sisple Needle Cell ... .............................. . A Modified Needle Cell (This work) ....................... . Cell Construction (A and B Cells) ... .................. . The New Cell Construction (Cell C) ........ . .......... .. iv .0 Instrulent Modifications ..................................... 87 Target Gas and Cell Support ............................... 87 Vacuul Modifications ...................................... 91 Collision Cell Differential Pusping Addition .. ......... 91 Direct Probe Modification . ................................ 93 Softknobs Interface . ...................................... 95 References ................................................... 98 CHAPTER V. PERFORMANCE . ................................... ..... 99 Instrusent Operation ......................................... 99 A CAD Experisent ........................ ................. 100 Cospnter Acquisition of Daughter Spectra ............... 102 Efficiency Calculation ....................................... 103 Collision Cell Efficiency Determination ...................... 105 Cell A.With Heliul ........................... ..... ........ 105 Cell 9 With Helius ............................. .... ....... 111 Cell C Nith Helius .... .................................... 113 The Nature of the Target Gas ...................... . .......... 118 Target Gas Conparison ..................................... 121 Nitrogen ...................................... . ........ 121 Deuteriul .............................................. 121 Cell A .............................................. 122 Cell 9 ..................................... ......... 122 Cell C .............................................. 12? Pressure Effects .............................. . .............. 127 Target Gas Dependence .................................... . 129 IBM Analyses ......................................... 131 N—butylbenzene ............................................ 131 5-Nonanone ................................................ 136 Phenetole ................................................ . 136 Conclusions .................................. . .............. . 142 References ................................................... 144 CHAPTER V. FUTURE PROSPECTS .................................... 145 Present Probless With Trina .................................. 145 Deal Deflection ........................................... ... 147 Tile-Array Detection ......................................... 148 References ............................................ . ...... 151 LIST OF TABLES 5.1 Peak widths in tise deternined by tile—sweeps of the H-t data field ..................................... ..... 107 5. 2 Pressure changes within the systes as the target gas pressure is increased ................................ 128 5. 3 Intensities of the two peaks observed for CAD of 5-nonanone ............... ..... ............ ............ 137 vi 1.1 1.2 1.3 1.4 2.1 3.1 3.2 3.3 3.4 LIST OF PI“ An MS/MS spectrus is typically obtained by fragrenting a selected ion and collecting the resulting ion current. In this way a daughter sass spectru- is produced, providing another di-ension of inforsation ......... . . . 6 56/16 can provide an added diaension of inforsetion pro- ducing a 3-disensiona1 ”fingerprint” of the coupound. The complete MS/MS data field is shown for the pro- tonated solecular ion of dinethyl sorpholino phos- phoresidate (MA) . . . . . . . ............................. 7 In an MS/FB instmnent, two stages of sass separation Ion are emloyed. A sequential aanner of analysis with ion forsation, parent ion selection, parent ion dissociation and daughter ion selection followed by ion detection is achieved, yielding either additional structural inforaation or aixture comonent identification . ................. ... ...... ............ 8 separation by aoaentus and velocity in TRIMS. Daughter ions fro. the sale parent appear at the some arrival tine but are dispersed accordingly to their soaenta. Stable ions have shorter arrival tines than daughter ions of the sme soaentun because of their greater initial velocity ............................... 10 Schenatic diagraa illustrating the cosponents of a MIXES instrusent ............. . ....................... 16 Energetics of collision-induced dissociation. The energy of excitation 0 represents the endothersicity of the ion-aolecule reaction and T represents the kinetic energy release accoapanying fragsentation .......... . . . . 26 Illustration of the types of seasureaents possible for CID peaks. In the absence of a kinetic energy loss the average translational energy of la is (Ia/I1) that ofm+ ............................ ...... 29 Origin of the kinetic energy release (T= T' + accoapanying dissociation after a collision . . . . . . ..... 37 Energy versus ion—target distance R for ground (l)and excited (2) states showing adiabatic (dashed plus solid line) and diabetic (solid line) state behavior .. 43 vii 4.1 4.2 4.13 4.14 4.15 4.16 4.17 4.18 4.19 4.20 4.21 4.22 4.23 4 24 A block diagran of the data acquisition systes and control hardware for TRIMS .1.... ............................ .. 49 H-t diagran for several types of B—TOF scans. The scan represented are a) stable ion scan; b-d) parent ion scans (constant daughter) for daughter ions 100 asu, 144 asu, and 196 see, respectively, e) daughter ion scan for parent ion 400 sun [1] ...................... 51 The conductance-pressure relationship defining gas flow regions........................ ...... 55 A diagras illustrating thee various pressure and con- ductance regions of cell C ...................... . ..... 59 A circuit representation of Figure 4.4 .................... 61 A redrawing of Figure 4.5 indicating the conductance values ......... ...... ............ ... ....... . .......... 61 A collision cell of early design for target gas con- fine-ant .............................................. 65 A short collision chalber placed at the focal point of the electric sector and incorporating the source slit [7] .............................................. 67 A diagras of the floating all to discrininate against . fragmentation occurring outside the cell [9] ... ....... 71 A diagran of a solecular been collision cell [10] . ........ 72 A diagram of a.sinp1e needle target gas inlet systes ...... 74 A diagraa illustrating the principal coaponents of cells A and B described in the text. Slit disensions in cell A were 9.5 as x 1.5 In; those in cell 8 were 8.73 II x 0.95 .- ..................................... 76 A cross-section through the upper portion of cell C drawn to scale ............................... . ....... 79 A cross-section through the lower portion of cell C drawn to scale ........................................ 80 Another view of the upper portion of cell C (not to scale ................................................. 81 Another view of the lower portion of cell C (not to scale .1 ............................................... 82 A sore detailed drawing illustrating the slit apertures and target gas inlet ....... . .......................... 83 A diagran of the top view of the lower portion of cell C .. 84 An andron view of the top portion of cell C ............... 85 Vacuus and inlet plusbing connections for the various cells used ..................................... .....a 88 Collision cell flange positioning on the analyzer ......... 90 The added diffusion pulp for the collison cell assesbly ... 92 Direct Probe inlet sodification ................... . ...... 94 Schesetic of the softknobs interface board ................ 96 viii (nous “NH QQGOI-b a 999 m “a m HQ O ...: H 5.12 5.13 5.14 5.15 5.16 5.17 Frag-entations of the methane molecular ion . . . . . . . . . ..... . 106 A raw data sweep of methane molecular ion CAD .......... ... 109 Efficiency curve for methane relative to pressure. The pressure at approximately 753 attenuation is 5 x 10'3 ............... ............................ 110 Collection, fragmentation and overall CAD efficiency curves for the nethane aolecular ion ....... .... ...... 112 Efficiency curve for sethane CAD employing Cell 3 ......... 114 Efficiency curve of methane CAD showing the improved values elploying the newly designed cell ............. 116 A raw data plot showing the daughters of methane observed when using Cell C ....... . ............................ 117 Methane CAD with deuterium target gas and Cell A .......... 124 Methane CAD with deuterium target gas and Cell E .......... 125 Fragsentation, collection and overall CAD efficiency curves for Cell 9 with deuterium as the target gas ... 126 A linear relationship is shown for the sethane- experiments indicating single collisions were maintained ... ....................................... 130 Raw data plot indicating the 92 metastable deco-position product of the molecular ion of n-butylbenzene ....... 133 Raw data plots indicating the progressive increase in the 91 daughter of the n-butylbenzene solecular ion with collision gas pressure ............... . ...... 134 A plot of the 91/92 peak height ratios ...... . ........ ..m.. 134 Fragsentat ion sch-s for 5-nonanone ....................... 138 Daughter scans of m/z 122 and 94 parents with helium as the target gas ....................................... 139 Partial fragmentation scheae for phenetole ...... . ......... 141 ix A remarkable new instruent has been developed, providing the capability for nus spectrometry/sass spectrometry (BB/DB) [1]. The instruaent aploys ion source pulsing with time resolved detection through a aagnetic sector; the technique has been coined, "tins—resolved ion muentt- spectrmstry,” (TRDB). Ions that frag-ant in the field-free region between the source and the ngnetic sector nintain the velocity of the parent, but due to the loss of use, have a lower aomenttl. At the setting of the ngnetic field at which they appear, they have a longer flight tine than ions of the em moaentu which had not fragmented. Thus flight tine resolution is file to separate all fragmentation products from stdzle ions a well as provide clear identification of the parent and daughter ion relationships. Compared to sore conventional ABM techniques, TRIIB offers several advantages. However, it is also plagued with several limitations. One of the limitations will be addressed in this thesis, n-ely, the lack of an operationally adequate collision cell. 2 Presently, the experimentation to test the theory of the technique and evaluate the performance of the instn-ent has been done ninly with n—dscane [2]. This cupound exhibits a large mier of met-tfile peeks but only a few were used in the experiments. Fraasntation of n-decane through metastable decuposition in the field free region preceding the magnet allowed for ABM data acquisition. Unfortunately, this configuration is useful only for cupounds that yield a large under of metastable decoqositions. In order to extend the usefulness, of the technique, the addition of a collision cell is necessary to enhance dissociations in the field free region. The goal of the present research is to construct and inlaent a nodular collision cell for the TR“ instrinent to provide an adequate collisionally activated deer-position (CAD) efficiency and also solve the difficulties associated with the gm load. Collisional activation efficiency in high energy instruents has long been and still is a major hurdle to the ABM technique [3,4]. Scattering of the daughter ions from the .gnet’s line of focus results in significant losses in sensitivity. The sensitivity is already less than that found in the conventional non-pulsed LEE-9000 mass spectrometer by the factor of the ion source duty cycle (the ratio of the time the ion be- is "an” to the pulsing period). It is therefore inortmt to maximise the collisional efficiency and reduce the scattering of fraaentation products as mach as possible to gain 3 the full advantage of the multi-dimensional data provided by the TRDB technique. Another problem inherent in the LEE-9000 is its lack of a differentially M ion source. This necessitates the development of a differ-mitially pimped collision cell in order to circI-vent the gas load difficulties. Gas escaping to other regions of the mass spectrometer increases the probability for scattering of parent ions as well as daughter ions thereby affecting the 3/2: ratio dramatically. Pressure plays an iaportant role in the CAD process and is a par-eter which mist be controlled for efficiency determinations. Therefore, the pining and delivery rate of the target g- mst be adequate to similtaneously produce .7. high pressure collision cell region and a low pressure throughout the r-ainder of the nos spectrueter. This involves ming speed and conductance considerations that not only influence the overall exhaust rate for a particular gas but also the final cell efficiency. This thesis is chiefly concerned with the design md performance of a collision cell for TRIBE. The birth of CAD, the history of its ' deveth and theory as it is presntly understood, is discussed to provide a foundation for understanding of the CAD process. Various collision cell designs including that of the new cell are presented. Collision cell performance is cowered .ong the cells with respect to efficiency. Also, ABM analyses are illustrated with an Qhasis on 4 estfilished syste- for qualitative cQarisons of CAD data. Finally, the future prospects for TRDB is presented in light of a rapid scanning data acquisition schae for gas chrcmatou-aphy mass spectrometry (GO/IS) . INA-3mm” The TRIBE technique shares my of the attributes cm to those of conventional BE/BE techniques. Nith BE/BE it is possible'to generate daughter mass spectra for each ion in a stfile ion spectri- [5—9]. An eque of this is shown in Figure 1.1 [5]. These daughter spectra can be used to elucidate the structure of a particular parent ion .and thereby provide an added diaension of information (see Figure 1.2) [10]. BE/BE also has some features analogous to a separation technique, similar to those of gas chromatography in GC/BE. One emloys the first BE to separate a mixture of ions according to mess, aid the second BE to identify each coqonent. Figure 1.3 illustrates a sass separation sch-e for structural elucidation and mixture analysis achieved by BE/BE. It is not surprising then, that tandem mass spectraetry has bears a powerful analytical technique having a wide variety of ch-ical, biocbaical, and environmental applications [ll-18]. TRIBE differs in many respects from the norm and therefore offers the potential for a praising new analytical approach to BE/IE. Conventional BE/BE instruments usually consist of a series of two or more mass selective devices in tend. with one or more collision chdaers. An alternative approach is equified with TRIBE. A fit" ' :: MASS INSPIRE!“ .— IMIZATIM -78 MASS SPECTRUM MS/MS SPECTRUM Figure 1.1. An BE/BE spectrtm is typically obtained by frag-eating a selected ion and collecting the resulting ion current. In this way a daughter mass spectm is produced, providing another dimension of infomtim. Reprinted from Cooks, R.G.; Glish, G.I.. Chm. and Eng. News 1981, 30, 40. ‘ I I x SPECTRUM OF PARENTS CID I SPECIRUM (x DAUGHIENS YIELDING all! as FROM nasmluul' Figure 1.2. MS/MS can provide an added dimension of information producing a 3-dimensional ”fingerprint" of the compound. The mlete MS/BE data field is shown for the protonated molecular ion of dimethylmorpholi L L '4-‘- (MA). Reprinted fru Dawson, P.H.; French, 3.3.; Buckley, J.A.; Douglas, D.J.; Si-ons, D. Org. Mass Spectre... 1982, 17, 212-219. 63:03:63: use-sonic“. 0.55.:- 90 sowuoluowsm Hon—525: ...-confines .553 2:33» .3322. S 83033 :3 3 38:8 .838?- sow evacuees use sogasfiooeemv no“ «sense .sowuuouoe new unease Sodas-Lou so.“ :33 32323 no access demusgvoe e 6231.! one souasusaoe see! he senses o3» Jana—Cues: gm: so an .MA 0.5m: ; -mem>exmm annexe: - - A)nwAfl_..n.u.lw\IVH«1/Mummu.zwamn~.z..o.nlzt\rl_t.m.fl.._ ”mnuhALz ...-.o; , . 1Io_ ~ zo_~uueuo zoap_oined with the aforoentioned propertiee the poeeibility for extracting eore infomtion fre- collieional activated demeitiee etill exiete. Surface-induced dieeociation (811)) ie a relatively new technique whoee infor-tion content ie etill euewhat unexplored. 10. 11. 12. l3. 14. 15. 16. 17. 18. 19. Tho-on, J.J. laye of Poeitive Electricity and Their Application to Chaical Analyeie, Long-one, Green, London, 1913, p. 94. Coob, 8.0. Collieion Spectroecopy, New York, 1978. Aetee, 8.". Proc. C‘ridge Phil. Soc. 1919, 19, 317. Swth, 8.1). Phye. Rev. 1926, 26, 462. httauch J.; Lichtbleu, 8. Phye. Zeit. 1939, 40, 16. 8ipple, J.A.; Pox, 8.8.; Condon, 8.0. Phye. Rev. 1946, 68, 347. 8oeenetock, 8.11.; Melton, 0.8. J. Ch. the. l$7, as, 314. 8npriyznov. 8.8.; Perov, A.A. Rune, J. Of Phye. Chen. 136, 39, 871. Jenninge, 8.8. Int. J. bee Spectroa. Ion Phye. 138, l, 227. McLefferty, 52".; Schudduage. 8.0.8. J. Ae..Ch¢I. Soc. 1%9, 91, 1%6. Van de Sande, C.C.; McLafferty, 8.". J. An. Chen. Soc. 1975, 97, 4617. Van de Sande, C.C.; McLafferty, 8.8. J. An. Chen. Soc. 1975, 97, 4613. Van de Sande, C.C.; McLafferty, PM. J. An. Ch- Soc. 1975, 97, 22%. Leveen, 8. Org. Maee Spectroe. 1975, 10, 43-54. Beynon, J.8.; Cooke, 8.0. Bee. Develop. 1971, 22, 26. Beynon, J.8.; Cooke, 8.8.; An, J.8.; Baitinger, 8.8.; Ridley, T.Y. Anal. Chen. 1973, 46, 1023A. finch, T.; Dante, P.F. III; McLafferty, 17.“. Int. J. Knee Spectroe. Ion Phye. 11972, 9, 333. Cooke, R.G.; Hendricke, L.; Beynon, J.8. Org. Hue Spectra. 1975, 10, 625-638. lie, 8.0.; “clatter, 11,; Beynon, J.8.; Cooke, 8.8. Int. J. he Spectroe. Ion Phye. 1974, 16, 23-36. 20. 21. 22. 31. 32. 21 Cooke, 3.0.; Kin, x.c.; Beynon, J.8. Int. J. an. Spectra. Ion am. 1974, 15, 245-254. Terwilliger, D.T.; Cooke, 8.0.; Deynon, J.8. Int. J. Mae Spectra. Ion Phye. 1975, 18, 43-56. Cooke, R.G.; Doynon, J.8.; Litton, J.8. Org. the epectra. 1975, lo, 503-506e Cooke, 8.0.; Aet, T.; Deynon, J.8. Int. J. thee Spectra. Ion Phye. 1975, 16, 348-352. leough, T.; Deynon, J.8.; Cooke, 8.0. Int. J. thee Spectra. Ion Phye. 1975, 16, 417-429. Terwilliger, 0.'l‘.; Elder, J.8. Jr.; Beynon, J.8.; Cooke, R.G. Int. J. thee Spectra. Ion Phye. 1975, 16, 225-242. Dayna, J.8.; Aet, T.; W. T.; Cooke, 8.0. Int. J. Mae Spectra. Ion Phye. 1975, 16, 343-347. McLafferty, F. 8.; Kornfeld 8.; Haddon, 11.0.; Leveen, 8.; Sakai, I.; Dante, P.F.III; Teai, Shie-Chuan; Shuddaage, 8.0.8. J. An. Chen. Soc. 1973, $, 388. Cooke. 8.0. Collieion Spctroecopy, New York 1978. Dayna, J.8.; Cooke, 8.0. J. of Phye. 8 '1974, 7, 10-18. McLllckey, S.A.; Oleverkerk, 0.8.0.; Boerboa, A.J.8.; Kieteeaker, P.0. Int. J. Meee Spectra. Ion Phye. 1&4, 59, 85-101. Dace, 8.8.; fulford, J.8. Int. J. bee Spectra. Ion Phye. 1&2, 42, 195-211. Daeon, P.8.; Sun, Wing-fung Int. J. Maee Spectra. Ion Phye. 1m, 44, 51-59. Daon, P.8. Int. J. the Spectra. Ion Phye. 1983, 50, 287-297. Yoet, 8.A.; Bake, 0.0.; lchilvery, 0.0.; Seith, 0.; Morrieon, J.0. Int. J. Maee Spectra. Ion Phye. 1979, 30, 127-136. Yoet, 8.A.; 8nke, 0.0. Anal. Cha. 1979, 51, 1251A. mm III CAD neon! "848100411“ Briefly, the CAD proceee can be deecribed a the interaction of an ion with a neutral target epeciee (anally a gee) reeulting in the albeequent frageentation of the ion. The baeic reaction ie ehoun in mtion (l). alt ) er + as (3a) where av ie the parent or reactant ion, 8 ie the neutral target, er ie the aughter or product and a: ie the reeulting neutral (repent. Typically, al’ ie for-ad in the ion eource of a aaee epectraeter and ie accelerated out of the eource to a conetmt knotn energy. In the reactione concerned here, the accelerating energiee are in the kilovolt range and it ie the high energy collieione which will be of concern. Eighenergyione antertheanalyaerregionofthenaee epectraeter with a dietribution of internal energiee ad encounter a region of higher ataic mar deuity relative to the reet of the 22 23 inetraat. The high alder deneity ie due to the addition of a target gee to a epacific preeeure. Thie region ia the collieion region and it ie here that all the iaortant evente occur for CAD. In thie region the ion acquiree an aunt of internal energy which will detereine mother or not it will {repent into a detectale, characterietic, ad lower-aaee ion. Accaanying thie frmentation proceee ie the releeee of internal energy ae the relative tranelational energy of the frmente. The daughter ione are then detected at an apparent aaee very cloeely reeealing that which would be obtained by ealoying the equation deecribed earlier for aetaetable decaoeitione [1]. The fratetion of the high relative velocity ion into a neutral ad a ion of lower aaee providea a "nee epectrtn” of the parent. ion. The added dieeneion of inforeatia ie what ie taken advatage of with CAD. Thie ie the eialifiad picture of what occure in CA0. Many reeearchere [2] have inveetigated the proceee thoroughly and their reeulte ad ineighte are preeented. A mm wine: flow that a eketch hae bea eade concerning the CAD proceee, it ie tine to ineert the detaile. Sae aeemtioa are ialicit in the iaae preeented. Firet of all, the proceeeee that occur at high relative energiee involve einiaal aaenta tranefer with eaell 24 anglee of ecattering. Then, too, the interaction between the electrone of t1. ion ad target occure without affecting the relative notion of the ion ad target except ae the coneervation lava dictate. Energy coneervation daende that appropriate adjuetaenta in traelational aergy occur. Thie energy change ie expreeeed aleoet entirely in the kinetic aergy of the feet epeciee. The relative traelational aergy of the colliding particlee ie both the eource of the energy of reaction ad a eignificat deter-inat of ite croee eection. The baic reaction (3a) above ie revieed now to include the (oration of a traeiat calex, an“. Reactione (& ad 3c) daecribe the achaiaa: )a".+ N (31)) el" > eat + a: (3c) The overall reaction can be broken into two etepe. Reaction (3b) deplete the collieional activation etep. Reaction (3c) deplete the unienlecular deco-poeition etep. The aechaniee of collieional activation can be thought to proceed via a traneient coeplex in which electronic excitation occure. Electronic excitation ie ueually followed by rapid intraeelecular energy tranefer to leave the ion vibrationally excited [3]. The target can be thought of ae a collection of electrone available for inelaetic collieione with the electrone of the ion. The generally accepted procedure for deecribing what occure in the collieion region ie to divide the entire proceee into two eeparate proceeeee ae etated above. The firet ie called bianlecular excitation which in then followed by the eecond proceee, uniaelecular fragaentation. Although collieional activation ie neceeaarily endotheraic ad fraaentation ie neceeaarily exotheraic the overall reaction aay be either. Tranelational energy auet be converted into internal energy in the firet proceee and than the internal energy of the ion euet be releeeed ae relative tranelational energy of aeperation of the frag-ante in the uni-olecular deconpoeition etep. Thie ie ehown in Figure 3.1 below [4]. In addition to the fact that can providee an added diaeneion of infbraation there are aleo phyeical aeeeureaente that can be aade to aid the inveetigation of the proceee iteelf. KINETIC INHIIY’IOBS One of the aeaeureahle quantitiee with.CAD ie the difference between the tranelational energy (E) of the reactant ion and that of the product ion. 26 \ I T x / m2+‘"‘3*N (final stole) Potential Energy / _L_ ml’+N (inniol state) F... lignre 3.1. Energetice of collieion-induced dieeociation. The energy of excitation O repreeenta the endotheraicity of the ion-aolecule reaction and T repreeente the kinetic energy releaee accoapanying fragaentation. Reprinted fra Cooke, 8.0. Collieion Spectroecopy, New York, 1978. 27 O = 8 (parent) - E (daughter) = E (loee) (3d) If the kinetic energy of the daughter ion ie corrected for the aee aeeociated with the neutral frag-ant“), then the difference between the reactant ad product kinetic energiee ia the kinetic aergy loee. The value of O repreeente the endotheraicity of the reaction. In general, for polyataic ione, the reactant ion baa will include ione in aeveral vibrational ad perhape ova electronic etatee. . Thie eituation will yield a dietribution of kinetic aergy loeeee in the activation etep. the nature of which will depend on the reaction at had. The kinetic aergy of the daughter ion ie aaeured relative to ite expected value in the abeence of energy loee or gain. For the reaction ehown in equation (3a) above, thie aergy ie all: of the energy of the parent. Inn-lemmas: Another quantity of intereet that can be deterained with CAD ie the traelational energy,T, releaeed in the fragaentation. Thie releaeed energy will reflect the daughter ion tranelational energy dietribution ad reeulte when a 'ieolated ion, aoving at high velocity, 28 frageente. Frageentation ie ieotropic in the center-of-aee coordinate eyeta ad the aergy releaee ie aalified on converting to laboratory coordinatea [4] . For polyataic ione the intraolecular energy traefer and dietribution require are tia than a eingle rotational period, ad therefore a ieotropic dietribution of fraaent ione reeulte. Thie eeene that an increeeed broadening of the -’ energy dietribution ie obeerved relative to that for al’ ad therefore all center-of-aee energy releaeee ca be aaeured. The kinetic energy releaee, T, ie a function of the obeerved energy dietribution ae given by: T = alsza -/ lfiaaE (3a) where al, a2, and a3 are the parent, daughter, and neutral aeeee, respectively, E in the energy of the etable ione, and 48 ie the corrected energy dietribution of the daughter. ione [5,6]. 8.0. Cooke adde, " an extra calication of 010 la that two dietributione, one due to energy releaee ad the other aeeociated with energy loee, are convoluted together in the experiaental ”all ehape. " [4] ' A diagra of the different typee of aaaureaente that ca be ade on a 010 peak ie ehown in Figure 3.2. 29 fD , - , reloflve ctoss section r/ .4, a. ‘ \ \ peou height) / C 3‘ T, kinetic energy release //' -. (peak width) _, '1 ‘/_;.,...—- O, lunatic energy loss /’ ,,/ ‘\ (peat DOSINOI‘) LL. \— -..—a Kinetlc Energy Figure 3.2. Illuetration of the typee of aaeurenente poeeible for CID peake. In the abeence of a kinetic energy loee the average tranelational energy of a' la (ea/ll) that of al*. Reprinted fra Cooke, R.G. Collieion Spectracopy, New York, 1978. 30 0mm In addition to the kinetic aergy loee ad kinetic aergy releaee aeauraente poeeible with CAD, the etructural elucidation value of the technique ie very iaortant. The energy coneiderationa aid in the underetading of the phyeical chaietry involved but the real pay off, fra an analytical point of view, ie in the ability of CAD to fragaent ione further and thereby help to fit together the puzzle piecea repreeented by an unknom aolecule’e fraaente. The aalytical applicaility of CAD for 18/58, eepecially in etructural probla, ie beaed on the principle obeerved by McLafferty et al [7] , whereby the relative aundance of a CAD fraaent forad via a high energy proceee ie deterained by the etructure of the precuraor ion, not ite internal energy before collieion. Nith CAD, each particular ion can be fragaented an‘d thie inforaation (with the help of a eubetructure library) [8,9] can lead to an unaiguoue identification. Decauee of thie ability, it ia iaportat to inveetigate the kinetice of the collieion not only before and after the cell but, aleo during paaeage through it. An analyeie of thie nature will provide a are coaplete picture of the cell iaortance in relation to the CA0 proceee and therefore be of a direct benefit to thie theeie. 31 nears name! 1'- WLLISIN CELL If we*were to take three enapehote of the CAD proceee the firet one would be approxiately fra 10" to 10“ eeconde after the ionization event. Typically the ionization-fragaentation tiae ie alecule—dependent, but for‘purpoeee here, let ue eeeae that in one aicroeecond the ion or ione we‘wieh to frag-ant further have been aade. The ione preeent at thie point in tiae are baeically of two typee: (a) etable ione which, in the abeence of collieion gee, will reach the detector and be recorded, and (b) eetaetable ione, the internal energiee and ratee of frag-entation of’whichwwill, in the abeence of collieion gee, cauee thee to frag-ant in the field-free region or in later parte of the inetruaent. Netaetable ion abundancee are typically a few ordere of eagnitude lower than the signal due to etable ione [10.11]. Becauee of thie low abundance, it ie probable that eeeentially all collieion-induced dieeociation ie a reeult of excitation of ione which are etable to uniaolecular fragaentation. Stable ione generally have energiee below the critical (activation) energy for the loweet-energy reaction. Therefore, we can aeeuae that the reactant or parent ion bea ia med ainly of stable ione in aeveral vibrational and perhape even electronic etatee prior to collieional activation. 32 ammonium The eecond enapehot would have to occur 10” to 10’13 eeconde after a collieion with the target in the cell. Thie tia would approxiately encaaee the excitation or activation tia. Thie enapehot would yield a picture of the biaolecular excitation etep deecribed earlier. Aeide froa the tia apent in the cell, the internal energy acquired in the collieion will govern eubeequent evente. Thie acquired energy ie not eaeily acceeeible but it in known that energiee in the range 1-10 eV are repreeented [12]. The reeult of the. ‘excitation in the for-etion of the traneient ceaplex diecueeed earlier. An electronic reorganization occure which can lead, ea the epeciee eeperate, to regeneration of reactante or on to producte. It ie in .thie reaction that energy euet be coneerved and therefore adJueteente are aade to the available'eaergy. Thie energy change ie expreeeed aleoet entirely in the kinetic energy of the feet epeciee. It ie iaortat to nation that the entire diecueeion thue far hae been concerned with reeulte obtained with zero angle ecattering. Cooke et a1 [13-16] have etudied high energy ion alecule reactione at nonzero ecattering aglee. Kinetic energy loee epectra can be deterained for particular ecattering aglee ad reaction croee eectiona for particular eleaentary reatione can be followed a a faction of ecattering angle. 33 Sm WATIN (Dnialecnlar Decaoeition) The fragaentation event ie the neat crucial for eucceeeful 0A0. Our enapehot in thie caee will be 10" or 10" eeconde after the bialecular acitation aking the calete rage of ion lifetiae ealed to be free 10“3 to 10"5 eeconde. In general polyataic ione will have a relatively long interaction tia (due to their elower velocitiee) a coaered to a vibrational period. Polyataic ione dieeociate fra the ground electronic etate after eufficient vibrational quanta becae localized in the bond being broken [3]. For kilovolt beae the aeparation of the activated calex fra the neutral target ie Juetified except for polyataic targete whereby the influence of the target on the. fragaentation proceee auet be coneidered. In the reactione etudied in thie theeie. only all diataice or eonatoaic gaee were ueed ae targete. Decauee the eize of the target ie eaell, the high velocity excited ion can eeeentially behave ae an ieolated eyetea. If the etatietical theory in aeeuaed ad coaariaone are node with the large exieting aee apectral data baeee, it ca be predicted that a "network of caeting ad coneecutive uniaolecular reactione will follow collieional activation." [2] Deceuee of the relatively large increaee in internal aergy, the coaetition between reaction pattevaye i‘e controlled largely by the frequency factor rather than the critical energy ad therefore there are likely to be an priary reaction pathe for highly excited ione tha for thoee of lower energy. 34 The priary path are thoee of eiale cleavage ad thie fact hae been applied in aeveral etudiee utilizing CAD [l7-20]. Stadard uniaolecular reaction theory norally ialiee that coalete internal aergy radaization occure in the energized epeciee ad the aethod of foreing the energized epeciee ie only eignificant if it producee the characterietic internal aergy dietribution of the ione. A carieon of electron-iaact (81) with CAD ca yield qualitative eupport, ad etudying the kinetice of the fragaentation by CAD ca yield quantitative eupport for the aeelation that the quaei- equilibria theory (GET) ie followed. 35 mum Ol' CA1) ‘10 81 One of the fortunate qualitiee of CAD in that the fraaentation to deughtere hae been found to be eiailar to the fragaentation that occure in an El. eource of a nae epectraeter [4]. With aee analyzed ion kinetic energy epectraetry (MIXES), for exaple, caee the ability to eelect the ion aout which are inforation ie deeired, to fregaent that eelected ion in the field-free region preceding a electric aector ad eubeequently to energy aalyze the daughtere. With T8118, the arrival tia of the parent ie equal to that of ite daughter and therefore by aetting the arrival tia equal to that of the parent ion, the daughter aeeee are deterained by ecaning the agnetic ‘ field. - In MIXES it hae been ehown that CAD epectra cloeely reeeale EI epectra (eee chapter 2). The can ie true for TRDB ae well [21], but are work euet be coaleted in thie area. The point ie, if the hypotheeie that aoet 0A0 reactione can be deecribed in tera of etatietical uniaolecular reaction theory ie true, then the aee epectra aeeociated with 81 ad CAD auet be at leaet qualitatively eiailar. Aleo, the fact that the epectra are eiailar both eupporte thie hypotheeie ae well ae providee inforation for etructural elucidation ad aixture aalyeia. 36 One of any reaeone, why EI ad CAD are eiailar ie that a the rage of ion lifetiaee eapled ie ceaared (lo-13 to 10" eecona), the two athode produce approxiately the aae profilee. Ae aentioned above for can, fregaentation any occur 10" or 10"5 eeconde after excitation ad when coained with the tia for excitation to occur, the E1 ad CAO eapled lifetiaa coincide. The depoeition of internal aergy in both 81 ad CAD largely governe the zaeaquent fragaentation ovate. An approach to characterizing the kinetice of fraaentation by CAD involvee aeeuring the kinetic energy loee eo ae to deteraine the internal energy of the frag-eating ion. In a etudy by Cooke et al it wae ehown that the energy loeeee aaeured for particular ione correlate with their internal energiee given in the breakdown curvee. 8e eaye thie ”confirae that ione of the aae internal aergy forad by alecular iaact ad by electron iaact undergo the aae reactione, ad it ie therefore probale that collieional activation reeulte in fragaentation fra the aae etatee ae are involved in ion-eource reactione” [22] . In wring 0A0 with EI, kinetic aergy releaee aaeuraente can be mod to an advantage. The difference between the kinetic energy releaee due to collieionally activated ione ad the kinetic energy releaee due to ataetale ione undergoing the cane reaction will yield the energy partitioning occurring in the highly excited ione. Figure 3.3, taken froe Cooke et al [4] ie than below. The difference betwea the appeerace potential of the [II-81* ion and the ionization 37 Potentnol Energy 4/ Reaction Coordinate Figure 3.3. Origin of the kinetic energy releaee (T=T' + '1‘) accapanying dieeociation after a collieion. Reprinted fra Cooke, R.G. Collieion Spectraecopy, New York, 1978. 38 potential of athanol ie equal to the al- of the critical energy. en, for the forward reaction ad of, the nonfixed energy appropriate to freaatation in the ion eource. The internal energy of the ione undergoing collieion ie aeel-ed to be 0.4 eY, and thue, taken in conjunction with the aaeured energy loee O’ of 1.7 eV fixea the internal energy of the fragaenting ione ee 2.1 eV ad thue their nonfixed energy, at, a approxiately 1.2 eV. The ataetahle ion reaction in athenol ie accoaanied by an average kinetic energy releaee of approxiately 10 a7. The correeponding CAD wee found to releaee 400 aeV. The experiantal reeult ie now directly morale to etatietical calculatione regarding partitioning of the nonfixed aergy of the activated coalex. Thie releaee ie coneiatent with the predictione of etatietical partitioning ad therefore lende are credibility to the fact that collieional activation reeulte in frag-entation .fra the aae etatee a are involved in electron iaect. 39 WWI“ cm mm The extent to which a given proceee occure on collieion ie conventionally expreeeed ae ite croee eection. Thie ie a aeeeure of the probability that interaction will lead to the epecified producte. Croee eectione are giva in unite of area per ata or aolecule. The utility of the CAD technique ie dependent on , the feet that croee eectiane are generally large. A typical value for the croee eection in a eethane reaction whereby ite aolecular ione collide with a target gaa to produce daughtere ie of the order of 10'15 ca? [23]. For a been of inteneity I, paaeing through a gee of nueber deneity n, the decreeent d1 in been atrength ie proportional to the nueher denaity, the dietance d1, and the croee eection 8, ie., -d1 = Infidl OE‘ I/Io = e <-nm (31‘) where la la the inteneity at 1:0 [4]. 40 The firet ad aJor point to be ado ie that there ie a croee eection aeociated with each atop in our two atop proceee. The croee eection for CAD ie not that for a eingle elaentary reaction. Rather, it ie the reeultat of the ratee of an ion-aolecule reaction ad a aialecular dieeociation. Therefore, 6 (CAD) = O (excitation) a O (fragaentation) (3g) with O (fragaentation) being dependent on the internal aergy of the ion after excitation ad on the ratee of ay caetitive franentatione or other energy loee achaniaa. Thie equation ropreeente the dependence of the croee eection for CAD on the probaility of excitation of the ion by a appropriate aount ad aleo the rate of uniaolecular frageontation of the excited ion. For the production of abundat CAO fraaente, it ia deeirale to obtain the largoet croee eection poeeible. A difficulty arieee in that the relative croee eoctione for excitation ad frag-entation for caeee in which the product a’ ie forad will tend to oppoee each other. Thue, the greater the croee eection for excitation, the lower ie the aergy of the collieionally excited ion ad the ealler ite rate of fraaentation. The experiantal reeulte of McLafferty and ca-workere for collieion in the low kilovolt energy region ehow tho probaility that the ion will be excited by 5 all or are in lower than the probaility for 41 eaallor «citatione [12] . The available aaeureaente of the aergy loee for CAD of polyataic ione confira thie concluzion [22]. A well lmown principle which an account for the reeulte above and relate croee aectione to ion energy or velocity ie the adiabatic criterion of Naeeey [24,25]. Thie criterion coneidere three ragee of ion velocity : (a) At low velocity the electrone ca adJ'uzt adiabatically to the perturbation reeulting froa the interaction between ion and target, aking a traeition unlikely; (b) at high velocity the electronic traneition tia ia long capared to the collieion tia, which again akee a traeition unlikely; (c) at interadiate velocitiee the collieion and electronic traneition tiae are'coaarahlo, which axiaizee the traneition probaility. In equation fora, thie criterion ia repreeented aa: v ’5’ a “a / h (3h) where v ie the ion velocity, "E ie the «citation energy (or difference in energy between the two adiaatic energy etatee), a in the adiaatic paraeter (the dietaco over which motion occure—of the order of 7 X ), ad hie Planck’e content. The axil- croee eection ie lean-ed to reeult when the electronic traeition tia ad the collieion tia are equal. In thie treeteent the croee eection ie a function of the velocity of approach of the ion and target. 42 Looking clonely at the theory developed for nono- and diataic ion nynte- one can nee it nu be helpful to explain none phenonena occurring with polyatonicn. Menney’ n explanation of the collieion event in well accepted and in ehown in figure 3.4 below [3]. Shovel are two energy ntaten repreeentative of my ion-target cmlex. Ountl- nechanicn ntatee there in no ion-target dintance at which both ntaten would be repreeented by the n-e energy. The calculated curven would follow the danhed and nolid linen in the figure, thoee being the electronically adifiatic energy leveln of ntaten l and 2. If ion and target were deecribed by ntate l at infinite neparation, and approWd each other infinitely .nlowly, the electron dennity of the nynt- would adJunt adidlatically to naintain the lowent total aergy poeeible. In thin‘cane no curve cronnin¢ and no electronic excitation would occur. If the ion-target dintance decreanen no rapidly that the electrone cannot adJunt adidiatically, the nyntu nay cronn fron etate l to etate 2 following the nolid (didiatic curve) line. The probability of thin curve croeninc in related to the tine apent in the interaction mien, and in thun a function of the radial velocity with which the ion and target nuclei approach each other. If the approach in too rapid, no net excitation will occur. Internediate velocitien are favored for curve cronning with electronic excitation. Evidence nupporting the above nechaninn in found by whim of the croee nectionn for ionn for-ed via CAI) an a function of ion _ kinetic energy [26]. Erl- the adiabatic anti-1n rule, the relative 43 EhERGY {y Rth ' IAROEY Figure 3.4. Energy vernun ion-target dintance R for ground (1) and excited (2) states showing adiabatic (danhed plun nolid line) and diabatic (nolid line) etate behavior. Reprinted fron McLafferty, P.Vl., Tanden Mann Spectronetry, John Niley and Sonn, New York, 1983. 44 Mility for curve cronningn with larger valuee of "B increaee - the ion velocity increanen, no that higher energy frag-ante (ex. 03’ fru nethane frag-notation) nhould be relatively lore intenne with increaning precurnor ion kinetic energy. Thin in found to occur «perinentally [27] . At nufficiently high ion velocitien, all ntatintically nignificant curve cronningn nhould be «haunted and the total CAD cronn eection nhould decreane. Thin in obnerved for GB.” of kinetic energiee >50 lav [27], well have the energiee of neat BB/DB «perinantn. If cronn-nectionn are neenured by neann of equation (31") the reeulte are likely to be low. Becaune of the two-ntep nature of the reaction the direction of nation of the product will be controlled by both the ecattering angle upon collinion end the angular propertien of the frag-antation etep. Thin equation annuen that the fraaent ionn are collected with the nane efficiency u the reactant ione and it in not apparent that thin in true. Although the collieion can be confined to the collinion chater, the nubnequent fragnentation can occur at any elbnequent point. Neverthelenn, relative cronn eection for the overall CAD nequence nhow dependencen upon reaction channel and «perinental variablen and add to the underntanding of the proceee. Bxperinent varifilee (target preenure, target identity etc.) which effect the cronn eection, and therefore CAD, will be dincunned later in thin thenin in light of the reeulte obtained. 2. 3. 9. 10. 11. 12. 13. 14. 15. 16. 17. m Cooke, 8.0.; Reyna, J.8.; Caprioli, 8.51.; hater, 0.8. Metenteble Iona, Blnevier, Aaterda, 1973. ‘ Levnen, K. Fundaental Anpectn of Organic Mann Spectraetry, Verleg Chaie, New York, 1978. “cufferty PM. , Tenda Mann Spectraetry, John fliley and Sonn, New York, 1&3. Cooke, 8.0. Collieion Spectroncopy, Na York, 1978. Beynon, J.8.; Caprioli, 8.8.; Baitinger, 8.8.; Any, J.8. Org. Mae Spectra. 1970, 3, 661. Selma, J.I..; Terlolnv, J.8. Org. Han Spectra. 1m, 15, 383-33. Manfferty, PAL; Hirota, A.; Berbala, M.P.; Peguen, 8.!‘. Int. J. than Spectra. Ion Phye. 1980, 35, 299-303. Crone, 8.8.; Bake, 0.0., preeented at the 32nd Annual Conference on Mann Spectraetry and Allied Topicn, San Antonio, '15:, May 27-June 1, 1M. Giordani, A.B.; Gregg, 8.8.; Hoff-an, P.A.; Crone, 8.8.; Beckner, C.l'.; Bnke, C.G., prenented at the 32nd Annual Conference on the Spectraetry and Allied Topicn, San Antonio, '15:, thy 27-June 1, 1m. McLefferty, F.W.; Fail-weather, 8.8. J. An. Chen. Soc. 1%, so, 5915. We "are; WffOrty, Fawn m1. Ch-e 1m. ‘1’ 31’36. McLafferty, 3.70.; "ache, '1'. Int. J. then Spectra. Ion Phyn. 1977, 23, 243-247. .Ant, 1.; Terwilliger, n.'r.; Cooke, 8.0.; neynon, 3.3. J. Phyn. Cha. 1975, 79, 708. Pranchetti, 7.; Car-Ody, J.J.; Erma, D.A.; Cooke, 8.0. Int. J. than Spectra. Ion Phyn. 1978, 26, 353-358. Laraee, J.A.; Carnody, J.J.; Cooke, 8.0. Int. J. than Spectra. I” Me 1979, 31. 333.—me lhnon, 8.8.; Fez-node, M.J.; Jeaingn, 8.8.; Cooke, 8.0. Int. J. thee Spectra. Ion Phyn. 1982, 43, 327-330. Yaaoka, 8.; Dong, P.; Durup, J. J. Cha. Phyn. 1969, 51, 3465. 18. ‘ lee 21. 22. 23. 24. 26. 46 Howe, I.; McLafferty, PM. J. An. Chan Soc. 1971, 93, 99-105. lie, M.S.; McLafferty, 3.8. J. Phyn. Chen. 1978, 82, 501. Ill, “.8.; 0m, 8.0.; McLafferty, 7.". J. Al. Cha. Soc. 1978, 100, 4600. Stultn, J.‘1'.; Myerholtz, C.A.; Newcae, 8.8.; Rake, C.C.; Ballad, JJ'. Rev. Sci. Inntra., in prenn. Cooke, R.G.; Hendrickn, I..; Beynon, J.8. Org. Mann Spectra. 1975, 10, 625—638. Ronantock, 8.M.; Melton, 0.8. J. Chen. Phyn. 1%7, 26, 314. hnney, 11. 8.". Rep. Prog. Phyn. 1949, 12, 248. McLefferty, 7.11.; lie, ”.8. J. An. Chen. Soc. 1978, 100, 3279. McLuckey, S.A.; Olnverkerk, 0.8.0.; Boerboa, A.J.8.; Kintaaker, 8.0. Int. J. Mane Spectra. Ion Phyn. 1%4, 59, 85-101. Va de Sade, C.C.; McLafferty, PM. J. An. Cha. Soc. 1975, 97, 4613. motion All GILL 0381“ The inntrunent on which all of the experinente in thin thenin were done wen developed by J.T. Stultn, 0.0. Bake and J.F. Holland [1]. It in a nodified LIB-9000 aaee npectroneter with a 60 negnetic nector of 0.2 n rediun and an electron inpect ion eource operated at an accelerating voltage of 3500 V. An nentioned in chapter one, T8116 in an BM technique. Tine-of-flight (TOP) manure-ate are utilized in addition to nagnetic dinpereion of the ione. The TOP niaal ban a repetition rate of 10 k8! and requiren 10 an renolution over the range of 1-50 an to enconpenn a nann range of 1-800 daltonn. Tine-elice detection in qloyed. Thie in a tined alitude neanurenant whereby the’alplitude of the ion nignel in neenured during a narrow tine-window at a npecific delay tine after the ntert nignel. 8y nonotonically increaing the delay tine after nuccennive ionisation/acceleration puleen, a epectra ca be acquired [2]. 47 48 7b.! ecaning, only a eingle tia alien in necennery at each agnetic field netting. A boxcar integrator [3] controln the naled arrival tia ad providee nignal averaging for the low-level nignele that are often obeerved in IBM «marinate. A nicrocauter baned on the Intel @88- nicroproceneor providee control of both the boxcar integrator ad the agnetic field ntragth an well 8 ion eource pulning. In Figure 4.1 a block diegren of the general featuren of the T8138 nynta in illuntrated. Bach cannat in deecribed are fully in reference 1 . The conputer laguage adopted for the T8116 nynta in FORTH. Thin laguage allown new "wordn" to be added to the language that perfor- functionn npecific to the inntraent. Thene npecific functionn ca be linked to perfor- a tank ainly by conntructing another word that in ade up of any other wordn already in the F0811! dictionary. For the CAD efficiacy calculatione ad peak integrationn, a routine wan written that aloyn the 8087 co-procennor for high npeed floating point naing. A routine for naving date to a 8-inch floppy wen aleo written. The latter wan done to provide a additional back-up aclainn. Sweep data (ra data over a particular raga-peak finding in not ealoyed) connaen are of the availale recordn in a file tha nca date (peek finding in perforad ad connequently there are aee 49 .25. now 03.3: ”chance can loaehe nowuwewscoo nan—o 23 mo loaves moods < .ué 0.58: 3:23“— nesnefi sateen _ . a 33.52.. . 1:3.- Efi: . :22. e3. 1:. 39:: no» . no: ... no. no. r M w 4. 28382.3 ? M _ M H U m _ H E . :3. ione: 393:8 iiiiiii no: ...... . . z 3332...... M W M w M W M w M w .... K0040 0‘. Ox. unlinked“... pun-M. we: .5..th .6149... new: A N L W {5&4 W 58;: M 5.16528 22:332. . Sud x :38 one .3» snug—.5: A1 3.3 m anemoneflum 4:65 :9... 3...: e VJ: E sen. Cahnutu" >9_ » c. up: _¢u:.r!c 84:..— q. IIIII I IIIIIIIIIIIIII . lllll I n . A _ . .3533 3:352 Sneeze. _ 8:49 _ .31: _ . . . . 55¢... — I \\\\\ \ I - _ 235533 q .2233 _ . 529i _ I-ll'lll"'||'I"l|l‘||ll|'l‘ 50 ad inteneity pairn for the peak only) ad in cenen where nweep data wan acquired the archival routine cleared nae epace. A routine which allown accenn to the parenetern of a particular nca or nweep wa written. Thin involved one of the Variable Device Par-tar Tale (VDPT) which holdn the veluen for the particular per-etern «played in a nweep or nca. A routine to perfor- a ntale ion nweep wen written an wel-l. In thin inntace, the true utility of FORTH an a laguage nuitale for inntrI-at control nurfaced. The nweep in entitled 88'“? ad wee eaeily written aloying «inting ‘ ”wordn”. The data lint dinpleyn the tia-of-flight valuen ad with a BDISP coaand the data in dinplayed in inteneity vernun DAC valuee. Thin type of nweep facilitaten .the dinceraent of the nae npectrel tia-agnetic field nap. Mich of the noftware need in thin data nyeta hen baa adapted fra noftware originally written for a niniler nynta ’ on a triple quadrupole nann epectraeter [4]. The three nont coaon ncan perforad by coaercial 38/58 iatraentn are the parent, daughter ad neutral lone ncan in addition to the nolal ntale nca. The B-t diagra for the nca typee in TRIMS in ehown in Figure 4.2. A ntale nca in accoalinbed in the T8DB inntraent by keeping the ratio of the agnetic field (B) to tine (t) conntat (at e value that correnpondn to the accelerating voltage) while both 8 ad t are chaged. A 51 :2“- 110m‘ 1m-I Hogan”: HeId (goon) VI ' I 1 F at I v I v v.' v I v I v Yf'l to is an 25 so 3: Time-ot-Uignt (pace) 0 "—1 I ' r I I 1 o ‘ 0 IN 200 300 no 500 500 70° 0'20 Potent ion mass Tim-e 4.2. B-t diegra for aeveral types of B-TOF scan. The ecann repreeented are a) etable ion nca, b-d) parent ion ecann (conntat daughter) for daughter ione 100 an, 144 enu, ad 196 an, renpectively, e) daughter ion nca for parent ion 400 an [1]. Reprinted fra Stultn, J.T.; Bollad, J.F.; Bnke, 0.0. Anal. Chen. 1983, 55, 1323-1330. 52 daghter nca in achieved by netting the tia of arrival for the parat ann, ad ecaning the aaetic field. A parent nca in achieved by ecaning the agnetic field ad arrival tia each that their product (at) raainn conntat at the value correnponding to the fighter ane. The peakn are thoee parent ione that dieeociate to produceeaneequalto thevalue forthenelecteddaughterann. A neutral lone nca in a coalex linked nca ad the peekn correnpond to all the parentn that frag-at to lone a particular neutral ane. Daughter ncenn ad nweepe coarine a ajority of the data promoted in thin theeie. For «ale, eince athae ga hen been uned to gage CAD efficiency, ite CAD reactione can be nonitored by netting the arrival tia for ann 16 ad ecaning the aaetic field to lower 0A0 valuen. 8y obnerving the daughtern of the athae aolecular ion a value for the CAD efficiacy .ca be obtained. mm consummate Before conntruction of a workale collieion cell could begin, engineering of the gen delivery ad evacuation nyeten wan required. Once the idea of a ”aolecular bea-contaiaat” type of cell wa drawn up the flow anaicn ad plaing characterinticn had to be connidered. A book by Duel-a [5] aided connideraly in the underntading ad aplication of the rulen governing gee flow. Conniderationn of prennure rage ad flow conductacen greatly influence the final arragaent for plaing. If, for «ale, there in not a nufficiant 53 vacu- on top of the collieion region than the cell will not have accalinhed the goeln denired. For caletenenn, a brief diecueeion of nae gen flow parantern in preeented. .-.... an The dencription of gen flow in vacuun nyntea in generally divided into three partn, the divinien being epecified by three ragee of veluen of a dineneionleen peraeter called the [mafia naer. Thin maer in giva an: 80 i/d (48) ad in defined an the ratio of the a free path (i) of a nolecule to a characterintic dineneion (d) of the channel through which the ga in flowing, for «ale, the radiun (in a) in the cane of a cylindrical tale. The free path in defined an the average dintace a nolecule traveln before it colliden with anther nolecule. In the high-prenaure rage of vacui- nynta operation, where the a free path in all capared to the characterintic diannion of the channel (i << d), collieione between noleculen occur are frequatly tha collieione of noleculen with the channel welln. Thie region in deecribed an vincoln flow ad in characterized by all Knudnen maern. 54 At low preenuren, the aan free path in large cared to the characterintic dinennion ad the flow of ga in linited by aolecular collieione with the welln of the chanel (i > d). The aalynin of the flow in priarily a geaetrical problen of deteraining the rentrictive effect of the welln on the free flight of a nolecule. Thin flow at large lnudna maern in texad aolecular flow. The traneition fra vincoln flow to aolecular flow occure at internediate valuen of the Knudnen maer ad nai-apiricel equationn are and to dencribe thin type of flow. A niale approxiation to the type of flow ad therefore to the - correnponding equationn one nhould une, in ehown below. Since the aan free path in related tothe preenure (P = preenure in nicronn (u)), we have: when dPu > 500 ,1 the flow in vincoue. When dPu < 5 , the flow in aolecular. Wha 5 < dPu < 500 , the flow in in the traeition rage. A diagra of the 3 typee of flow ad their preenure region in ehown in Figure 4.3 [6]. It in iaortat to note that in aolecular flow the conductace in independent of the preenure. Conductance, men/sec Figure 4.3. The conductace-preneure reletionnhip defining gen flow regionn. Reprinted fra Mcfadden, H. Techniquen of Gained Gen Chroatography/Maee Spectraetry, New York (John Wiley ad Sonn, Inc.) 1973. A flow rate, 0, will be mad throughout the calculatione ad repreeentn a throughput or the product of the volaetric flow rate dV/dt acronn e plae, ad the preenure P at which it in aaeured. The quatity which relaten the throughput relative to the preenure differential in called the conductace, 0. The value 0 in defined an: 0 = 0/ (Pa-Pl) (4b) where pa in the upntrea preenure ad pi the damntrea («it) preenure. The value 1/0 ay be treated an a renietace, 0 a a current ad (pa-pl) a a electrical potential. In thin anner, a nerien of tiling connectionn ad prenauren ca be treated a a electrical circuit. The nont failier of the vim-flow equetionn in the Poineuille equation for the flow through a ntraight tube of circular croee-eect ion: 0 = To P. (Pa-PU/Snl (4c) 57 where a in the ttbe radian; l, the tale lengthfll, the vinconity of the gen; and Po, the aritl-etic a of Pi ad P2. Therefore 0 in: C= hem/8711 (4d) Modificationn are derived in Daaa for canon that ay deviate fra thin general equation. lnudnen proponed the fundaentel relation for aolecular flow ad it followed that in thin region the flow conductance in independent of preenure. For a long tube of length 1, varying cronn eection A, ad periater 8 the conductace in l 0 = 4/3 va / I 8/A3 d1 (4e) 0 where v in the aolecular npeed. For a cylindrical time of radiun a, 0 = 30.48 a3/1 (T/MP/z l/nec (4f) Through orificee end nhort tuben: C = 3.64 A (1711)"2 l/nec (4g) 58 The eaetionn in thin eection were thoee uned for the collieion cell denign calculatione. In nae caeee eneationn and epproxiationn were ado a deecribed in the following nectionn. T- ” 031.1. (C) commune ANALYSIS Before ay gen in allowed into the eynta, a conductace aalynie ca be perforad. An initial aetation in that the ga in entering a aolecular flow region ad aolecular flow will continue to the diffaion plan. Thin in a good approxiation connidering the fact that the preenuren in the cell will rage fra 0.1 to l x 10" nicron. A diegra of the cell nhowing the preenure regionn ad conductace regionn. dincaeed in ehown in Figure 4.4. Two are aeration are ade initially. nu firnt in that the preenure at A (PA) will equal the preenure at the point the gen exitn the inlet tube. Alno it in aneaed that the preenure at M (Pu) in equal to the preenure at the point betwea the elite ad in equal to the preenure recorded at the Penning gauge. When the relative dineneionn are connidered theee anationn appear to be valid. The vectore nignify half (for eialicity) of the direction optione open to a gen nolecule. It in deeirale to nininize the preenure in the chaer (Po). The preenure at the lower diffunion pta ad the 59 |"" P SO I I I ’4 ' A 8 013% . V C p l - - - PENNINO GAUGE Figure 4.4. A diagra illuntrating the varioun preenure ad conductance region of cell 0. 60 aflitianl diffunion pin in eignified en ho ad 9.. renpectively. A circuit ca be drum for the diagra ad in ehown in Figure 4.5. lquetioa ca be developed fra thin ad are an follown: 0a in =Gln u 9 u (PA - Pno)Ctop 4’ 2(PA - FNMA where (PA - Px)CA (Pu - Pc)Ce + (Pw - Pto)Cbotton If the aolecular flow ennlation in correct than a caarinon betwea celln A ad 8 (whone conntruction in deecribed below) ad cell 0 ca be nade with purely conductace coneideretione. A redraing of Figure 4.5 in ehown in Figure 4.6 with the indicated valuen for-the conductacen calculated fra the aove equatione. The value. of 24 l/nec arieee by coneidering the conductacen of the cell, the tube to the diffunion pl-p ad the nouth of the diffunion pin. The nlit conductacee are epproxiately 5 l/eac. The value of 14 l/eec includen the conductace of the region betwea the elite, the cell ad the conductace value calculated for the total dintace to the large eource diffaion pn-p. If we annae that 100 gen noleculee enter thin differatial eynten, Pc will be a great deal ealler tha in cell A or 8. lith cell A or B there in no option for the gen. It at pane through the elite ad enter the chaer along the ion bea «in. For cell 0, approxiately twice an .1011 gun will pane to the all diffunion pa an will pane to the large diffunion pa. 61 Figure 4.5 A circuit representation of Figure 4.4. COLLISION CELL C CONDUCTANCE REPRESENTATION 241/9“: (reassuec or SMALL) so (onrrusuou we?) Pc (cweuaEe) ‘ Sl/sec (Pezssuez) p (PRESSURE) SVm’ P. L0 (“t f" '3" (PRESSURE) I‘M/sec. ("555"“) (CELL sure) (AT LARGE) U" "Mm” (aurrusuou Pun?) (GAUGE) owE HALF or COMPLETE REPRESENTAUON Figure 4.6 A redrawing Figure 4.5 indicating the conductance values. 62 The quantion raeining in: will the diffunion pl-pn be able to accaodnte the flow denired ? Fra «periantn with celle A ad 8 a reenonable eetiate an to the preenure in the cell for CAD in in the l x 10“ to 1 x 10" torr rage. Meet other «periantern operate in thin preenure rage en well. In order to aintein a preenure in thin vicinity, with the. overall conductance calculated to be approxiately 10 l/nec, the throughput nhould not greatly exceed 1 x 10"2 torr l/eec. The overall conductace connidern the conductace to both diffunion plan fra the point at which gen entere the cell. A 20 a inlet caillary tube of 0.010” i.d. with a head preenure of approxiately .100 torr will nupply 1.3 x 10‘? torr 1/nec helia. Thin throughput will yield preenuren in the denired CAD rage. The inlet ttbe ca be further reduced in the lent 2 a by innerting a ealler i.d. tlbe into the capillary tube. Thie will effectively lower the throughput. Alno the capillary tube ca be increaned in length if a further reduction in throughput in necennery. Therefore, the aewer to the quentionoftheadequacyoftbept-pinyee, anlongenthethroughput in kept ealler tha approxiately 0.1 torr l/nec. The throughput will have to be aaller than thin to effect a working preenure of l x 10" to l x 10'3 torr innide the cell. With theee conditione the diffunion pan nhould not even cae clone to «periacing difficulty. If e eituation nhould arine whereby throughput requiraente are exceeded, the cell in conntructed' no that chagen ca be ade quite eaeily to nininize down tia of the ane npectroater. COUJSIW can. DESI. The heart of ay 18/36 technique lien in ite ability to further frag-ant a ion of intereet ad effectively collect ite frag-ante. Thin aility in characterized by fregeenting a known parat ion, naing the daughter ion aundancen, ad calculating a CAD efficiency value. The procedure followed by neveral authore in the pat to obtain CAD data wan to eialy allow a target ga to fill up the entire field-free region. In a defile-foaming inntraat, the field-free region and for target gen in that volla betwea the agnetic ad electric nectorn. In T898, the field-free region in that volae between theeourceadagnet. Saeauthornhaveevenuneda controlled bake out achainn to fill the field-free region with renidual gen thereby increaeing the probability for a precurnor ion collieion. The overall CAD efficiencien reported were ueually very low for theee "celle" ad therefore the DB/DB technique nuffered. For the high aergy inntn-entn it wen difficult to achieve overall efficiencien nuch greater tha 1 or 2 percent. An we nhall nee, highly efficient collieion celle (>118) have already been developed ad are in uee currently [10]. The eie of thin theein in to build a cell for T8113 that will yield coaarale CAD efficiacien ad connequently allow tB/FB data to be obtained on a eingle focueing inetruaent. 64 A anon 031.1. ”I“ Inetead of introducing collieion gen into the entire field-free region preceding the electric nector of a double-focuning aaee epectraeter, aother idea involven the confinaat of the target gae to a are npecific region conniderebly ealler than that previounly aed. Thie type of denign in ehown in Figure 4.7. Collieion gen atere the region ad in contained naewhat to increaee the probaility for collieion ad ntbeequat fregnatation. The advatege to thin early denign wan that it wen eiale. The relative contaiaent in depadent on the conductacee of the nynta. Thin denign could be fitted to ay field-free region. The parent ione entering the cell frag-ented to a certain «tent ad undienociated parate an well an daughtern were trenenitted. There are nay dieadvategen to thin denign. The firet in that a collieion reaction can occur at ay point along the length of thin region ad ay involve coneidereble agular deflection. A direct connequence of thin probla in that. the focun to the eubeequent aalyeer becaee poorly defined. Thin ca ceune traaienion lonnee ad a decreaed nan reeolution. Alno, a preenure gradient ca axint in thin deeign leading to other effectn that will be dealt with later. A further undenirable connequence of long collieion chaere in the beckncattering ad detection of ione which have kinetic aergy releanee of nuch agnitude ad direction that they would otherwine be dincrininated againnt. Thie leadn to a lone of aergy reeolution. It in not eurprining that thin type of denign yieldn low efficiacien. «a, to 10 cw—u- ANALYZER REGION V l PARENTS _._____,. AND DAUGHTER; : l J GAS IN Figure 4.7. A collieion cell of early deeign for target gen confinenent. 66 Specifically, a epatielly enell collieion region, located at a.bean fool. fra the ion eource, ha clear advategen with renpect to nininining avoidale traeninnion loenen ad nann renolution deuadatia. A M COLLISION CELL J.8. Deynon et al, in 1973 denigned a nhort collieion region located at the object elit of a electric nectar. A diagra of the cell in ehown in Figure 4.8 [7]. Thin deeign avoia any of the probla prenated for the firet deeign. Since the eource elit in norally nituated at the focal point of the electric eector, they incorporated thin elit within the collieion chaer. The chaer in a tube of 0.25 inch diaeter ad the eource elit wan replaced by two elite of 0.005 inch width cut into opponite niden of the tube. Surrounding the chaer in a eecond concentric tube denigned to provide differat iel ming. Figure 4.8. -..-a ' ' F ' C ' O H ' _.-4.Ja...e._.aa'rldn.(“u¢‘llaol--‘-.l.‘.ll__.1 ' B l ,Reeolvmq Sms -... ___, Ion SOU'CE ' l FITS? Fund-Free Region .'.«‘3I-3‘1 ~)(33 \. ~.\ Inlet System A nhort collieion chalber placed at the focal point of the electric aector and incorporating the eource elit [7]. 68 The cheaer wen ade only 0.25 inch in length to circavent the probla encountered with nont other denignn. Now, ennentially all collieione occur at the focal point. Nha perforning kinetic aergy aauraentn, one in concerned with the connequencen of reaction at a certain point in the inntraent giving rine to a product ion of a epecific energy ad direction. If all reactione of intereet occurred at the focal point of the kinetic aergy aalyeing device then product ione would be focueed at the detector ueing electric eector voltagee dependent on their energiee ad indepadat of their direction of notion. Thin nhort chaer yielded up to a order of agnitude inrovaent in naeitivity for CAD over that obtained in the entire field-free region. In addition, giva reactione occurring at the focal point, the recorded kinetic energy dietribution would then provide a accurate napping of the energy dietribution of the reaction productn. with the nhort deeign each incaing ion in likely to nuffer only one interaction in the collieion region. Thie fact in iaortat fra the thereochenical ntadpoint whereby one in inventigeting aergy proceenee for npecific reactione. Cooke and Beynon were inventigeting charge tranfer reactione ad denired a are accurate kinetic aergy dietribution. They had found a iarovaent in aergy reeolution. Nclefferty et al, have ehown that altiple collieione ca be of baefit to the ann npectroatriet. Precureor ione that undergo altiple collieione not only increaee the CAD efficiency, but aleo increaee the intanity of thoee fregnent ione fomd via high energy proceenen [8]. Therefore, thie deeign wen tailored to Cooke and 69 Deynon’n «perinate eince they deeired eingle collieione for their aauraentn. Another advatage to thin denign wan in tera of control of target gen preenure. with the all region the gen contaiaent wan ach better ad there wen lane of e chace to develop preenure gradiatn. Thin deeign aleo nininieed the enomt of target ga that could diffune to other parte of the inetrI-at ad therefore decreaeed the chacen of inelaetic ecattering either before or after a reactive collieion. One difficulty noted with thin denial wee that even with the differential pulping nynten, nae leakage of the collieion ga into the . eource ad into the firet field free region did occur. Thie preenure probla in one of the aJor difficultiee in denigning a collieion cell. The L88 9000 in not a differentially pl-ped inntraent ad therefore to prevent target ga leakage, a differentially plaed cell in. necennery. 70 ‘1. mm GILL A couple of yearn later Cooke ad Dayna developed a cell that dincrininnted egainnt framentetionn occurring outeide the cell [9] . In etudying the factorn (inetraental ad aergetic) detenining pea ehapen for collinionally activated decaoeitione they applied a potatial to the chaer an ehown in Figure 4.9. The potential within the chaer in variable no that reactione occurring there can be dintinguiehed fra thoee occurring elnahere in the field-free region. TB mm mm CELL (0M0) The aove anime repreeent e are ntatic type of collieion cell. McLafferty et a1, [10] have taka the idea of conntructing a all collieion region for focuning purponen and developed a dynaic type of cell ehown in Figure 4.10. Thie denim ealoye a ga jet target rather than the conventional etetic cell. Lena A foclnen the precurnor ion baa onto the collieion baa. In the collieion bea analy, helia atoa which effune fra the nozzle have velocity vectore preferentially directed toward the large diffunion pla. Thin will reduce the diffunion to the rent of the inntraent. The heliun atern fra a 25 a elit opponite a 6000 l/eec pla ad producen a collieion ribbon orthogonal to the ion baa. With efficient ‘71 5— Annuod :1 -\ Potential / Electreeehc { Analyzer Mons Resolving Sm S 0020. 0.0I0' m . | '\ Ion been \ / ‘/:————t . Slit: 0.020' .—-—-—— I2'————-‘— O.875'—~———— 6.56'-—-——-b Figure 4. 9. A diagra of the floating cell to diecrininete egainnt fregnentationn occurring outeide the cell [9]. 72 I tone IOLECULAR It“ COLLISION .. CELL . 83 ea aniline IO 00"“ NON 90.? Figure 4.10 A diagram of a molecular beam collision cell :10}. 73 differential pulping a pressure of approximately 1 torr could be naintained at the be- center. so that lost collisions took place within a very narrow region near the focal point of the ion lens systel. Bxperinentally they have deterained that >95: of all collisions and frag-entations occur within 2.5 II of the helius ribbon. In conventional collision chasbers the gas effuses preferentially along the ion been axis. This effect was silulated by closing off the systel which directly pulps the helius been; this caused the collision efficiency to drop sevenfold. A SIMPLE NEEDLE CELL Another even si-pler design than the earlier systess was developed by Glish and Todd [11]. With the idea that a single point of focus is desired for opti-n perfomnce they constructed the cell in Figure 4.11. Ideally, an infinitely thin collision cell at the object point would be desireeble. This design introduces gas through a needle via a Cajon Ultra-torr adaptor with a 1200 l/sec pulp directly underneath it. The overall CAD efficiency found was between 7 and 8 percent for the sethane aolecular ion at 70-803 parent ion been attenuation. 74 NEEDLE CELL tARGET / 2 .E— x Y "new! nous mum-area IONS Ewrnwcs sur TO nus srzcrsowsrcn 1. TO 1200 L PER SEC DIFFU$!ON PUMP Figure 4.11 A diagram of a simple needle target gas inlet system. 75 A mum mu GILL (THIS mm This latter design was evaluated on the m 9000 but due to the lack of any differential pining there was a noticeable build-up of pressure throughout the analyzer. Since the initial experiments pt-ping changes have been sade. The needle was eodified by encasing it in an aluminum sheath to attempt better containment of the target gas. A diagram of the cell that ha been used for a eaJority of the work that will be presented, is shown in Figure 4.12. One problem with this type of cell is that the gas must escape through the ion entrance and exit of the collision chainer, producing gas plumes in either direction from the chadmr along the ion beam. Thus even with a minimum-width chafier, a substantial proportion of the collisions may be occurring outside the cell. It should be noted that in this case the cell 'is not at an object point for the magnetic sector. Nevertheless, the cell is placed directly after the deflection lens assafily and is as close as is physically possible to the entrance slit of the mass epectraeter. Cell A is the first of two collision cells tested in this thesis. It was constructed from a 25 ul Hamilton syringe that has the top "T" portion removed and fire polished. The syringe outer diueter is 7.50 -. The length of the glass portion of the syringe is 7.40 cm and the 76 cuss / same: 4, , ~<——— EPOxY ALUMINUU 2.05” l n . g COLLISION CELLS ’ A AND I figure 4.12. A diagram illustrating the principal coeponents of cells A and 3 described in the text. Slit dimensions in cell A were 9.50 - x 1.50 -; those in cell 0 were 8.73 - x 0.95 -. 77 needle extends 3.30 on into the almimn sheath. The aluminum is ccented to the glass with epoxy. Outgassing of the epoxy was no aajor problu. The total length of the sheath is 5.90 ca. The slits were cut by our machine shop to be 9.50 - in height and 1.50 - in width. Since this was the first cell, the slits were made relatively wide. This cell was grounded by allowing a long thin wire, attached to the base of the cell, to dangle in the sass spectre-star and touch the metal vacuum Mar wall. This arrangement was changed in cell B. Rotation was possible in either direction by means of a 7 .94 - 'diueter CaJon fitting. The cell also had the capacity for vertical and slight lateral adjust-ant implemented by movement of a support bracket attached to the mass spectrometer frue. Cell 8 is similar to A except for a few ainor changes. The slit area is decreased by 443. The cell B slit dimensions are 8.73 - x 0.95 -. Cell 8 also has a thin copper connection to the alt-inl- sheath to provide better grounding and therefore minimal charge build up. he volume in which collisions can occur in the cell is 0.083 cm’. If we assume rooe temperature and a pressure of 1 x 10" torr helium in the call, there will be approxinately 3.24 x 10*12 atoms of helium per ethic centimeter available for inelastic collision. 78 TH! NII’CBLL.OONBTHDCTTOI Both cells A and 8 described above provide average CAD efficiencies, but a different design could potentially increase these values. The data fer the two cells are presented in chapter V. A new cell has been constructed in our machine shop that incorporates the design features presented in section 2 of this chapter. A diagram of this cell, cell 0, is shown in Figures 4.13-19. This new cell has many positive attributes. The most beneficial attribute is that it is differentially pumped. This helps to defray the gas load and provides a narrow region for collision. The cell is modular. The reason for this is that it is desirable to change certain features easily and to experiment with the changes. The bottom section of the cell physically turns into an upper section (as shown by Jagged ends). The upper section is 11 cm in height and the bottom section 4 cm excluding threads. The cell is 31/32" in outer diameter to fit into a CaJon-type fitting for the mass spectrometer. There is a short concentric tube with slits cut into it for ion entrance and exit. All the slits are of the dimensions of cell 8, because cell 8 was found to exhibit a slightly higher CAD efficiency. The gas entrance will be along a narrow internal diameter tube (0.018 cm) of 19 cm in length. A curved piece of tubing is fit over this to permit experiments with different nozzle tips. Initially, the curved piece is simply an extension of the inlet tube. 79 b— $151—$— 2/4 " ------- ---------—-—----------.----- .-----'----”-----‘-------------.‘ 4‘. $31 a 7 ' . LI (4%. 14' E u/n' ‘ :H p:' J u 31/ c H I-— 31——-u| b—zfif—d Figure 4.13. A cross-section through the upper portion of cell 0 drawn to scale. Vv“ - O. Figure 4.14. A cross-section through the lower portion of cell 0 drum to scale. 81 g2. ..l... ‘Z UPPER saw on: J cou. IS|ON cw. “c” T l as": in i J : F—’/; ——-1' figure 4.15. Another view of the upper portion of cell C (not to scale). To Top of Cell %; T *w‘ +___—- Remova ble Inlet Line M434 Exit ‘9 /Slit Ion Beam Lfi __’ ," " "1“: .'"‘ 1| ——5 Parents and Entry : r— . : Daughters Exit / 1"“"Ll;'.'l“ "7 Entrance ‘ Slit l . I To 600 R/sec Diffusion Pump figure 4.16. Another view of the lower portion of cell C (not to wane). L .900" II Jr. I I -| P________"4W I ' to center w I ly__--___,l 3". figure 4.17. A sore detailed drawing illustrating the slit apertures and target gas inlet. .o deco no nonetoa gases on» no zoe> so» on» no nuances < ee.e ensues Amcam em>o mange amen snowshoes: Ammomvpnoaasm xmeo nee newscmeee one one; umHQH eon mac ..1 le Ill. NM? ... ll; . = Heme no .oe wywl. bi” ‘ Stabilizer Bar 49'. 4—— Bored Swage Fitting 5,1. for 1/16" Inlet Line figure 4.19 An end-on view of the top portion of cell 0. 86 The center ring is 5/16" in diameter and it is possible to decrease the top opening to provide more containment of the gas if necessary. The inner slit distance can be changed by modifying only the lower portion and not the whole cell. Another advantage is that the complete design is constructed out of brass and therefore grounding is not difficult. Also a pressure gauge will be fitted to the top of the cell to allow monitoring of the collision pressure. An overall classification of this cell design, relative to the others described, would be of a "static-dynamic" nature. It is designed to use the velocity vectors of the incoming gas to an advantage. In this way the gas leaving the tube creates a dynamic systa. If a smaller diameter ”washer” is used for the top of the center tube in figure 4.18 (thereby reducing the "basket containing the slits” dimension) then a certain amount of gas containment is achieved. In this way somewhat of a static system is developed. The target gas leaves the tube at the base of the slits and passes either to the small diffusion pump, the large diffusion pump, or the analyzer chamber. These considerations have been discussed earlier in section 2a. of this chapter. 87 new MITIGATION rammmarmcsusm In order to provide a gas inlet line to the collision cells a supporting piece of angle iron was mounted to the frue of the mass spectrometer. The angle iron was slotted vertically and the connecting bracket fixed to the frue was slotted horizontally. This arrangement allows for a coarse adjustment of the collision cell. Afixed to the angle iron is a 3/16” thick aluminum plate with 1” slots for fine control of cell positioning. Afixed to this aluminum piece is the necessary vacuum and inlet plumbing connections. A diagr- is shown in figure 4.20. The idea was to design a gas inlet that is adjustable, sturdy and of low volume. The lower voline allows fast gm equilibration and therefore better flow control. A Swagelok dual inlet needle valve allows annual control of the flow into the apparatus. Two on-off valves are used to allow evacuation of this line. These valves are inortant because a build up .of pressure behind the valve closest to the duel nudle valve could be disastrous for the ion source. Became of this the inlet forevacuum valve is always opened before the other valve. Helium, deuterium, or other collision gases enter fru the left on the diagm. A ”T” connector between the two valves provides direct connection to the collision cell. A Swagelok connection is eade with cells A and 8 using graphite ferrules. e. \n ‘.‘ figure 4.20 Vacuu and inlet plating connections for the various cells used. 89 Optin- ce11 positioning is achieved by maximizing the signal for an ion of medium abundance. first, a slight movuemt of the angle iron and second, an adjustment of the slotted aluminum plate will usually suffice for proper be. focus. Once a maxim is found, the alumimn plate is fastened into position with bolts and the same is done for the mgle iron. Rotation of the cell is easily achieved by applying an open end wrench to the swagelok connection (and turning slightly. The ion peak can be observed to increase and decrease in intensity throughout the adjustment period. The time required to locate the maxi- ion current varied between 10 and 15 minutes. This time should decrease for cell 0 because it will be constructed better (no epoxied sections that may be off center slightly). Nevertheless, once the cell is positioned properly no further adjustments will be necessary. The stability of the positioning is very good. Slight vibration effects were noted in cells A and B, but cell 0 fits closer to the CaJon fitting and thereby reduces vibrations. A diagr. of the collision cell flange addition is shown in figure 4.21. .nvw>~n:n one :o wlw:c_a«moa nmcn~m -oo scandaaoo ~N.e ansmmm £331 £053.55; Qua cut ... 000 O.— I FIFI 3(uu to. Alllllsn.ulf —I||u / moceom woz GZ< m>J<> 1(u4 Z. 2(UJ whr4(2( 95 was connected to the extension "T" and to this a needle valve. An on-off valve follows the needle valve with the associated evacuation txbing. A short piece of 0.0625” o.d., 0.010" i.d. stainless steel (EPIC) tibing was placed before the on-off valve to restrict the flow to the needle valve. A piece of glass 2.5 cm in length is inserted into the probe tip to electrically insulate the probe from the ion source box. In the experiments performed to check the probe as a s-ple leak device, the variation in the parent ion signal was approximately +—58. This variation was from one data collection to the next over a short period of time. Greater variations occurred when starting and ending values are compared after 1.5 to 2 hours of operation. Thisprobe provided a simple means of introducing methane into the mass spectrometer and was used for a majority of the CAD efficiency experiments . for convenience in focusing of the magnet a ”softknobs" interface was constructed that allows the control of the digital signal sent to the MC. The softknobs interface was designed by Bruce Navcome, built by me, and progra-ed by J .T. Stults and Carl Myerholtz. A diagram of the schematic is shown in Figure 4.24. The softknobs are rotary shaft .23.. 83333 Boast... 2: no ofloflfia as... 83: mufmmHZH mmOthow £73m, - 5:3 .- Six/$3 _ - 13.1...- ...... ..<.=u,awn i. -. . : -. A 756...: A. .. Queue..- .. 2%.... . 1-; -, ELFWWWE; 1a. 22.x. l Ln hLCS'N. OOOOOOOO 97 encoders (Panelcoder model 62, Disc Instn-ents, Costa Mesa, CA) that provide manual control of the magnet. The knobs output two quadrature encoded digital signals, 1A and 18, for example, each associated.with an edge-triggered flip-flop. The waveform will be + or - sonar... out of phase depending on whether the knob is turned clockwise or counter clockwise. The rising edge is used to sample the state of the other pulse. Thus one flip-flop is set for a clockwise rotation, the other for counter clockwise rotation. The Q outputs of the two flip-flops set bits in a latch. The -0 outputs are NANDed and fed to a monostable. The monostable provides a delay and a rising edge that strobes the latch. This rising edge also sets an interrupt flip-flop that signifies valid data has been latched. Next an interrupt routine reads the byte out of the latch and clears all ths_f1ip-flops. The . value read by the interrupt routine is decoded and the corresponding device is incremented or decremented. 3. 4. 10. 11. Stults, J.T.; Holland, J.F.; Enke, C.G. Anal. Chem. 1983, 55, 1323-1330. Stults, J.T.; Myerholtz, C.A.; Newcome, B.8.; Enke, C.C.; Holland, J.F. Rev. Sci. Instrum., in press. Ramalay, L.; Ling, E.; Burkholder, 0.; Ieki, M.; Jones, W.E. Chem. Biomed. Environ. Instrum. 1979, 9, 335. Myerholtz, C.A., Ph.D. dissertation, Michigan State University, Department of Chemistry, East Lansing, Mi. (1983). Dushman, 3.; Lafferty, J .14. Editor, Scientific Foundations of Vacuum Technique, 2nd ed. Neew York (John Wiley and Sons, Inc.) 1%. . McFadden, H. Techniques of Codained Gas Chrcmatography/Mass Spectrometry, New York (John Wiley and Sons, Inc.) 1973. Beynon, J.8.; Cooks, R.G.; Eeough, T. Int. J. Mass Spectrom. Ion Phys. 1974, 14, 437. McLafferty, F.W.; Todd, P.J. Int. J. Mass Spectrom. Ion Phys. 1981, 38, 371-378. Beynon, J.8.; Cooks, R.G. Int. J. Mass Spectrom. Ion Phys. 1975, 19, 107. McLafferty, F.W.; Todd, P.J.; McGilvery, D.C.; Baldwin, M.A. J. Am. Chem. Soc. 1980, 102, 3362. Glish, 0.; Todd, P.J. Anal. Chem. 1982, 54, 842-843. cunnv. mm The crux of this thesis lies in the question: can the cell achieve CAD efficiencies that are cmpardule to those found by the molecular beam and needle designs discussed in chapter four. The evaluation of the cell performance on be divided into four major categories: (a) calculation of its efficiency, (b) comparing 'the effects of different target gases on the operation and overall efficiency, (c) comparing the effects of pressure on the system, and (d) comparing the results of the [48/56 analysis on other established chemical systems. The final comparison in (d) includes the qualitative- agreement of spectra with those of other authors. All three cells will be discussed in light of the above areas. It should be noted that TRIMS is still in the developmental stage and therefore only well-established chemical syste- were investigated for this thesis. new NATION The TRILS instrument is of an average complexity in operation and therefore a brief discussion of how data are actually acquired will prove useful. The instrument start up is the same as that necessary for acquisition of a normal aass spectrum. The maaetic field is scanned a few times to set up the fluxes. Suple is introduced 99 100 through the direct probe inlet and the source pressure is adjusted to approximately 6 x10” torr on the Penning gauge. The filament is turned on first, then the high voltage acceleration potential, the multiplier, the magnet current power supply, the scope and the .plifier power supply are turned on, respectively. For a normal sass spectrum the detector amplifier is an LF351 op amp with a 10 megohm feecback resistor. The ban is focused and either a calibration can be performed (by closing off the direct inlet and introducing a calibration compound) or pulsing can be readied. hhen setting up for pulsing mode, the multiplier is turned off and the detector amplifier is switched to a fast pulse nplifier (Comlinear Corp.). The multiplier is then turned on along with the Wavetech pulse generator for the extraction lens pulsing. (More details can be found in J.T. Stults’ Ph.D. thesis from MSU, 1%). The extraction voltage is manually adjusted simultaneously with the magnetic field to locate peaks. Once a peak is located it is focused by adjusting the focusing plates in the source. At this point CAD experiments, data sweeps, etc. can be done. ACADM If a CAD experiment is to be performed, the source vacuum guard is first turned off. Because the LED lacks a differentially pumd collision cell the pressure in the analyzer will rise considerably and 101 the vacuum interlock will consequently turn off the electronics. By operating the dual inlet valve, after evacuation of the collision cell line, the target gas is permitted to flow into the cell and produce a concaittant decrease in the parent ion intensity. For example, consider the case in which sufficient methane gas is ionized to produce a 25 mV signal for the molecular ion peak. After adding collision gas to the cell to a pressure of approximately 2 x10'5 torr (uncorrected Penning), the molecular ion peak signal drops to 15 mV. At this point the daughters of CE" can be observed. When reducing the magnetic field to lower mus values two peaks arise on the scope: the parent ion which is the molecular ion of mass 16 and its daughter of mass 15 which is transmitted at the magnetic field strength that corresponds to a calculated apparent m/z value of approximately 14. This implies that the m/z 15 daughter of m/z 16 arrives at the detector at the same time (estdalishing the parent-daughter relationship), but at a lower magnetic field strength corresponding to the apparent mass. This checks with the B-t theory. Proof that the collision cell is the cause of what is observed is found by simply removing the target gas and searching in the sue manner as described above. Other daughter experiments were performed in this manner and documented by computer acquisition. 102 CUIPUTEEIAOOUISITIOI’OF DAUGHTER SPECTRA When acquiring data by computer, the Wavetech function generator must be triggered by the computer. The ion signal is connected to the gated integrator and various commands in FORTE are entered. To initiate a data acquisition, a file must be entered first, followed by , the name of the experiment to which the data will be written. The parameter editor, FED, is then invoked to prepare for a sweep or a scan. For a daughter sweep the DAC value for the parent mass must be found in order to determine its time-of-flight. The softknobs instrument addition described in chapter four is employed to find the DAC value for the peak. Once the DAC value is known, a time sweep can be performed.(no collision gas) by setting the magnet at the determined DAC value and sweeping the time over a set interval. Currently, the maximum number of data points obtainable with a sweep is 2000. Therefore care is needed in assigning the interval. This sweep will yield a display of the ion peak intensity versus time. A command, DLIST (data list), will yield the time-of-flight (tof) values. Once the tof is known for the parent, the daughter ions can be found quite easily. Collision gas is added and P80 is entered. The current time is set at the tof of the parent and the magnet is scanned corresponding to the set interval. The sweep data are’written to the selected experiment and can be displayed simply by selecting the scan. The parent peak tof has been found to drift minimally (+ or - 2 to 3 as) over the course of one hour. Day to day flight time variations 103 have been found to be very small as well. Most of the data presented were found in this manner. The direct probe inlet facilitated sample entry and the softknobs interface facilitated DAC control. EFFICIENCY'CALCULETION The major determinant of the value of a particular collision cell design to an DS/IUB researcher is the nuflier representative of the collision cell’s efficiency in fragmenting parent ions into detectable daughter ions. Collisionally activated decompositions yield not only a greater number of fragmentation pathways but usually increased fragment abundance as well as fragments that are more structurally significant. The overall CAD efficiency depends not only on the cell design, but also on the performance characteristics of the instrument used. A particular cell design may be very good in principle but if, for example, the ion optics are poor the efficiency for the CAD process will suffer. CAD efficiency is composed of two terms: the fragmentation efficiency and the collection effeciency [1]. It is the product of these two efficiencies that accounts for the overall CAD efficiency. 104 The fragmentation efficiency (Er) is defined as: Eur = if: / E (F: + P) (58) where ft is the abundance of the ith ion (i is summed over the range of fragment masses observed) and P is the abundance of the parent ion after the collision cell. The denominator represents the exit ion flux or the total ion current. The collection efficiency (Re) is defined as: Ec =Zm + p) / Po (5b) where no represents the ion current when the pressure is equal to ”zero” in the cell (i.e., without target gas). 'The overall CAD efficiency is simply the product of the two and relates the total fragment ion current to the ion current of the parent under the original pro-CAD conditions, Ecru = 2m / Po (5c) The collection efficiency is found to decrease as the percentage of attenuation of the parent ion beam increases and has been explained to be due to scattering losses and charge neutralization. The fragmentation efficiency, on the other hand, increases due to the increasing probability of interaction of an ion with the collision gas. Diagrams displaying the counteracting efficiencies will be shown below. 105 Since fragmentation of the molecular ion of methane has become the bencl-rk for evaluating collision cell efficiency [2] , it is this CAD process that is dealt with most in this thesis. A diagram of the fragmentations found for methane is shown in Figure 5.1 [3]. A time-sweep with a constant magnetic field for each.mass was performed for the ions of’methane. Beam deflection was employed here to encode time. The results are shown in table 5.1. The data indicate that there is a considerable overlap on the time axis due to the broadening of the low mass peaks. This broadening may be due to the pulsing mechaniem. Nevertheless, this broadening of the stable ions does not interfere with the mass assignment when sweeping both 3 and t. COLLISION'CILL EFFICIRIUY’HBTBIIEIIIIOI The first efficiency experiments were done with methane as an analyte, collision cell A, and helium as the target gas. Helium gas was chosen initially due to its established high efficiencies (discussion below) in comparison with other target gases [4]. Cell A was the first design to be investigated (its construction is discussed in chapter four). The fragmentation of the molecular ion of methane has been docunented extensively. Therefore literature 106 '1 .1», + + + + + CH4—s- CH—..CH—.-CH' C L3 VJ] Figure 5.1 Fragmentation of the methane molecular ion. 107 than: 5.1 PEAK MASS DAC TOF Q6601 BASE 91360) MN 14620) 16 4614 12.94 .56 .17 15 4435 12.60 .56 .17 14 4243 12.32 , .04 .28 13 4045 12.06 1.06 .33 Tile 5.1. Peak widths in time determined by time-sweeps of. the B-t . data field. 108 comparisons can be made. The parent mass, 16, was focussed upon, collision gas was then introduced into the cell and the fragments were feund by scanning the magnet at the arrival time of the parent. The raw data were then analyzed by peak height or peak area. For methane, the l5, l4 and 13 daughters of mass 16 were usually observed, but due to sensitivity limitations the n/z of 12 daughter was in some cases difficult to discern from background. An example of a raw data sweep indicating the 15 and 14 fragments of 16 is shown in Figure 5.2. The pressure is 4.3 x10‘5 torr as indicated by the Penning gauge. The actual pressure in this value divided by 0.21 (for helium). The maximum CAD efficiency was found to be 2.8!. The data are plotted in Figure 5.3. The error was estimated to be approximately +- 0.1: by comparison to duplicate trials. The parent ion intensity entering the collision cell was monitored by closing off the target gas inlet which allowed the pressure to return to 1 xlO" torr as indicated on the Penning gauge. No netsstables were observed, therefore the assumption that there is a very little metastable contribution to the CAD efficiency, is valid. Peak heights were measured and scaled appropriately. Intensity 100 60 ‘0 20 109 M ETHAN E MOLECU LAR ION CID WITH HE P:- 4.3X10-5 T .... A J L. ‘1' I1 I I I l T ‘ I I T ‘ T ‘ I I I 380 390 400 410 420 430 440 450 450 MAGNETIC FIELD (DAG UNITS) I ‘ l 470 480 Figure 5.2 A raw data sweep of methane molecular ion CAD. 110 METHANE MOLECULAR ION WITH HE TARGET 3.00 T 2.30 5 . E T 5 T LT. 2.60 LI. 11.1 o i a: o '— .— 1.21 2.40 ! g T LIJ 0. 220 P l I I l l L l l l P _‘ 1 ._L _L J 20° 'FTIrr'IfiifffrTi 20.00 40.00 60.00 80.00 100.00 PERCENT ATTENUATION Figure 5.3. Efficiency curve for methane relative to pressure. the . pressure at approximately 75: attenuation is 5 x10'5 torr. 111 Fm sweeps resedpling Figure 5.2, the collection efficiency and the fragmentation efficiency could be calculated. Figure 5.4 illustrates the resulting curves for cell A. The trends correlate with that obtained by other investigators. The precision of the CAD experiment is affected by parameters such as target gas pressure, analyte been variation over time, and the ion optics of the instrument. As a check on the reproducibility of the results, the same-experiment was performed one month later. The results parallel the trends observed earlier with the percent CAD efficiency values differing by approximately +- 0.13 in the vicinity of the maximum. At lower pressures, the error is closer to +- 0.52 which may be due to inconsistencies in reading the peak heights at the low fragment ion currents. This error was later reduced by use of a FORTH program to calculate the peak areas and the corresponding efficiency values while taking into consideration detector gain. It should be noted that when the parent ion intensity is low enough in the beginning of the experiment, daughters can be found employing the sue detector gain. Sose later experiments were performed by increasing the gain and correcting for it in the efficiency calculation. Collision cell B resembles cell A except fer the fact that its slits are 448 less in total area. This was done to investigate whether the containment of target gas could be achieved.nore ‘EI‘lIt'Itl I“ ‘ 112 METHANE MOLECULAR ION FRMflIBflWflON GMUEBHON1NID(”EDEHHGENCES 1GBJ.A pa 8 .1 $00 $00 L70 L00 L50 '. [I E wH—1—H—1—i-+~+—+++—H~+—+—1~ LOG EFFICIENCY §§§§§§§§§§§ \ T A + 1 111 L11 14 T 11 Ft] r_1 1 I 3.00 4.00 4.40 4.00 0.20 0.00 0.00 0.40 0.00 7.20 7.00 PRESSURE HE (TORR X10 +5) Finn-e 5.4. Collection, fragmentation and overall CAD efficiency curves for the methane molecular ion. 113 effectively and thereby reduce the probability for ion-molecule reactions outside of the cell. A narrower slit should provide approximately the same CAD efficiency, but allow for a lower gas throughput due to a decreased conductance to the collision chamber. This, in fact, was found to be true for helium target gas. The data for cell B are plotted in Figure 5.5. The CAD efficiency was found to be approximately equal to that of cell A at 2.5 x10‘5 torr on the Penning gauge. This experiment though, had some difficulty in target flow control which could have been due to the neans by which the inlet needle had been cut. Nevertheless, the same general trend in the data is observed as with cell A. As described in chapter 4, collision cell C provides differential puming to the mass spectrmeter syst- enabling more efficient exhaustion of the target gas. The efficient removal of target gas should provide a better focal point for collisions as well as distribute the gas load to other pumps besides the large diffusion pump on the analyzer region. Again, methane molecular ion was passed through the modified direct probe to provide a relatively constant sample back pressure. Conditions in which the other cells were used, were duplicated as best as possible and CAD efficiencies were calculated at various target gas pressures. PERCENT CAD EFFICIENCY 1° P .u a: O O O O O to 3 irr-I-d—i—I—I-H—rI-Jr-I-r‘rH-I—H'Ir 2.20 2.00 P o o 114 METHANE MOLECULAR ION HEUUM TARGET CEU. B PERCENT" ATTENUATION 42.00 44.00 46.00 48.00 50.00 i I l T I T w- . . . 1 T l— r 2.25 2.50 2.75 3.00 PRESSURE HE (TORR x10 +5) Figure 5.5 Efficiency curve for methane CAD esploying cell B. 115 The results of the experiments with helium are shown in Figure 5.6. The target gas flow was easier to control because the gas was distributed throughout the delivery system much more effectively than through the ”needle” cells A and D. This may not be a great advantage in perforning 08/38 experiments, but it facilitated pressure monitoring. A maximum CAD efficiency for helium is approximately 5.5x at 4.0 x10’5 torr. The same general trend is observed as before, but cell C provides a significant improvement over cells A or D. A plot of the raw data indicates a feature that was not observed in previous cells with helium (see figure 5.7). With the improved efficiency the 16->12 peak is discernible from background at this detector gain. Although this peak is discernible, the efficiency calculations did not use this peak area value. The precision of these efficiency measurements is approximately +- 0.053 and is due mainly to better control of target gas pressure. In order to further gauge the performance of these cells more experiments were performed. Initially the type of target gas was varied and a few documented chesical fragmentations were studied. 116 METHANE MOLECULAR ION HEJUM TARGET 08.1. C PERCENT ATTENUATION 20.00 40.00 60.00 80.00 1 00.00 1.— 1— T r T j T r F r 1 I —I' I *1 5.00 i -I- ‘2 + DJ 4.00 as T E 9 .300 o i 1.. z . ‘5‘ a: 2.00 11.1 n. 1.00 1 L 1 1 1 . L . 1 1 1 . 1 . 0.00 T I I 1' 1 T I r U r I l T r r 0.00 2.00 4.00 6.00 8.00 PRESSURE HE .(TORR X10 +5) Figure 5. 6. Efficiency curve of methane CAD showing the improved values employing the newly desimed cell. 117 CAD OF METHANE MOLECULAR ION 100 15 14 13 AND 12 DAUGHTERS HE TARGET " 'I Intensity . ’15 2° 13 14 ’12 ‘ L 0 TrTrrjer‘T'l'l'I'l'l'l'l'l‘Trl'l'IEI'lrl 270 280 31 0 330 350 370 380 41 0 430 450 470 MAGNETIC FIELD (DAC UNITS) Figure 5.7. A raw data plot showing the daughters of methane observed when using cell C. 118 TI! NATUII OI’TII TARGET GAS A'great deal of the work on CAD of polyatomic ions has been done with target gases such as argon, helium, nitrogen, and even air. When comparing different targets, relative cross section measurements or efficiency calculations are usually employed. One might initially suppose that since CAD theory indicates the occurrence of an electron cloud interaction, the greater the electron cloud the greater the cross section. Therefore, gases such as xenon should be esployed. However, the energy transferred in helium collisions is greater because the electrons involved are of higher energy [5]. In addition to energy effects, charge exchange and scattering are to be considered when employing a target gas for CAD. Energy deposition is one of the important aspects of CAD. Ordinarily, it is desirable to maximize this energy in order to increase the degree of fragmentation and to favor highly endothermic decompositions which are most useful in distinguishing structural isomers [6]. The most widely employed methods of altering the amount of energy deposited into the ions are variation of the kinetic energy of the ions and variation of the target gas density. The latter method operates by varying the number of collisions that the parent ion and, possibly, fragment ions are likely to undergo [7]. II 1 119 Another aspect of CAD that affects the efficiency is the propensity for the target to charge exchange with the parent ions. The reaction is: >m1+N* (5d) l1*+N McLafferty et a1 [8], in a study on CAD, observed that the efficiency of the process was related to the ionization potential (IP) of the target gas. The higher the IP,‘the less energetically feasible charge exchange, and hence the more favored the competitive process of dissociation. It was found that ionized methane shows a neutralization cross section that is approximately linearly related to target ionization potential. They conclude that if the desire is to minimize ion current due to charge exchange, one must choose a target gas with a large ionization potential. Therefore, of the gases studied, helium provides the least amount of neutralization. A third important detriment to CAD efficiency is scattering. The general equation is: ln* + N > s1+ (scattered) + N (5e) In this process an ion is scattered beyond the acceptance angle of the instrument. Two trends were observed in a study [4], when argon ions and methane ions are subjected to target gas interactions: (1) the amount of scattering increases with collision gas pressure and, (2) the relative amount of scattering increases with an increase in target 120 ass and/or size. It was determined to use a target gas of saall size and of low mes such as helius. Thus, it has been deterained that helim offers the most benefits for high energy CAD efficiency. The reasons for CAD inefficiency are the large abundance of scattering and charge exchange relative to dissociation. There is a greater efficiency with heliu because these two factors are attenuated. 121 um GAB 00901113011 In order to investigate the target gas effects, nitrogen wu med in cell A under the same conditions as heliln. In this experiment, the detector gain was calibrated and raised to 7 for the daughters then lowered to 4 for the parent ions (with nitrogen renoved). At a pressure of 4.9 x10" torr as indicated on the Penning gauge (actual for nitrogen) the CAD efficiency was 1.7:. The trend is roughly the sale as with helitn, producing a maxinn in the vicinity of 5 x10‘5 torr. The efficiency rises gradually to the maxinn and then drops off substantially (3 CAD efficiency = 0.93 at 6.5 x10"5 torr).. This result agrees with earlier results obtained by McLafferty et al [8] in ' that the efficiency, at high energy, decreases as the molecular weight of the target gas increases. Also, with the gain at 7 for the daughters, the CAD fragment of less 13 could be observed in addition to those of mass 14 and 15. When cell D was employed the maxinn efficiency was 1.18. Another gas, deuteritn, was chosen to observe the effects on the CAD efficiency of a diatcnic with the sane mass as helitn. Deuteritn has two rotational and 1 vibrational degrees of freedom whereas a single etc. of helitn has neither of these. This fact may be responsible for effecting the overall cross section. 122 CELL A The resulting plot of the experimental results of'methane CAD with deuterium under the same conditions as for nitrogen above, is shown in Figure 5.8. The maximum efficiency of 2.22 occurs at a lower Penning gauge pressure. This efficiency is slightly less than that achieved with helium and could possibly be due to increased scatter or charge exchange effects. Collision cell D was also investigated using deuterius collision gas. An interesting result is that the efficiency maximum was approximately twice that achieved with deuterium in cell A, but still less than that achieved with helium in cell C. A plot of the data is shown in Figure 5.9. McLafferty et al [8], noticed an increased efficiency due to deuterius at low relative kinetic energies compared to those with helium. Collision cell A may not have allowed for a sufficient ion density along'the slit axis. Since cell D had a reduced slit size, there may have been fewer collisions or scattering outside of the cell. From the experiments with heliul, it was shown (see above) that cell A required a higher pressure than cell 8 to produce a sufficient number of fragments. With deuterium, at the approximate pressure of the maximum 123 for cell A, the collection and fragmentation efficiencies both improved and so does therefore, the overall CAD efficiency (see Figure 5.10). 124 METHANE MOLECULAR ION WTARCT CELA PERCENT ATTENUATION 40.00 50.00 60.00 70.00 80.00 I L 1; >- i C; 0 g 2.00 4- E . 1.1.1 3 '— Z 1 8 1 a: 1.50 -- 1 DJ D. L l L _l_ l___ ‘ LA— 1000 r I“ r I I 2.00 3.00 4.00 5.00 6.00 PRESSURE (TORR x10 +5) Figure 5.8 Methane CAD with deuteritn target gas and cell A. 4.50 4.30 4.10 3.90 3.70 3.50 3.30 CAD EFFICIENCY 3.10 2.90 2.70 2.50 45.00 125 METHANE MOLECULAR ION “LLB . 50.00 55.00 60.00 65.00 70.00 75.00 1 -__ _-_ ..---- -..--.__---. ...... ,1 . - +~ 1 1— . «1 .1— ‘— ‘- d— -_ / / 1P / .11... / l l l L 1 I L k l l l f l r l l l 1 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 PRESSURE 02 (TORR X10 +5) Figure 5.9 Methane CAD with deuteritn target gas and cell 8. 126 METHANE MOLECULAR ION Illflflllflfllil(XILECHONIAND GNDIETIIDUIEB IllL.I L00'n- 1.00 ~1- 1.70 -- 1.00 A 1.00 -J 1.40 4— 1.30 — 1.20 d- 1.10 —b ..Ih fl 0A0 -r- 1 1 1 1 1 11 1 1 1 11 “-39 r I 1 l 1 I 11 I 1 [fl wamawwmmewuomwwamw PRESSURED2 CTORRXIO-l-S) LOGEFFICIENC‘Y §§§§§§ l I Figure 5.10. Frauentation, collection and overall CAD efficiency curves for cell 8 with deuteritn as the target gas. 127 GILL C Collision cell C has yet to be tested with deuterium. In addition to the most important aspect of CAD efficiency, the means by which each cell deals with pressure with respect to the mass spectrometer is important as well. It was very pleasing to find the differentially pumped cell far superior in removing the target gas from the systes. A series of data is shown in table VD. As the pressure on the Penning gauge rises, the pressure at the backing pump. of the large diffusion pump rises as well. It is desireable to keep this backing pressure < 0.1 torr otherwise the large diffusion pump may send some of its oil back to the analyzer region of the mass. spectrometer. Table 5.2 shows that when we are colliding analyte ions with target gas at 5 x10'5 (Penningbfor helium), the backing pump rises to approximately 50 millitorr. This is very good when compared to the values from celle A or D. The theoretical analysis in chapter four closely follows the results obtained. In theory, for cell C, the small additional diffusion pump positioned at a specific distance from the call should exhaust approximately four times the entering gas as does the large diffusion pump. In reality it is more like a 2:1 ratio which may be due to inefficient exhaustion of the gas to the 128 Till! 5.2 BACKING PUMP COLLISION PENNING FORE VACUUM DETECTOR FOR LARGE CELL PIRANI DIFFSTACK DIFFUSION (micron) Ctorr) (micron) (torr) (micron) 4 - 2.5x10‘5 12 2.8x10‘7 8 4 7.0x10‘6 15 5.0x10‘7 12 5 2.2x10“ 17 7.8xlO'7 18 6 3.03410“9 21 1.0x10“5 39 9 4.01410'3 34 2.0x10‘5 18 II 5.0x10“ IO 2.6x10‘5 62 ‘ 14 6.011105 50 3.0x10‘5 Table 5.2. Pressure changes within the system as the target gas pressure is increased. 129 forevacuum m. When the small diffusion ptnp is closed (eiailar to employing cell A or E) the pressure climbs to ) 0.1 torr on the backing pump at this pressure. Of course, this valve is only closed for a short time at these target gas pressures. Another point that should be noticed from table 5.2 is that the earlier assumption of molecular flow is a good one. The collision cell gauge is situated above the reaction region and it indicates approximately 10 millitorr only at the higher collision pressures. Also, experiments on phenetole indicate that effective production of daughters occurs at a lower value on the Penning gauge and therefore the overall pressure is much lower, further reducing pump difficulties. TAIGIT’GAS DEPENDENCE Pressure measurements aid in distinguishing single collision or multiple collision reaction products [9]. If a plot of In P/Pb versus target gas pressure yields a linear relationship, first order kinetics are followed. Figure 5.11 shows that, indeed, single collisions are occurring in cell A over the range of pressures tested. This is important for the comparison of CAD efficiency results with those obtained by other investigators. Usually single collision conditions are specified as standard. 130 11.1 0 Z --3, LLI 1n 0 e Z L11 1' a] .. c3 --§ 0) ' 3’: <5 . . 4. o (D; o 2 1—5 ° 5 1.11 "§ E3, 05 w 25 . 1—33 Q 0.. 1:1: 2 s 11.1 I“ Q 0: .. O l.— 12 LE- -— § § §. §. §. «82 L'- ° ‘1’ '1‘ '1' ‘1‘ 1‘ [aosanoaad ONIHELNBNHOSHOOBHd 0N111x3] N'l Figure 5.11 A linear relationship is shown for the methane experiments indicating single collision conditions were maintained. 131 MEANS ANALYSIS The fourth major category in the evaluation of cell performance involves a comparison of the results of an MS/MS analysis with a chesical fragmentation example in the literature. The goal of this endeavor is to establish qualitative agreements between data fros this system and that of others. In this way the performance of the cell with regards to resolution and sensitivity can be qualitatively assessed. The cospounds exasined were n-butylbenzene, methanol, ethanol , 5-nonanone, n-decane, and phenetole. Him Nebutylbenzene was chosen because the fragmentation of its molecular ion has been studied both by photodissociation and CAD [10,11]. With this comparison, one can determine the energy acquired by the ions interacting with target gas in a collision cell. The reaction of concern is: Ciofli4*- ----> C780’ + CaHv- or (6a) m/z = 91 --—--> Gina’- + 0330 (083-03432) (&) n/z = 92 The energy acquired represents the average increase in the internal energies of ions as the result of collisional excitation. Calculations by Kim and McLafferty [6] on the deposition of energy 132 into the ion in collisional excitation suggest that the average increase in excitation energy is considerably less than 10 eV for 8 keV ions. It was found that the relative magnitudes of the peaks due to the 91+ and 92+ ions are sensitive to the photon energy used [12]. Since these experiments have provided a sensitive and accurate measure of the effect of changing the ion’s internal energy, the ratios of the peak heights resulting from the above competing reactions for collisionally excited ions are measured and compared with the photoexcitation data. P. Brown has determined that reaction 6a and 6b have critical energies of the order 2.6 and 1.4 eV respectively [13]. The rearrangement product 92+ is observed as a metastable with no 91+ present. Figure 5.12 shows the 92+ metastable of the molecular ion of n-butylbenzene. As the pressure inside the collision chamber increases, more energy is transferred to the molecular ion and simple cleavage of the C387 moiety begins to occur (see Figure 5.13). By measuring the peak height ratios at various pressures, a direct comparison to photoexcitation energy can be made (see Figure 5.14). It can be estimated that the average input of excitation energy is between 2.2 and 2.4 eV. Griffiths et a1 [10], obtained his data with 5 keV molecular ions and the results here are from 3.5 keV ions. An additional experiment by Griffiths et al proved that with helium target gas, the peak height 133 3:. .... use; at: 0:56.44. 8: 22 one. nae. 8.: ...3 can nae 8e r . . _ a b . . r e on a . r. 3 4 _ ... 3 k r. as can. u. ~ F 2: 3:. one :J ..5 gammammdknfi NO e+h:e1:me:5m12 .SLDZL ’4 Figure 5.12 Raw data plot indicating the 92 metastable decomposition product of the molecular ion of n-butylbenzene. 134 N—BUWLBENZENE’HE mo 91 AND 92 DAUGHTERS .°=v 2,2 x10" 1: 3475 Intensity S .4 I I p... 2.0 x10'5 1: 2000 IA} 1 100 50} ° r h A1 I T 213‘ § 100 P' 3-0 310" I= 1437 3‘ '5' so :5 .° I j 100 P-3.8 110" I: m 4? g 50 :5 o I I 100 P" 4.2 x10" I: 514 5‘ § so E ° 1 l 000 020 050 070 1000 1020 1000 1070 1100 1140115310 FIELD (DAC 01075) Figure 5.13. Raw data plots indicating the progressive increase in the 91 daughter of the n-butylbenzene molecular ion with collision gm pressure. 3+ e. .x «mob Banned and 8% 8n SN 8. w p H O.«.l| 0.81 t. «I 135 :3 Rho-l 8a asuuoxu Boa—(...! gal-«swung MOP/E HIDE: x HI 3 . . Cd3 // + CH“ m/B=UZ Figure 5.15 Frag-entation ache-e for 5-nonanone. 139 PHENETOLE ‘00 - 80 60 40 lntenshy 20 4,14.l .1 Al 1 ° 'I'I'I'r'r'l-F'I'I'I'I'fi mo HELI‘JM AT 2.5 X10-5 TORR PARENT 122 {=103220 so -I so- Intensity 20 - ° ‘I'III'I'I‘I'I‘III'I’IIj 00 UNATTENUATED PARENT 94 I=154620 1 so so Entensity 1 11.1 11 .l HEUUM AT 2.5 X10-5 TORR l=39840 q. 100 e.- 80 60 4O Intensity l l l l I l l A I 20 20 40 60 80 300 120 140 Figure 5.16. Daughter scene of n/z 122 and 94 parents of phenetole with helium ee the target gee. 140 {reputation pathway is enhanced. The Cefiot- parent ion fraguents to I]: 66 as a.-etastable deco-position. When collision gas is added, the intensity of 66 doubles relative to the metastable peak. (This plot is unrealized to less 66.) Peaks at a/z 55, 51, 40 and 39 are also observed. A depiction of the frag-entation pathway [15-17] is shown in Figure 5.17. ' Maquestiau et a1. [15] elployed a double focusing instrulent with air as the target gas. Abundances were recorded relative to the total ion current excluding that at ll: 66. They observe a peak at n/z 65 in a high relative intensity whereas I do not observe this peak. Also, I do not observe the ion of nass 98 with CAD. The unattenuated 93 ion represents 0.7x of the total ion current when heliul is not present. Nevertheless, I do observe nasses 55, 61, 40, and 39 of which 55, 51 and 39 are observed by Mbquestiau to as low as 53 of the total ion current after 66 has been subtracted. 141 +. /O — CH2 - CH3 m/a = 122 +_ \H’ +. OH / / . . . o ‘3 fl + CH2 CH2 m/a 3 9“ isomerIXes -CO 4‘6 unsaturated———+ m/a» 66 Aldehyde. “Suggested Structure E5] Figure 5.17 Partial frag-entation schene for phenetole. 142 CONCLUSIONS _ To fully asses the newly designed cell’s overall perforuance, an analysis of the data as well as an evaluation of instruuent control benefits and upgrade is necessary. The four saJor categories, sentioned earlier, can be discussed in light of the data recorded. In addition, with the construction and inplelentation of an adequate collision cell, TRIIB can be more fully utilized as an [IS/MS analytical tool. The calculation of the CAD efficiency indicated that cells A and 8 produced poorer results than that of collision cell C. For exauple, with heliul target gas cells A, B and 0 produced respective efficiency uaxisa of 2.8, 2.9 and 5.5 X. All three cells yield the sale trends as found in the literature for the efficiency curves (8c, Br and Solo). With cell C the 16-->12 frag-eat is observed indicating that energy deposition is sore effective. Efficiencies for methane CAD achieved by McLafferty indicate that this cell is perforning quite well. McLafferty et a1 [9] have elployed a heliuu beau collision interface as well as 10 kV ions. A direct colparison is not possible, but it is interesting that an efficiency of 68 was noted at 10 kV, whereas the results here (at approxiuately the sale beau attenuation of 603) are at 3.5 kV. Since higher CAD efficiencies are possible at higher accelerating potentials, this indicates that the perforsance of this cell is good. 143 In addition to the fact that varying the target gas changed the spectra in a predicted manner, the pressure effects agreed with the theory quite well. As discussed above, the conductance theory was obeyed to a large extent. The design facilitated target gas passage and proved to be quite effective in renoving the gas load (see table 5.2). The final aspect of the collision cell design concerns the quality of the IKE/DB spectra produced. The n-butylbenzene data indicate that the 91* and 92* daughters of the nolecular ion can be partially resolved. The daughter ion resolution depends to a great extent on the ion source but, nevertheless the cell’s capacity to ‘ induce this frag-entation at an energy eiailar to that achieved by others, is pleasing. Although sensitivity is also very nuch source dependent, one can deduce frol the 5-nonanone and phenetole data that the sensitivity is quite poor. The two most intense peaks in the daughter spectrul of 5-nonanone are recorded in table 5.3. The data fro. phenetole is very prouising though. Masses 65 and 93 could possibly be present, but the width of the peaks at aVz 66 and 94 most likely prevents their detection by the data systel. These data prove that the sensitivity varies for different cospounds and is most likely due to the CAD processes (energy deposition, etc.) occurring. Overall, the cell has performed adequately. The current source pulsing inple-entation (see J.T. Stults, Ph.D. thesis, DBU) presents proble- but, these may be overcoee in the future. 9. 10. 11. 12. 13. 14. 15. 16. 17. 144 mm Yost, R.A.; Enke, C.C.; McGilvery, 0.0.; Snith, D.; Morrison, J.D. Int. J. Mass Spectral. Ion Phys. 1979, 30, 127-136. Glish, 0.; Todd, P.J. Anal. Chen. 1982, 54, 842-43. Todd, P.J.; McLafferty, P.W. Int. J. Mass Spectron Ion Phys. lmlg 38, 371-3780 Lara-ea, J.A.; Cueron, D.; Cooks, R.G. J. An. Chen. Soc. 1981, 103, 12-17. . Cooks, R.G. Collision Spectroscopy, New York, 1978. Kin, 31.3.; McLafferty, P.W. J. As. Chen. Soc. 1978, 100, 3279. Ouwerkerk, C.B.D.; McLuckey, S.A.; Kistesaker, P.G.; Boerboon, A.J.H. Int. J. Mass Spectron. Ion Phys. 1984, 56, ll-l3. McLafferty, F.W.; Dente, P.F. III; Kornfeld, R.; Tsai, 8.; Howe, I. J. Am. Chen. Soc. 1973, 95, 2120. McLafferty, F. W. , Tandu Mass Spectraetry, John Wiley and Sons, New York, 1983. . Griffiths, I.W.; Mukhtar, 8.8.; March, 11.3.; Harris, F.M.; Beynon, J.8. Int. J. Mass Spectroa. Ion Phys. 1981, 39, 125-132. Harrison, A.G.; Lin, M. 8. Int. J. Mass Spectrou. Ion Phys. 1983, 51, 353-356. Mukhtar, 8.8.; Griffiths, I.W.; Harris, F.M.; Beynon, J.8. Int. J. Mass Spectron. Ion Phys. 1981, 37, 159. Brown, P. Org. Mass Spectrom. 1970, 3, 1175. McLafferty, F.W.; Kornfeld, P.; Haddon, W.F.; Levsen, 3.; Sakai, 1.; Bente, P.F. III; Tsai, 5.; Shuddesage, 8.0.8. J. AI. Chen. Soc. 1973, 95, 3886. Mequestiau, A.; Haverbeke, Y.V.; Fl-ang, 8.; DeMeyer, 0.; Das, R.G.; Reddy, 0.8. Org. Mass Spectroa. 1977, 12, 631. Borchers, P.; Levsen, 8.; Theissling, 0.3.; Nibbering, N.M.M. Org. Mass Spectrom. 1979, 12, 746. Maquestiau, A.; Huang, 3.; Pauwels, P.; Vallet, P.; Meyrant, P. Org. Mass Spectron. 1982, 17, 643. CHAPTER VI With the addition of an adequately functional collision cell, the future of TRIMS is bright. MS/MS information can now be readily achieved. The phenetole exalple in chapter four provides direct evidence of this fact. Previously, the CAD products of the aolecular ion of phenetole were not observed (personal conuunication with Dr; JIT. Stults), but with the improved pulping efficiency CAD spectra can now be recorded. Sole remaining difficulties and possible solutions to the. are given below. PRESENT PROBLEMS WITH TREES Of course, sensitivity and resolution are two parameters which every mass spectrosetrist, or analytical che-ist for that latter, would like to maxiaize. Unfortunately, the two have an inverse relationship to one another and, at best, experinents aredevised in which, one of the two is maxinized depending on the type of 145 146 infer-ation sought. TREMS in its present configuration of superilposed analyzers, unfortunately, cannot escape the clutches of this ever present situation. With the current pulsing technique (source trapping followed by an extraction pulse) both resolution and sensitivity can vary. A couplication with TREMS is that it is affected by both the magnetic field resolution and the time-of-flight resolution. Good resolution in both are required for accurate energy independent sass assign-ant. The total energy an ion receives frou acceleration is a function of the extraction voltage for a fixed accelerating voltage. With a higher extraction voltage (e.g., 100 V pulsed negatively fros 3500 V) there results a decrease in the available energy to isobaric ions. A decrease in energy leads to a decrease in velocity. This high extraction pulse induces a larger energy spread and a corresponding velocity spread which ultilately lowers the magnetic field resolution. Simultaneously, at the other "corner" of this hypothetical data field an increase in the extraction voltage decreases the flight time and reduces the flight time spread because the ions spend less tine in the source. This results in an increase in the tine-of-flight resolution. The paragraph above illustrates the effect spatial parameters of the ions has on velocity and tine ueasureuents. The initial energy distribution turns out to influence the velocity only slightly when coupared to the spatial effects. The turnaround time (the difference in flight tines between ions with different initial energies) is the most significant contributor to the flight time difference. Thus, at 147 any one magnetic field strength, the peak width in time is duel-ainly to initial ion kinetic energies. Therefore, as the extraction voltage increases the turnaround time decreases and the flight tine peak widths decrease. I These events occur in reverse for low extraction voltages. In this case, more energy is available to the isobaric ions and therefore their velocity increases and the spread in energy is reduced. This results in an increased magnetic field resolution. ‘The time the ions spend in the acceleration region increases, the turnaround time increases, the peak width in time increases and consequently a decrease in the time-of-flight resolution results. BIN! DIFLICTION In order to circumvent these conflicting effects, beau deflection could be employed [1,2]. The flight time would be a direct measure of the velocity and no space focusing would be required. The ion source could be operated at the normal extraction voltage (10 V) to maxinize the magnetic field resolution. The time-of-flight resolution would depend on the duration of the intersection of the deflected beas'with the limiting aperture; the width of the resulting ion packet in time can be made quite short. The location for beam deflection will be at the exit slit of the LED 9000. This offers the advantage that the ions can be deflected 148 across the width of the exit slit. This avoids difficulties with pulsing at the entrance slit which must be along the long dimension of the slit image to avoid ion movement in the nonhomogeneous fringing field of the met. The simultaneous nature of the suentum-flight time measurement would be renoved, but the physics of the mass separation and mass assignment rennin completely identical. With this configuration both.magnetic field and flight time resolution can be optimized, but sensitivity still remains a ”formidable foe”. The sensitivity of this technique will decrease relative to that of the trapping method. The addition of the new collision cell has improved the ability to detect ions that were not observed previously (e.g., 16-->12 fragment in nethane). One of the limitations of the present instrument is the low amplification of the signal, forcing the gated integrator to be operated at its 50 mV sensitivity limit. Better amplification would relieve the strain on the gated integrator,'but large gains are difficult to achieve at the large.bandwidths necessary for amplifying the narrow thae-of-flight peaks. .‘ THIIPAIBAY’DITDCTTOI Time array detection (TAD) [3] should provide a significant improve-ant in the sensitivity for certain experiments. It is hoped that a phase I implementation of the integrating transient recorder (1TB) will be functional by the summer of 1986. If a version can be 149 built for the CVC time-of-flight, than a means of either moving it or building another one will be investigated. While the ITR is being built and optimized, there is still plenty to do on the TRIBE instrument. It is planned that initially a single set of deflection plates will be constructed. A pulse generator with a 5 us or less rise time will be necessary, supplying pulses at a rate of 10 kHz. During this construction period, software improvcsats can be made to the data systen. A hardware smoothing routine, and a new FORTH processor built by NOVIX is being investigated [4]. NOVIX has built the hardware to serve the software and not vice versa as is co-only found. This is an expensive venture though and say not be realized. One must not forget the overall picture that represents the analytical problem many of us here at MSU are trying to solve. This problen centers on the fact that with the advent of high speed chro-atographic systems, detection systems have been hard pressed to collect the couplets Its/m data field from transient sanples. Capillary. chronatography peaks are generally 1 to 2 seconds in width and this narrow time interval can strain lost detectors to yield any kind of significant data. Of course, the problen with the deluge of data that will inevitably arise with the ITR also must be overcose. 150 Laser desorption [5] is a technique that may be particularly advantageous with TRIMB. The feasibility of this is being investigated presently. 151 Im BIkkOl', J.Mene Jo Pb”. Rs 1973, 6’ 785-7890 Dakker, J.M.B. J. Phys. E. 1974, 7, 364-368. Holland, J.F.;Enke, C.C.;Allison, J.;Stu1ts, J.T.;Pinkston, J.D.;Newco-e, D.;Watson, J.T. Anal. Chen. 1983, 55, 997A—1112A. Murphy, R.W. Journal of FORTH Appl. and Res. 1985, 3, 185. Conzenius, R.J.;Capellen, J.M. Int. J. Mass Spec; Ion Phys. 1980, 34, 197-271. E "'IIIIIIIIILTIIIIIIIIIIIIIIIIIIJIII“