an investigation of site-selective gas-phase reactions of amino … · 2017. 1. 29. · an...

An Investigation of Site-Selective Gas-Phase Reactions of Amino Alcohols with Dimethyl Ether Ions

Erika S. Eichmann and Jennifer S. Brodbelt Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, Texas, USA

The reactions of dimethyl ether ions with neutral amino alcohols were examined in both a quadrupole ion trap mass spectrometer and a triple quadrupole mass spectrometer. These ion-molecule reactions produced two types of ions: the protonated species [M + l]+ and a more complex product at [M + 13]+. The abundance of the [M + 13]+ ions relative to that of the [M + I]+ ions decreases with increasing formal interfunctional distance. Multistage collision-activated dissociation techniques were used to characterize the [M + 13]+ product ions, their reactivities, and the mechanisms for their formation and dissociation. In addition, molecular semiempirical calculation methods were used to probe the thermochemistry of these reactions. Reaction at the amino alcohol nitrogen site is favored, and the resulting [M + 13]+ addition products may cyclize for additional stabilization. Comparisons were made among the behavior of related compounds, such as alcohols, diols, amines, and diamines. The alcohols reacted only to form the protonated species, but the diols, amines, and diamines all formed significant amounts of [M + 13]+ ions or related dissociation products. (1 Am Sot Mass Spectrom 2993, 4, 230-241)

S ubstituent effects, which may be due to steric hindrance within a molecule or to electronic interaction between two or more functional groups of

a molecule, have been shown to play important roles in the outcome of organic reactions, both in solution [l-4] and in the gas phase [5~8]. In the condensed phase, the nature of the solvent can have a significant effect on the reactivities of molecules [9, 101, whereas these interactions are typically absent in the gas phase. In fact, gas-phase reactions have been studied for ions with varying degrees of solvation, and the results have shown that coordination with even a single solvent molecule substantially changes the chemical properties of the ions [ll]. Thus, more information about the intrinsic behavior of the molecules can be obtained from gas-phase studies [ 121.

Several studies of the gas-phase chemistry of di- functional molecules, such as amino alcohols [13], dials ]14], diethers [15], and dicarboxylic acids and their derivatives [16, 171, have been reported. In these studies, as well as several involving molecular orbital calculations [18, 191, correlations of the functional group interactions with basicities, ion stabilities, and solution-phase thermodynamics were made. In these

Address reprint requests to Jennifer S. Brcdbelt, Department of Chemistry and Bimhemistry, University of Texas at Austin, Austin, TX 7871%3167.

cases, the favorabilities of competitive reaction and subsequent dissociation or rearrangement channels showed marked changes with variations in the nature and separation of functional groups within the molecules. For instance, it was noted [ZO] that the extent of dehydration of simple protonated c(, w-amino alcohols was largely dependent on the length of the carbon chain and that internal interactions prevented ammonia loss from compounds with short interfunctional distances. The competitive dehydration and deamination processes were shown to proceed via cyclization of the thermodynamically favored products. Under collision-activated dissociation (CAD) conditions, however, the dominant products were found to be those formed by the pathways of lowest activation energies [20, 211.

To date, the majority of gas-phase ion studies concerning functional group interactions have involved the interpretation of fragmentation patterns of open- shell ions formed by conventional electron ionization [22-291 or the evaluation of the dissociation of closed- shell ions generated by simple proton transfer reactions [30-371. Few studies of the effects of functional groups on more complex ion-molecule or dissociation reactions have been reported. From an analytical per- spective, however, an understanding of such effects is critical for the development of more sensitive, selective, and structurally informative chemical ionization techniques and for the interpretation of dissociation

0 1993 American Society for Mass Spectrometry 104&0305/93/$6.00 Revised October 28,1992

Accepted November 2,1992

J Am Sac Mass Spectrom 1993,4,230-241 SIT!-SELECTTVEGASPHASEREACTIONS 231

patterns of increasingly complex molecules. As an initial effort toward this goal, we undertook a study of the bimolecular and dissociation reactions of amino alcohols and related compounds. In addition to allow- ing characterization of the fundamental nature of substituent interactions, these studies provided a basis for future evaluations of the gas-phase chemistry of more complex, biologically active amino alcohol systems, such as the antibiotic drugs spectinomycin [38, 391 and kanamycin [40, 411.

The reagent chosen for this preliminary work was dimethyl ether (DME), selected primarily because it was previously shown to have striking functional group selectivity as a chemical ionization reagent for other organic systems [42]. The reactive ions formed are protonated DME, at m/z = 47, and the methoxymethylene cation CH,OCH:, at m/z = 45. Ion-molecule reactions between selected amino alcohols (Ml and these DME ions in an ion trap mass spectrometer or in the conventional source of a triple quadrupole mass spectrometer typically produced [M + l]+ and [M + 13]+ ions. The [M + 13]+ ions result from methylene substitution by the DME ions at a heteroatomic site [43], whereas the [M + l]+ ions are formed by simple proton transfer from the reagent ions. To determine the structures of these ions and elucidate mechanisms for their formation, each product ion of interest was examined by CAD [44]. This technique involves selection of a particular ion, followed by acceleration and collision with an inert target gas to induce dissociation of the ion. The fragmentation patterns observed can then be used to aid in the structural characterization of the precursor ion. In some cases, a second stage of CAD was possible and was likewise helpful in the characterization of the fragment ions formed in the first stage of activation. To supple- ment the information obtained in this manner for the amino alcohols, the analogous bimolecular and dissociation reactions of a variety of amines, diamines, and diols were also studied. In some cases, compounds that suitably modeled the product ions in question were commercially available, and comparisons to au- thentic structures were made. Finally, we used semiempirical energy calculations to model the thermochemistry of the reactions in question. Although the values obtained are only rough estimates, they provide a reasonable qualitative basis with regard to the ener- getics of the reactions considered in this study.

Experimental

All organic substrates were purchased from Aldrich Chemical Co. (Milwaukee, WI) (purity 297%) and used without additional purification. DME was ob- tamed from MG Industries. Ion-molecule reactions were performed in the chemical ionization (CI) source of a Finnigan triple-stage quadrupole mass spectrometer 0) and in a Finnigan ion trap mass spectrometer (ITMS) (Finnigan-MAT, San Jose, CA). The CAD re-

sults obtained from both instruments were qualita tively and quantitatively similar; however, the initial abundance of the ion-molecule reaction products was generally greater in the higher pressure source of the triple quadrupole instrument. Because the results from both instruments were similar, the discussion of all other points is based on the results obtained from the ITMS.

For the TSQ experiments, all samples were admitted to the ion source on the tip of a solids probe, absorbed to a small plug of glass wool. The DME reagent gas pressure was typically approximately 1.5 torr. Argon was used as the collision gas in the second quadrupole, and typical pressures were l-2 mtorr. These pressures result in multiple-collision conditions. Collision energies ranged from 1.5 to 7 eV in the laboratory frame. These experiments were performed primarily to demonstrate the probable origin of the [M + 13]+ product ion.

In the ITMS experiments, samples were admitted either through a leak valve or on the tip of a solids probe, and typical sample pressures were 1 X 10e6 ton. The nominal DME pressure was 1 X l@ torr, and helium buffer gas was admitted to 1 mtorr. A low-energy ( 5 70 eV) electron beam was applied for 0.5 ms to ionize the CI reagent, A variable reaction period (O-100 ms) following the ionization pulse was used to allow ion-molecule reactions to occur between the reagent gas ions and the organic substrate. The radiofrequency (rf) voltage applied to the ion trap was typically 300 V,,_,, during these two intervals. A selected ion was then isolated by application of appro priate rf and direct current (dc) electric fields [45]. During this interval, typical dc voltages were -40 to -100 V, applied for 5 ms, and rf voltages were generally 300-900 V,, r), applied at a frequency of 1.1 MHz. The selected ion was then allowed to react for a variable amount of time (10-100 ms) with the neutral species present in the trap, and the resulting products were observed. Alternatively, the isolated ion was activated at its resonant frequency to induce dissociation for further characterization. The supplementary activation voltage used for CAD experiments was generally 0.5 V(r_r, and applied for 5 ms at the axial frequency of the ion (Le., 155 kHz for m/z = 88 at 4, = 0.4). Quantitative information is reported to the nearest 5% to account for the average variation in results. In all cases, at least 100 scans were averaged to obtain the final values.

All thermochemical values that were not available in the literature [46] were estimated by using a combi- nation of molecular modeling and semiempirical calculation methods. The molecular modeling program PCMODEL (version 4.4) and the semlempirical program MOPAC (version 6.0) were purchased from Serena Software (Bloomington, IN1 and run on a Mac- intosh IIsi personal computer. The semiempirical program AMPAC was run on the University of Texas VAX system. Preliminary structure minimization was

J kn %c Mass Spectrom 1593,4,230-241 SIT!-SELECTIVE GA!?-PHASE REACTIONS 233

structures and the possible mechanisms for their formation were examined. Because two reactive ions are produced from DME, the first step in this mechanistic study was to determine which ion gave rise to the [M + 13]+ ion. Two reactive ions, C,H,O* (1) and C,H,O+ (2) are initially formed when DME is ionized. The ion-ejection capability [45] of the ITMS was used to isolate each of these in turn by ejecting the un- wanted ions. They were then each allowed to react with the neutral amino alcohol substrates. Whereas reaction with 1 gives only the protonated substrate, reaction with 2 gives a mixture of a small abundance of [M + l]+ ions and a larger amount of [M + 13]+ product ions. Thus, 2 is the reactive species of interest in this study of the formation of [M + 13]+ ions.

H,CGO% 2 1 2

This result is confirmed by examination of the CAD spectra for the adducts of 2, with the amino alcohols occurring at [M + 45]+. In the quadrupole ion trap, this adduct is typically not seen, presumably because it is too energetic to survive and spontaneously fragments to a more stable product ion; however, in the higher pressure source of the TSQ, the [M + 45]+ adducts are adequately collisionally stabilized prior to dissociation and can be further analyzed by CAD. As expected, mass selection followed by collisional activation of the [M + 45]+ adduct ion leads to exclusive

formation of [M + 13]+ products. This confirms that the reactive ion of DME is indeed 45+ and suggests that formation of [M + 13]+ occurs indirectly through an intermediate [M + 45]+ ion.

In accord with these results, two mechanisms for the formation of the [M + 13]+ ion are proposed. Al- though the structures are shown specifically for 3- amino-1-propanol (31, analogous mechanisms can be written for all of the compounds studied. The following reaction depicts the formation of the [M + 45]+ collision complex 4, which is presumed to be one possible precursor for the final product ion of interest: +

3 2 4

From there, two different paths can be envisioned for its dissociation to the [M + 13]+ product (Scheme I). Both involve nucleophilic interaction of the basic sites of the amino alcohol with the positively polarized carbon atom of the 45* ion 2. Path A describes one route in which the nitrogen atom attacks the positively polarized carbon atom of 2, resulting in the covalently bound species 5a. A similar path can be written in which all steps are the same, except that the attacking site is the oxygen atom (path B), and the structure obtained is 5b rather than 5a. Intramolecular proton transfer within 5 then gives 6, from which methanol is eliminated, giving the [M + 13]+ ion 7.

Listed in Table 2 are the heats of formation for the neutral amino alcohols, as well as for the lowest en-

r-l HO: $4J n

3 HCJ: +

UJ,,, - PNXH

0' -CH30H HZC

3 H ’ hi,

5a 6a 7a

5b 6b 7b Scheme I. Formation of [M + 131+ ions.

234 EICHMANN AND BRODBELT J Am SW Mass Spectrum 1993,4,230-X1

Table 2. Estimated thermochemical values for [M + 13]+ formation reactions”

Heat of formation Hear of reaction

[M+ 131+

Compound (MI Neutral [M + 451+ [M f 131’ [M + 451+ Stepwiseb Overallb

B-Amino-1-hexanol -67C 14 84 -78 22 -54

5-Amino-l-pentanof -75 21 94 -61 25 -36

4.Amino-I-butanol -57c 30 98 -70 20 -50

3.Amino-1-propanol -5Z5 37 106 -68 21 -47

2-Amino-l-ethanol -4t3= 46 117 -61 21 -40

2-(Methylamino)-ethanol -58 52 120 -47 20 -27

2-(Dimethylamino)-ethanold -49c 64 -44 -

aValuss expressed m kilocalories/mole and rounded to the nearest umt. bThe stepwise values represent the dissociation of [M+451+ already formed; the overall values

represent direct formation of the ion from M end 45 I, CExperimentat values (see ref. 21). dThe [M f 131+ ions cannot be formed in this case from the lowest energy [M + 451t ion. they are

only formed from rhe higher energy oxygen site adduel.

ergy forms’ of their [M + 45]+ adducts and the resulting [M + 13]+ product ions. From these values, the estimated heats of reaction for the formation of both the [M + 45]+ intermediate and the final [M + 13]+ product were calculated and are also listed.’ The last entry in Table 2 requires particular comment. In all cases, including that of Z_(dimethylamino)ethanol, the lowest energy [M + 45]+ adduct is the one formed by addition at the nitrogen site; however, as will be shown, the elimination of methanol from the 2-(dimethylamino)ethanol adduct cannot occur without extensive rearrangement, and the [M + 13J’ product ion thus cannot form.

The heats of formation for the [M + 45]+ adducts are considerably lower than those for the final [M + 13]+ product ions (see Table 2). This is consistent with the previous assertion that the [M + 45J+ ions are too internany energetic to survive in the ion trap. The formation reactions for these [M + 45]+ adducts are exothermic by 30-80 kcal/mol (see Table Z), and the excess energy resulting from the reactive collision remains in the adduct as internal energy, Unlike substrates with high internal energy in the condensed phase, where the solvent can act as an energy sink, high-energy ions in the gas phase must relax by other means. One common mechanism in the ion trap is collisional cooling, in which the energetic ion under- goes a series of deactivating collisions with a neutral buffer gas. In the case of the [M + 45]+ ions, this cooling effect is not efficient enough to remove the large amount of excess energy, and the ions spontaneously dissociate before they can lower their internal

1 See Table 9 and corresponding discussion for a more complete ;xplanation of the different forms of these ions.

Heats of reaction were calculated using the following reaction as a basis and using the values 158 and -48 kcal/mol for the heats of formation for the 45+ ion and neutral methanol, resoectivelv (see rcf 46):

& ,

M + +0 H,C/ QCH2 - (M+13f + CH#i

energy. Although the heats of formation for the final [M f 13]* product ions are substantially higher than those of the intermediate [M + 45]+ adduct ions, and the loss of methanol from these intermediate ions is endothermic by 20-25 kcal/mol (see Table 2),3 this reaction invariably occurs. As shown in the last col- umn of Table 2, the overall reactions between the neutral amino alcohol substrates and the 45+ ion of DME are exothermic by 22-55 kcal/mol. An energy diagram for the reaction of 3-amino-l-propanol with the 45* ion of DME is shown in Figure 2 and summa- rizes these points.

Competition Between Reaction Sites

Preliminary investigation of the competition between the two reactive sites was accomplished by studying the reactive properties of a series of compounds in which the functional groups were modified. First, three compounds in which the amine group is increasingly substituted was examined. Although the difference between 2-amino-I-ethanol (81, 2-(methylamino)-l- ethanol (9), and 2-(dimethylamino)-l-ethanol (10) is only in the degree of substitution, this difference was expected to dramatically change the product distribution of the reaction.

8 9 ia

In simple amino alcohol systems, the nitrogen atom is the most likely site of addition because of its high nucleophilicity [47]; however, if the amine functionality is altered so that the nitrogen addition route be-

30vrrall reactions at the oxveen atom are much less endothermic (10 to -5 kcal/mol) and mi hr therefore be expected to occur more readily, but the necessary P M + 451* formation reactlen is much less favored for this site (-25 to 45 versus -40 to -80 kcal/mol).

J Am Sac Mass Spectrom 1993, 4, 23OE241 SlTEeSELECTWE GASPHASE REACTIONS 235

Reaction Coordinate

Figure 2. Schematic energy profile for the formation of the [M + 131’ ion of 3-amino-l-propanol.

comes less favored, the extent of competition between the two paths will change. Although the resulting oxygen site addition product ions would still be observed at a mass-to-charge ratio of [M + 13]+, these effects should be manifested as differences in the favored dissociation pathways followed by the collisionally activated product ions.

The increasing substitution should cause the nitrogen atom to be more nucleophilic, but the added steric bulk of 9 is predicted to entropically disfavor reaction at the amine site. Thus, the distribution between the two possible products should shift to favor the oxygen site addition product, corresponding to 7b. In fact, on the basis of the proposed mechanisms, the highly hin- dered N,N-disubstituted compound 10 would be expected to give exclusively the oxygen addition product. In addition to the steric effect mentioned above, elimination of methanol from the nitrogen site [M +45]+ . mtermediate requires the presence of at least one available proton on the amine group, a fea- ture that the [M + 45]+ adduct of 10 does not possess. Presumably, the nucleophilic nitrogen atom would still be attracted to the polarized double bond of 2, and the initial ion-molecule complex intermediate corresponding to 4 may in fact form, but the subsequent reaction as shown in path A of Scheme I cannot take place. The nitrogen site addition product would thus make no contribution to the amount of [M + 13]+ ion observed.

H C=NA 2 i OH

lla (R = HI 12a (R = CH,)

lib (R = H) 12b (R’ = H; R* = CH,) 13 (I? = R2 = CH,)

Although protonated 8-10 show virtually identical dissociation patterns their [M + 13]+ ions show very different fragmentation behavior (Figure 3). For the [M + 13]+ product of 8, the major fragment ions correspond to losses of water and formaldehyde. Dissocia- tion of the [M + 13]+ ion of 9 yields predominantly formaldehyde and vinyl alcohol losses. Finally, the [M + 13]+ ion for 10 shows only Ioss of 30 Da (loss of formaldehyde). These observations are best rationalized on the basis of the ring-opened structures shown by 11-13 for these ions, corresponding to structures 7a and 7b.

The fact that water loss is the predominant fragmentation pathway for the adduct of 8 indicates that the primary site of the reaction of 8 with the DME ion 2 is the nitrogen atom and that the operative mechanism is thus the one described in path A of Scheme I. Unless significant rearrangement of the ion occurs prior to fragmentation, the loss of water cannot occur from the oxygen site [M + 13]+ product ion. The hydroxyl group must be free to accept a proton, forming a hydronium ion that is easily cleaved to give the dehydration product. Therefore, the predominant form of the [M + 13]+ product of 8 is the nitrogen site addition product lla rather than llb. Loss of formaldehyde, on the other hand, is most easily explained by direct cleavage of the carbon-oxygen bond of llb. It should be noted that the same loss can occur from lla by a different mechanism that involves a 1,5-hydride shift from the oxygen atom to the olefinic carbon atom (see below, Scheme IVb and corresponding discussion); however, this rearrangement is expected to have a much higher energy barrier than the simple loss of water and is not likely to significantly contribute to the dissociation of the ions. The relative abundance of the ion resulting from loss of formaldehyde, which ap-

Figure 3. CAD spectra for [M + l]+ and [M + 13]* ions of (a) 2-amino-l-ethanol; (b) 2~-(methylamino)-l-ethanol; and (c) Z- (dimethylamino)_l-ethanol.

J Am Sot Mass Spectrom 1993, 4, X30-241 SlTl$SELECTIVE GASPHASE REACTIONS 237

56+

* {M+ 13)+ 44+ II I

n (M - 15)f X9+ H9-J k_P& I

Figure 4. Comparison of CAD of model compounds to CM + 13]+ ions of (a) Z-amino-l-ethanol, and (b) 3-amino-1-propanol. The * peaks are due to [M + l]+ ions that are present in the CI spectrum of the amino alcohols before activation and remain unchanged on CAD. Because these peaks arise on storage of the [M + 13]+ ION (see text), they cannot be removed from the spectrum by conventional isolation techniques.

that the structures are able to interconvert by the mechanism shown in Scheme ITI is reduced by the less energetically favored cyclization of the nitrogen site [M + 13]+ ions necessary to form 19b.

Likewise, the values for the acyclic oxygen site addition products were difficult to obtain because their structures consistently cyclized during optimization (see Table 3); however, in these cases values could be obtained by using the fully extended structure for the calculation. It must be noted that because the extended form provides only a local energy minimum, no quantitative conclusions can be drawn from this information. These values are considerably higher than those for the cyclic structure, and any oxygen site [M + 13]+ product ions formed are thus expected to exist as cyclic ions. Also, contrary to the conclusions presented for the nitrogen site product ions, the conversion of oxygen site [M + 13]+ ions to their nitrogen site coun- terparts is facilitated by the predominance of the cyclic species corresponding to 19b.

According to these results, it seems that the predominant form of the [M + 13]+ ion remains the nitrogen site addition product, even under thermodynamic

7a 19a

A Proton Transfer (I

'0 NCH NHz

- (‘1 2

J Am Smc Mass Spectrom 1993, 4, 2X-241 SITE-SELECTIVE GA!+l’HASE REACTIONS 239

22b 2Ib

Scheme IV. Formation of the 44+ fragment ion

alcohol (entries l-3) and 2-amino-l-ethanol (entry 5) products are similar, but no dehydration peak is observed at all for the 3-amino-1-propanol ion (entry 4). This can be attributed to two main factors: the favorability of the proton transfer step and the degree of interaction between the nitrogen atom lone pair and the developing positive charge on the carbon atom bearing the group leaving. These effects are illustrated in Scheme V.

The proton transfer step for the dehydration reaction of the [M + 13]+ ions of all of these compounds most likely involves the loosely bound proton-bridged structure corresponding to 7, rather than the covalently bound cyclic structure having the general form of 19. Thus, the more the equilibrium between the two alter- nate structures favors the covalent form, the lower the favorability of the dehydration reaction will be.

The cleavage step of the reaction is presumably the more important in determining the final product dis tribution. The favorability of the carbon-oxygen bond cleavage after proton transfer presumably

depends on the degree of assistance from a nearby electron-donating substituent [48]. Interaction between the nitrogen atom and the nascent positive charge stabilizes the transition state and facilitates the dehydration reaction. Therefore, this pathway should be observed to a greater extent for ions that allow for more favorable functional group interaction. The ab- sence of the dehydration fragment from the CAD spectrum of 7 can be rationalized by the fact that even though the proton transfer reaction occurs to a significant extent, interaction of the nitrogen atom with the developing positive center would necessarily be a four-membered ring, as in 24. Thus, bond cleavage to 25 is not expected to happen readily. Conversely, the CAD spectrum of the 5-amino-1-pentanol [M + 13]+ ion 27 shows a large amount of the dehydration product. A highly stable six-membered ring transition state 28 is formed, and dehydration to 29 is favored.

Studies on Related Systems

To support our conclusions, simiiar experiments were performed on related compounds. Several generaliza- tions can be made concerning the reactions of various amines, diamines, alcohols, and dials with DME ions, and these can be related to the bimolecular reactivity and dissociative properties of the amino alcohols examined.

Results from our laboratory of more extensive studies of simple alcohols and diols are reported elsewhere [43]. In connection with the present work, several additional experiments were performed. Two simple alcohols, 1-butanol and 1-propanol, were found to give only protonation on reaction with DME ions. In both cases, sample and reagent pressures and ionization and reaction times were independently and collec- tively varied to promote [M + 13]+ formation, but no

1 1 ’ I HO’- NH

:N 0

C ‘-6

8a 7a 24 25

26 27 26 Scheme V. Loss of water from nitrogen site addition products.

240 EICHMANN AND BRODBELT J Am Sot Mass Spectrom 1593,4,230-241

evidence of this reaction was ever observed. In con- trast, 1,3-propanediol and ethylene glycol were examined in the same way and were found to give abur- dant [M + 131 + ions under the conditions typically used for the amino alcohol reactions. Thus, these sys- terns also show a strong dependence on favorable functional group interactions for promotion of the nucleophilic methylene substitution reaction.

fion (CHE-91226991, and an American Society for Mass Spectrom- etry Young Investigator Award are gratefully acknowledged. Generous assistance with the thermochemical calculations from Professor John E. Bartrness is also greatly appreciated.

References

Studies with monofunctional amines yielded very different results. Whereas the simple alcohols do not exhibit any reaction other than protonation in the presence of DME ions, ethylamine, propylamine, and butylamine all give significant amounts of [M + 13]+ product ions. In addition, the only fragmentation pathway observed for these product ions on collisional activation is the formation of an ion at m/z 44. These observations confirm several points concluded from the study of amino alcohols.

1. Hughes, E. D. Q. Rev. Chem. Sot. 1948, 2, 107. 2. Brown, H. C.; Fletcher, R. S. J. Am. Chem. Sec. 1949, 71, 1845. 3. Perez, J. D.; Phagouape, L. M.; Davico, G. E. 1, Phys. Org.

Chem. 1989, 2, 225. 4. Wolf, R.; Gruetzrnacher, H. F. Nau 1. Chem. 1990, 24, 379. 5. Mishima, M.; Inoue, H.; Fujio, M.; Tsuno, Y. Tetrahedron Lett.

1990, 31, 685. 6. Manion, J. A.; Louw, R. 1. Chem. Sot., Perkin Trans. 2 195%

551. 7. Hrusak, J.; Tkaczyk, M. Org. Mass Spectrom. 1990, 25, 214. 8. Jerkovich, G.; Hankovszky, H. 0.; Hideg, K. Org. Mass

Specfrom. 1990, 25, 67.

First, the fact that [M + 13]+ formation occurs with simple amines and not with monofunctional alcohols indicates that the primary reactive group of the amino alcohols is indeed the amine functionality. Whereas assistance from a nearby nucleophilic substituent is necessary for reaction to occur at a hydroxy site [43], such assistance is not required for the same reaction at an amine site. Second, the exclusive fragmentation of the amine [M + 13]+ ions to 44+, even without the presence of a hydroxyl group within the molecule, once again indicates that the oxygen functionality is often a spectator and that the amine group is the site with dominant reactivity.

9. Dyumaev, K. M.; Korolev, B. A. Russ. Chem. Rev. 1980, 49, 1021.

10. Aleman, C.; M&eras, F.; Uedos, A.; Duran, M.; Bertran, J. J. Phys. Org. Chem. 1989, 2,611.

11. B&me, D. K.; R&shit, A. B.; Mackay, G. I. 1. Am. Chem. Sac. 1982, 104, 1100.

12. Olmstead, W. N.; Brauman, J. I. J. Am. Chem SOE. 1977, 99, 4219.

13.

14.

15.

Hour&t, R.; Rufenacht, H.; Stahl, D.; Tichy, M.; Longevialle, P. Org. Mass Spectrom. 1985, 20, 300. Meyerhoffer, W. J.; Bursey, M. M. Org. Mass Spectrom. 1989, 24, 169. Morton, T. H.; Beauchamp, J. L. J. Am. Chem. SW. 1975, 97, 2355.

16. Harrison, A. G.; Kallury, R. K. Org. Mass Spectrom. 1980, 15, 277.

Likewise, the behavior of the diamines contrasts sharply with that of the diols. As observed for a majority of the diols [43], the predominant product obtained for the reactions of 1,2-diaminoethane, 1,3- diaminopropane, 1,4_diaminobutane, 1,5-diaminopen- tane, and 1,2_diaminopropane with DME ions is the protonated molecule; however, another product at [M - 4]* is also typically observed. This product is presumed to arise from spontaneous loss of ammonia from the [M + 13]+ product ion. In fact, the CAD spectrum of each [M + 13]+ ion shows only one fragment, corresponding to loss of ammonia. For further elucidation, the [M + 13 - 17]+ ion for 1,4-diaminobutane was isolated and energized in a second activation step, and its CAD spectrum was recorded. Only two fragments are observed: loss of ethylene and loss of 42 u. These two fragments are observed in the same relative abundance in the CAD spectrum of the original [M - 4]+ reaction product for the same compound. Thus, the initial assumption that this ion arises from spontaneous deamination of the [M + 13]+ ion is supported.

17. 18.

19.

20.

21.

22.

23.

24. 25.

26.

27. 28.

29.

30.

31.

Acknowledgments 32.

Financial support from the Welch Foundation (F-1155), the Petroleum Research Foundation (22270-G5), the National Insti- tutes of Health (ROl_GM46723-Ol), fhe National Science Founds-

33.

Weinkam, R. J.; Gal, J. Org. Mass Spectrom. 1976, 11, 188. Aue, D. H.; Webb, H. M.; Bowers, M. T. [. Am. Chem. Sot. 1976, 98, 311. Amett, E. M.; Chawla, 3.; Bell, L.; Taagepera, M.; H&e, W. J.; Taft, R. W. 1. Am. Chem. Sot. 1977. 99,5729. Wysocki, V. H.; Burimky, D. J.; Cooks, R. G. 1. Am. Gem. Sot. 1985, 50, 1287. Brodbelt, J. S.; Wysocki, V. H.; Cooks, R. G. Org. Muss Spectrom. 1988, 23, 54. Cartes Cartes, E.; Cartes Romero, E.; Martinez, R. Org. Mass Spectrum. 1990, 25, 127. Decouzon, M.; Gal, J.-F.; Geribaldi, S.; Rouillard, M.; Sturla, J. M. Org. Mass Spectrom. 1990, 25, 312. Waight, E. S. Org. Mass Spectrom. 1989, 24, 565. Yanagisawa, T.; Tobita, S.; Tajima, S.; Yoshifuji, M.; Inamoto, N. Shitsuryo Bunseki 1989, 37, 229. Longevialle, P.; Bouchoux, G.; Hoppillard, Y. Org. Muss Spectrom. 1989, 24, 919. Erra-Balsells, R. Org. Mass Spectrom. 1989, 24, 956. Geoffrey, F.; Cloke, N.; Day, J. P.; Greenway, A. M.; Sheddon, K. P.; Shimran, A. A.; Swain, A. C. 1. Organomet. Chem. 1989, 372, 231. Yaylayan, V.; Pare, J. R. J.; Laing, R.; Sporns, I’. Org. Mass Spectrom. 1990, 2.5, 141. Kuck, D.; Thoelmann, D.; Gruetzmacher, H. F. J. Chem. Sot., Perkin Trans. 2 1990, 251. Diakiw, V.; Gordon, I’. M.; Kingston, E. E.; Shannon, J. S.; Lacey, M. J. Org. Mass Spectrom. 1989, 24, 615. Mandelbaum, A.; Mueller, D. R.; Richter, W. J.; Vidavsky, I.; W&z, A. Org. Mass Spectrom. 1989, 24, 857. Meot-Ner, M.; Sieck, L. W. ht. J. Mass Spectrom. ion Pro- cesses 1989, 92, 123.

J Am Sot Mass Spectmm 1993, 4, 230-241 SITE-SELECTIVE GASPHASE REACTIONS 241

34. Attina, M.; Cacace, F.; De P&is, G.; Grandinetti, F. Inf. J. Mass Spectrom. Ion Processes 1989, 92, 263.

35. Charles, B.; Cole, R. B.; Bruneton, J.; Tab&, J. C. Org. Mass Spectrom. 1989, 24, 445.

36. Gillespie, T. A.; Urogdi, L.; Katritzky, A. R.; Yost, R. A. Org. Muss Spectrom. 1989, 24, 817.

37. Struaro, A.; Traldi, I’.; l’aradisi, C. Org. Mass Spectrom. 1989, 24, 777.

38. Mason, D. J.; Die&, A.; Smith, R. M. Antibiot. Cbemafher. 1961, 11, 118.

39. Cochran, T. G.; Abraham, D. J.; Mar& L. L. 1. Chem. Sot. Chem. Cummun. 1972, 494.

40. Umezawa, H. Ann. N. Y. Acad. Sci. 1958, 76, 20. 41. Hichens, M.; Rinehart, K. L., Jr. J. Am. Chem. Sot. 1963, 85,

1547.

42. Brodbelt, J.; Lieu, J.; Donovan, T. Anal. Chem. 1991, 63, 1205. 43. Eichmann, E. S.; Alvarez, E.; Brodbelt, J. S. J. Am. Sot. Muss

Spectrom. 1992, 3, 535. 44. Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrome-

fry/Mass Spectrum&y: Techniques and Applications of Tandem Mass Spectrometry, VCH: New York, 1988.

45. Weber-Grabau, M.; Kelley, P. E.; Syka, J. E. P.; Brad&w, S. C.; Brodbelt, J. S. Proceedings of the 35th ASMS Conference on Mass Spectrometty and Allied Topics; 1987; p 1114.

46. Liz., S.; Liebman, J.; Levin, K. J. Phys. Chem. Ref. Dafa 1984, 13.

47. March, J. Ed. Advanced Organic Chemistry: Reactions, Me&- nisms, and Structure. 3rd Ed., Wiley: New York, 1985; pp 305-308.

48. W&&am, R. J. J. Org. Ctinz. 1978, 43, 2.581.

an investigation of site-selective gas-phase reactions of amino … · 2017. 1. 29. · an...

Documents