Fragmentation of Substituted Oxonium Ions: the Role of Ion–Neutral Complexes

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Fragmentation of Substituted Oxonium Ions: The Role of Ion–Neutral Complexes

Ya-Ping Tu and John L. Holmes Department of Chemistry, University of Ottawa, Ottawa, Ontario, Canada

ϩ ϭ Disubstituted trialkyloxonium ions R1OCH2O (R1)CH2OR2 (R1,R2 CH3 or C2H5) have been prepared by chemical ionization of dimethoxy- and diethoxymethane individually or as a mixture, and their fragmentation has been studied by means of metastable ion and collision- induced dissociations. It is found that when a methoxymethyl group is attached to the charged oxygen atom, the oxonium ions can fragment by C–O bond cleavage to generate a methoxymethyl cation/dialkoxymethane ion–neutral complex, in which methyl cation transfer occurs to expel neutral formaldehyde. When an ethoxymethyl group is connected with the central oxygen atom, a reaction channel involving loss of C2H4O is observed and found to be insensitive to collisions. This process is proposed to involve isomerization prior to fragmentation leading to methylated dialkoxymethanes coordinated with neutral acetaldehyde in Ϫ ion–neutral complexes; these ion–neutral complexes are estimated to be 35 kJ mol 1 more stable than the original oxonium ions. (J Am Soc Mass Spectrom 1999, 10, 386–392) © 1999 American Society for Mass Spectrometry

ecomposition of organic ions in the gas phase where contributions from permanent and induced di- has been shown over the past 20 years to poles and orbital angular momentum are successively Dinvolve intermediates having nonconventional given. A simple approximation [15] which only takes structures such as distonic ions and ion–neutral com- the first term into consideration has successfully been plexes [1–9]. For odd-electron ions, fragmentation utilized to estimate semiquantitatively the stabilization ␮ through both distonic ions [3] and ion–neutral com- energy by using the permanent dipole moment, D,of plexes [4, 7] has been well documented; in some cases the neutral and the charge, q, on the ion when they are ␪ pathways undergoing these two unusual intermediates separated by a distance r (where is the angle between compete with each other [10], and sometimes difficul- the dipole and the charge to dipole axis). Indeed, the ties in determining the reaction mechanism require neutral partner plays a critical role in both ion–neutral complexes and complex-mediated reactions. Among increasingly sophisticated investigations [11, 12]. For the various small molecules acting as the neutral com- even-electron ions [13], however, decomposition in- ponent in the complexes, carbonyl compounds have volving distonic ions is unlikely because the energy relative large dipole moments, which result in pro- required to produce the appropriate diradical species is nounced stabilizing effects. For example, formaldehyde too great, making an ion–neutral complex the only ␮ with a D of 2.3 D can stabilize its ionic partner in a Ϫ accessible low energy intermediate through which frag- complex by 70 kJ mol 1; its homolog, acetaldehyde, mentation takes place [5, 9]. ␮ having a D of 2.75 D, provides a stabilization energy of Ϫ The existence of a stable ion–neutral complex on the up to 90 kJ mol 1 [5, 9, 15]. Such features of the neutral potential energy profile of a fragmentation reaction is species in the complexes have attracted special attention attributed to the interactions between the two partners [10, 16]. within the complex, which include ion–permanent di- Oxonium ions are not only important species in pole and ion–induced dipole attractions. Theoretical solution chemistry [17], but it is apparent that they also treatment indicates that an ion coordinating with a play a significant role in the chemistry of ion–neutral neutral species is stabilized by an energy given by the complexes in the gas phase. Most oxonium ions studied following equation [14]: by mass spectrometry are equivalent to alkylated carbonyl compounds [18], alkylated/protonated ethers ϭ Ϫ␮ ␪ 2 Ϫ ␣ 2 4 ϩ 2 ␮ 2 [19, 20], or protonated alcohols [21], where the ion Es dq cos /r q /2r L /q r chemistry is centered at the oxygen. Unfortunately, there are no reports to date on the fragmentation of oxonium ions containing an additional functionality on Address reprint requests to Dr. J. L. Holmes, Department of Chemistry, University of Ottawa, 10 Marie-Curie, Ottawa, Ontario K1N 6N5, Canada. the alkyl group. Generally, oxonium ions are high- E-mail: [email protected] energy species. If an appropriate functional group,

© 1999 American Society for Mass Spectrometry. Published by Elsevier Science Inc. Received October 27, 1998 1044-0305/99/$20.00 Revised December 28, 1998 PII S1044-0305(99)00004-5 Accepted December 28, 1998 J Am Soc Mass Spectrom 1999, 10, 386–392 OXONIUM ION FRAGMENTATIONS 387

Table 1. Product ions generated in metastable fragmentation at different accelerating voltages and in collision-induced dissociation of the oxonium ions Precursor m/z (% ⌺) ϩ ϩ (m/z) Mode CH3OCH2 C2H5OCH2 –C2H5OCH3 –CH3OCH3 –CH3CHO –CH2O 1 MI 4 kV 45 (2.4) 75 (91.5) 91 (6.1) (121) MI 8 kV 45 (3.6) 75 (91.0) 91 (5.4) CID 45 (29.7) 75 (68.5) 91 (1.8) 2 MI4kV 45(Ϫ) 59 (5.5) 75 (12.0) 89 (44.0) 91 (37.0) 105 (1.5) (135) MI8kV 45(Ϫ) 59 (13.4) 75 (13.5) 89 (40.5) 91 (31.5) 105 (1.1) CID 45 (4.8) 59 (47.8) 75 (14.0) 89 (19.4) 91 (13.7) 105 (0.3) 3 MI4kV 45(Ϫ)59(Ͻ0.1) 89 (14.4) 103 (5.3) 105 (80.7) 119 (0.5) (149) MI8kV 45(Ϫ) 59 (2.0) 89 (24.0) 103 (5.0) 105 (68.6) 119 (0.4) CID 45 (2.0) 59 (27.5) 89 (27.6) 103 (3.6) 105 (39.0) 119 (0.3) 4 MI 4 kV 59 (0.3) 103 (25.0) 119 (74.7) (163) MI 8 kV 59 (0.5) 103 (32.9) 119 (66.6) CID 59 (15.0) 103 (37.5) 119 (47.5)

which may be expected to produce a polar neutral To maximize the yield of the oxonium ions the repeller species through isomerization, is introduced onto the voltage was set at almost zero. alkyl moiety of an oxonium ion, is it possible to obtain MI and CID spectra were acquired with the ZABCAT an isomer in which the charge on the central oxygen program [24] by averaging 6–8 consecutive scans, at an atom is stabilized by the neutral species in an ion– accelerating voltage of 8 kV unless indicated otherwise. neutral complex? Helium was used as the collision gas for all CID Ϫ It was reported [22] that in the chemical ionization experiments at a cell pressure of 5 ϫ 10 8 mbar, which (CI) mass spectrometry of dimethoxymethane the reac- causes 10%–15% attenuation of the main beam for these tion of the protonated molecule with a neutral molecule ions. To obtain the CID mass spectra of the primary leads to the formation of a dimethoxylated trimethyl- products, precursor oxonium ions were mass selected oxonium ion. In the present study, dimethoxy- and by the magnetic sector; primary fragment ions gener- diethoxymethane were employed individually or as a ated metastably in the second field-free region (FFR) mixture to prepare substituted oxonium ions 1–4 under were then transmitted to the third FFR, where the above CI conditions by the same reaction as outlined below. CID conditions were applied. Metastable peak widths The ion chemistry was investigated by means of meta- (w0.5) were determined using sufficient energy resolu- stable ion (MI) and collision-induced dissociation (CID) tion to reduce the main beam width to 3–4 V at half techniques. Reaction mechanisms involving isomeriza- height. The corresponding kinetic energy release values tion to ion–neutral complexes prior to fragmentation were calculated by established procedures [25]. are proposed following identification of the product Dimethoxy- and diethoxymethane were commer- ions and estimation of heats of formation for related cially available (Aldrich Chemical, Milwaukee, WI) and species. used without further purification.

Results and Discussion The MI and CID mass spectra of oxonium ions 1–4 are summarized in Table 1. For those containing only methoxy or ethoxy substituents (1 and 4), fragmentation proceeds via three channels, i.e., simple bond cleavage, ether elimination, and loss of a neutral species corre- Diagram 1 sponding nominally to formaldehyde or acetaldehyde. For the other oxonium ions, 2 and 3, in which both methoxy and ethoxy are present, the decomposition Experimental pattern is a superposition of 1 and 4. ϩ All experiments were carried out on a modified [23] The product ions, CH3OCH2 (m/z 45) and ϩ ZAB 3F tandem mass spectrometer with BEE geometry C2H5OCH2 (m/z 59), from the decomposition of 1 and (VG Analytical, Manchester, U.K.). In the CI mode, 4, respectively, are of low abundance in the MI spectra dimethoxy-, diethoxymethane or their mixture (ϳ1:1 but become more significant upon collisional activation; ratio) was introduced through the septum inlet into the their intensities are also slightly increased at a shorter ion source which was maintained at a temperature of metastable time frame (with a higher accelerating volt- Ϫ Ϫ 150 °C and a total pressure of 8 ϫ 10 5–1 ϫ 10 4 mbar. age). These are characteristic of a simple bond cleavage 388 TU AND HOLMES J Am Soc Mass Spectrom 1999, 10, 386–392 reaction, such as Reaction 1, which has a relatively high activation energy and a small kinetic energy release.

ϩ [CH3OCH2–O (CH3)–CH2OCH3] 3 ϩ ϩ CH2(OCH3)2 CH3OCH2 (1)

Loss of 30 u from 1–3 (m/z 91, 105 and 119, respectively) is of minor importance in both MI and CID mass spectra. This is in sharp contrast to the loss of 44 u from 2–4 (m/z 91, 105, and 119, respectively), which predominates in both MI and CID mass spectra (Table 1). Furthermore, a more significant difference was found between the simple bond cleavages and these two processes. The former can be enhanced by collisions; the latter, however, are insensitive to collisions. The dependence of the intensity of these product ions on the collision gas pressure is illustrated in Figure 1. Methylated carbonyl compounds, including the ϩ CH3OCH2 cation which can be viewed as methylated formaldehyde, have been demonstrated to be excellent Figure 1. Dependence of the intensity on the pressure of He in methyl cation donors in the gas phase [26]. For the the collision cell for (filled circle) a simple bond cleavage product simple bond cleavage of 1, for example, the incipient ion and (open circle) loss of formaldehyde or acetaldehyde from ϩ the oxonium ions. fragment ion CH3OCH2 could transfer a methyl cation to its neutral counterpart, CH2(OCH3)2, before depar- ture to form methylated dimethoxymethane (m/z 91) Reaction 1 and this is supported by the similar low because the acetal has a greater methyl cation affinity abundances of the two product ions (Table 1). There- than formaldehyde. This reaction is confirmed by the fore, it can be proposed that loss of formaldehyde from determination of the structure of the product ion, which 1, as shown in Reaction 2, involves an ion–neutral shows a CID mass spectrum (Table 2) identical to that of complex, 5, in which methyl cation transfer occurs to methylated dimethoxymethane prepared by using methyl iodide as a reactant. (Dimethoxymethane gives produce complex 6. With formaldehyde as the neutral rise to an ion of m/z 91 in the ion source, which could partner, complex 6 would be a stable intermediate. This be produced either by the reaction described herein or is in agreement with the observed small kinetic energy by reaction between dimethoxymethane and the release (Table 2) compatible with a reaction mediated ϩ by an ion–neutral complex [27]. The formation of the CH3OCH2 ions. To obtain methylated dimethoxymeth- 2 3 ane by direct reaction with methyl iodide, CD3I was m/z 105 and 119 ions from and , respectively, should used, and the CID mass spectrum of the m/z 94 ion was follow similar pathways as indicated by the small observed.) Thus the formation of the m/z 91 ion from 1 kinetic energy release in each case (Table 2) although ϩ should have an activation energy approaching that for the CH3OCH2 ion disappears in the MI spectra.

ϩ 3 ϩ [CH3OCH2–O (CH3)–CH2OCH3] [CH2(OCH3)2 /CH3OCH2 ] (2) 15 3 ϩ A [CH3OCH2–O (CH3)2 /O CH2] 6 3 ϩ ϩ [CH3OCH2–O (CH3)2] CH2O

The absence of ethylated dialkoxymethanes among stable isomer after overcoming a transition state that is the decomposition products from oxonium ions 2–4, energetically higher than or at least as high as the ϩ e.g., the m/z 133 ion from 4, means that C2H5OCH2 is dissociation threshold of this newly formed isomer. The not necessarily a good ethyl cation donor in these product ions of this reaction were shown to be methyl- systems because other reaction channels become ener- ated dialkoxymethanes; for example, the CID mass getically more feasible. However, it is surprising to find spectrum of the m/z 119 ion from 4 is exactly the same that the loss of neutral C2H4Oisinsensitive to collisions, as that of authentic methylated diethoxymethane pre- which implies that the decomposing ion rearranges to a pared by the reaction of methyl iodide and di- J Am Soc Mass Spectrom 1999, 10, 386–392 OXONIUM ION FRAGMENTATIONS 389

Table 2. Kinetic energy releases (T0.5) for the major MI fragmentations and the CID mass spectra of their product ions

Precursor Primary product T0.5 CID of the primary product (m/z) Neutral lost (m/z) (meV) [m/z (% ⌺)]

1 CH3OCH3 75 35 47 (80) 45 (15) 44 (5) (121) CH2O 91 21 45 (80) 61 (20) 2 CH3OC2H5 75 26 47 (78) 45 (17) 44 (5) (135) CH3OCH3 89 31 45 (72) 61 (28) CH3CHO 91 19 45 (87) 61 (13) CH2O 105 20 45 (60) 59 (40) 3 CH3OC2H5 89 30 45 (62) 61 (38) (149) CH3OCH3 103 38 47 (7) 59 (82) 75 (11) CH3CHO 105 22 45 (56) 59 (42) 75 (16) CH2O 119 20 59 (100) 4 CH3OC2H5 103 34 47 (8) 59 (76) 75 (16) (163) CH3CHO 119 22 59 (99) 89 (1) ethoxymethane in the ion source, which is dominated reaction. Based on these results and the estimated heats by an ion of m/z 59 (Table 2). At first glance it could be of formation of related ions (see Appendix [31]), a proposed that the reaction starts with an isomerization potential energy profile for the decomposition of 4 is ϩ ⌬ of the initially formed C2H5OCH2 ion to presented in Figure 2. The fH of the oxonium ion 4 is ϩ Ϫ1 CH3CH OCH3 followed by methyl cation transfer; this estimated to be 50 kJ mol and that of methylated isomerization has been shown experimentally [28] and diethoxymethane (the product ion from acetaldehyde Ϫ by theoretical calculations [29, 30] to have a high energy loss) is 270 kJ mol 1. Using an energy window of 50 kJ barrier. If this were involved in this reaction, then the for the observation of the competing metastable disso- ϩ isomeric ion [C2H5OCH2–O (C2H5)–CH(CH3)OCH3] ciations, the activation energy of the isomerization from should also be able to give rise to the same m/z 119 ion. the precursor ion 4 to intermediate 7 is estimated to Ϫ Unfortunately, this ion cannot simply be prepared from be ϳ 80 kJ mol 1 with a transition state being ϳ25 kJ Ϫ diethoxymethane and dimethoxyethane under CI con- mol 1 higher in energy than the final products (Figure ditions; instead, an isomer of the oxonium ion 2, ϩ ؅ [CH3OCH2–O (CH3)–CH(CH3)OCH3](2 ), was ob- tained from the mixture of dimethoxymethane and dimethoxyethane under chemical ionization conditions. However, loss of acetaldehyde from 2؅ does not occur under either MI or CID conditions. This implies that ϩ ϩ isomerization of C2H5OCH2 to CH3CH OCH3 is not involved in the loss of acetaldehyde from oxonium ions 2–4. Alternatively, the reaction could proceed with a 1,3-hydrogen migration in the ethoxymethyl moiety, which results in an ion–neutral complex with acetaldehyde as the neutral partner, as described in Scheme 1. Although structural information on the neutral lost is not directly available, loss of the neutral as acetaldehyde is energetically more favorable than that of other

C2H4O isomers; for example, formation of oxirane Ϫ would be 113 kJ mol 1 higher in energy than that for acetaldehyde. More importantly, as indicated above, acetaldehyde is capable of providing a substantial stabilization energy in the ion–neutral complexes [5, 9, 15], and so the reaction intermediate 7 would lie in a well on the potential energy surface. This is well supported by the observed lack of collision sensitivity (Figure 1) and the small kinetic energy release value (Table 2) of this

Figure 2. Potential energy surface for the fragmentation of Scheme 1 oxonium ion 4. 390 TU AND HOLMES J Am Soc Mass Spectrom 1999, 10, 386–392

Scheme 3

Scheme 2 of 4, the CID mass spectrum of the product ion from ether elimination (m/z 103) shows a dominant peak at 2). The reported stabilization energy provided by acet- Ϫ m/z 59 (Table 2), indicating overwhelming loss of aldehyde in an ion–neutral complex (90 kJ mol 1) [15] ⌬ ϭ Ϫ1 acetaldehyde, whereas the diethoxymethyl cation (m/z leads to a fH 15 kJ mol for the intermediate Ϫ 103, from triethyl orthoformate by loss of an ethoxy complex 7, which is 35 kJ mol 1 more stable than its group) decomposes by expelling a C H unit to form the precursor ion. The formation of such a stable complex 2 4 base peak at m/z 75 in the CID mass spectrum. This can would be the driving force for the isomerization in spite be rationalized by assuming that the ether elimination of the energy barrier to a 1,3-hydrogen migration. The from 4 involves a 1,5-hydrogen migration as shown in reaction mechanism outlined in Scheme 1 should be Scheme 3. also valid for loss of acetaldehyde from 2 and 3, the The above proposal that ether elimination via 1,5- reactions having similar features, although another hydrogen migration is preferred over the stepwise pathway could be operative for ion 2, as discussed hydrogen abstraction is further supported in the frag- below. mentation of 2. Stepwise loss of dimethyl ether from 2 Ether elimination is another major fragmentation of ϩ would produce a (CH O)CH (OC H ) ion (m/z 89), all oxonium ions. Dimethyl or diethyl ether is expelled 3 2 5 which is expected to lose ethylene to form an ion of m/z from 1 and 4, respectively, while losses of these two 61. However, the CID mass spectrum of the m/z 89 ion ethers from 2 and 3 compete with each other. The CID ϩ ϩ is predominated by acetaldehyde loss (45 :61 ϭ 72:28, mass spectrum of the product ion, m/z 75, of the Table 2), implying a structure such as CH3OCH2– reaction of 1 (Table 2) is similar to that of the dime- ϩ ϩ OCH CH which can be readily generated through a thoxymethyl cation, CH(OCH ) , generated by loss of 3 3 2 1,5-hydrogen migration similar to that depicted in a methoxy group from trimethyl orthoformate, showing Scheme 3.Asfor3, therefore, the higher abundance of the dominant ion at m/z 47 corresponding to proton- the m/z 89 ion compared with that of the m/z 103 ion ated dimethyl ether. However, the signal at m/z 45 in in both MI and CID mass spectra can thus be rational- the CID mass spectrum of the m/z 75 ion from 1, which ized by isomerization of the precursor ion to 3؅ (Scheme is absent in that of the dimethoxymethyl cation, indi- 4) through which elimination of methyl ethyl ether by cates the cogeneration of an isomeric m/z 75 ion, ϩ 1,5-H migration takes place. Wherever an ethoxymethyl presumably CH3OCH2–OCH2 , which can readily expel ϩ group is present, ether elimination via 1,5-H migration formaldehyde to give rise to the CH OCH ion (m/z 3 2 predominates, and is shown by the loss of acetaldehyde 45). These two isomeric m/z 75 ions could be formed from the product ions of this reaction (all m/z 89 and by a simple C–O bond cleavage followed by hydrogen 103 ions from this reaction of 2–4, Table 2). abstraction from different sites as shown in Scheme 2 or It is seen in Table 1 that not only are the intensities of by concerted hydrogen migration. Stepwise hydrogen m/z 89 and 91 ions for 2 close to each other, but they transfer is a nonregiospecific process, pathway (ii) in also change in the same manner when the accelerating Scheme 2 is therefore of lesser importance for 1 because voltage changes. This can be understood by assuming of the instability of the resultant primary carbocation. that both fragment ions are formed via a common Even for ethoxy-containing oxonium ions (2–4), where transition state. The mechanism for ether elimination the methylene hydrogen of the ethoxy group can mi- through a 1,5-hydrogen shift is appropriate for these grate to produce more-stable secondary carbocations, processes. As shown in Scheme 5, the dimethyl ether pathway (ii) should still not be competitive with (i) as initially formed is associated with its ionic partner as an the dialkoxymethyl cation involved in pathway (i) is ion–neutral complex 8, which isomerizes to complex 9 expected to be more stable. However, the ethoxy hy- prior to extruding acetaldehyde. By sharing the 6-mem- drogen-involving channel is prevalent when this group bered ring as the transition state, both the elimination of is present in the oxonium ion. For example, in the case dimethyl ether and the loss of acetaldehyde have the

Scheme 4 Scheme 5 J Am Soc Mass Spectrom 1999, 10, 386–392 OXONIUM ION FRAGMENTATIONS 391

Ϫ1 same activation energy. Unfortunately, this mechanism kJ mol ; from these the MCA of CH2(OCH3)2 and Ϫ1 is not applicable to the decomposition of the other CH2(OC2H5)2 are estimated to be 387 and 410 kJ mol ⌬ ϩ oxonium ions (1, 3, and 4), where the significant differ- and the fH of [CH2(OCH3)2 CH3 ] and [CH2(OC2H5)2 ϩ Ϫ1 ⌬ ence in abundance of the product ions from these two CH3 ] are 360 and 270 kJ mol . Comparing the fH of ϩ processes indicates that the two reaction channels likely the these two ions with that of (CH3)3O and ϩ have different activation energies. (CH3)2O C2H5, substitution of a methoxy or ethoxy group on the carbon adjacent to the central charged oxygen atom results in a change of Ϫ175 and Ϫ220 kJ Conclusion Ϫ mol 1, respectively, in the heat of formation. Hence for ⌬ ϭ ϩ Ϫ ϫ ϭ Fragmentation of substituted trialkyloxonium ions in oxonium ion 1, fH 536 ( 175) 2 186 and ⌬ ϭ ϩ Ϫ ϫ ϭ Ϫ1 the gas phase proceeds via three channels, i.e., simple for 4, fH 490 ( 220) 2 50 kJ mol . bond cleavage, ether elimination, and expulsion of formaldehyde and/or acetaldehyde. It is found that the aldehyde losses involve a prior isomerization leading to Acknowledgment methylated dialkoxymethanes coordinated with the The authors are indebted to the Natural Sciences and Engineering neutral aldehyde molecule in ion–neutral complexes. It Research Council (Canada) for continuing financial support. is shown that both aldehydes provide substantial stabilization energies to their ionic partner, and thus the References formaldehyde-involved complex is estimated to be as stable as the original ion and the acetaldehyde-coordi- 1. Williams, D. H. Acc. Chem. Res. 1977, 10, 280–286. Ϫ nated complexes are ϳ 35 kJ mol 1 more stable than the 2. Morton, T. H. Tetrahedron 1982, 38, 3195–3243. precursor oxonium ions. 3. Hammerum, S. Mass Spectrom. Rev. 1988, 7, 123–202. 4. McAdoo, D. J. Mass Spectrom. Rev. 1988, 7, 363–393. 5. Bowen, R. D. Acc. Chem. Res. 1991, 24, 364–371. Appendix 6. Morton, T. H. Org. Mass Spectrom. 1992, 27, 353–368. 7. Longevialle, P. Mass Spectrom. Rev. 1992, 11, 157–192. ⌬ To obtain the fH for the oxonium ions 1 and 4, the 8. McAdoo, D. J.; Morton, T. H. Acc. Chem. Res. 1993, 26, 295–302. ⌬ 9. Bowen, R. D. Org. Mass Spectrom. 1993, 28, 1577–1595. fH, PA, and methyl cation affinity (MCA) values are needed for CH (OCH ) and CH (OC H ) . The PA of 10. Traeger, J. C.; Hudson, C. E.; McAdoo, D. J. J. Am. Soc. Mass 2 3 2 2 2 5 2 1992, CH (OCH ) can be estimated from the difference in Spectrom. 3, 409–416. 2 3 2 11. (a) Holmes, J. L.; Lossing, F. P.; Terlouw, J. K.; Burgers, P. C. proton affinities between CH3O(CH2)nH and J. Am. Chem. Soc. 1982, 104, 2931–2932. (b) Burgers, P. C.; CH3O(CH2)nOCH3, which reflects the stabilizing effect Holmes, J. L.; Terlouw, J. K.; Van Baar, B. Org. Mass Spectrom. of the second methoxy group by forming a cyclic 1985, 20, 202–206. (c) Postma, R.; Ruttink, P. J. A.; Van Baar, B.; hydrogen bond in protonated dimethoxyalkanes. This Terlouw, J. K.; Holmes, J. L.; Burgers, P. C. Chem. Phys. Lett. Ϫ difference is 49, 40, and 27 kJ mol 1 for n ϭ 2–4, 1986, 123, 409–412. respectively. When n ϭ 1, the stabilizing effect of the 12. (a) Biermann, H. W.; Morton, T. H. J. Am. Chem. Soc. 1983, 105, 5025–5030. (b) Burgers, P. C.; Holmes, J. L.; Hop, C. E. C. A.; extra methoxy group should drop significantly, there- Postma, R.; Ruttink, P. J. A.; Terlouw, J. K. J. Am. Chem. Soc. fore the PA difference between dimethyl ether and 1987, 109, 7315–7321. (c) Cao, J. R.; George, M.; Holmes, J. L.; ϳ Ϫ1 CH2(OCH3)2 is estimated to be 25 kJ mol , leading Sirois, M.; Terlouw, J. K.; Burger, P. C. J. Am. Chem. Soc. 1992, ϭ Ϫ1 toaPA 815 kJ mol for CH2(OCH3)2. For CH3OCH3, 114, 2017–2020. (d) Audier, H. E.; Milliet, A.; Leblanc, D.; 1992, CH3OC2H5, and C2H5OC2H5, PA is 790, 808, and 828, Morton, T. H. J. Am. Chem. Soc. 114, 2020–2027. ⌬ Ϫ Ϫ Ϫ Ϫ1 13. McLafferty, F. W. Org. Mass Spectrom. 1980, 15, 114–121. and fH is 184, 217, and 251 kJ mol , respectively; each ethyl substitution for the methyl group 14. Su, T.; Bowers, M. T. In Gas Phase Ion Chemistry, Bowers, M. T., ⌬ ϭ Ϫ1 ⌬⌬ ϭ Ed.; Academic: New York, 1979; Chap 2. brings about a PA 19 kJ mol and a fH 15. Bowen, R. D.; Williams, D. H. Int. J. Mass Spectrom. Ion Phys. Ϫ Ϫ1 ϭ ϩ 33.5 kJ mol . Hence, for CH2(OC2H5)2,PA 815 1979, 29, 47–55. ϫ ϭ Ϫ1 ⌬ ϭϪ ϩ Ϫ ϫ 19 2 853 kJ mol and fH 347 ( 33.5) 16. (a) Longevialle, P. Rapid Commun. Mass Spectrom. 1995, 9, Ϫ 2 ϭϪ414 kJ mol 1 are derived from the corresponding 1189–1194. (b) Longevialle, P. Rapid Commun. Mass Spectrom. 1996, 10, 621–626. (c) Longevialle, P.; Lefevre, O.; Mollova, W.; values for CH2(OCH3)2. The uncertainty of the estima- Ϫ tion is Ϯ 10 kJ mol 1. Bouchoux, G. Rapid Commun. Mass Spectrom. 1998, 12, 57–60. ⌬ ϩ ϩ 17. (a) Perst, H. Oxonium Ions in Organic Chemistry, Verlag Che- The fH of (CH3)3O and (C2H5)3O have been Ϫ1 mie: Weinheim, 1971. (b) Perst, H. In Carbonium Ions, Olah, reported [19c, d] to be 536 and 406 kJ mol , respec- G. A., Schleyer, P. V. R., Eds.; Wiley–Interscience: New York, tively; assuming a linear change in heat of formation, 1976, Vol. 5. ⌬ ϩ Ϫ1 fH of (CH3)2O C2H5 would be 490 kJ mol and 18. (a) Bowen, R. D.; Colburn, A. W.; Derrick, P. J. Org. Mass accordingly the MCA of (CH3)2O and CH3OC2H5 are Spectrom. 1992, 27, 625–632. (b) Bowen, R. D.; Derrick, P. J. Ϫ estimated to be 375 and 388 kJ mol 1, respectively. The J. Chem. Soc., Perkin Trans. 2 1992, 1033–1039. (c) Bowen, R. D.; differences in PA between (CH ) O and CH (OCH ) Wright, A. D.; Derrick, P. J. J. Chem. Soc., Perkin Trans. 2 1993, 3 2 2 3 2 501–507. (d) Bowen, R. D.; Derrick, P. J. Org. Mass Spectrom. and between CH3OC2H5 and CH2(OC2H5)2 are 25 and Ϫ1 1993, 28, 1197–1209. (e) Bowen, R. D.; Suh, D.; Terlouw, J. K. 45 kJ mol , respectively; using simple proportions, the J. Chem. Soc., Perkin Trans. 2 1995, 119–129. difference in MCA between each pair of compounds 19. (a) Farcasiu, D.; Pancirov, R. G. Int. J. Mass Spectrom. Ion would be 25 ϫ 375 / 790 ϭ 12 and 45 ϫ 388 / 808 ϭ 22 Processes 1986, 74, 207–215. (b) Wang, D.; Squires, R. R.; 392 TU AND HOLMES J Am Soc Mass Spectrom 1999, 10, 386–392

Farcasiu, D. Int. J. Mass Spectrom. Ion Processes 1991, 107, 23. Holmes, J. L.; Mayer, P. M. J. Phys. Chem. 1995, 99, 1366–1370. R7–R8. (c) Zagorevskii, D. V.; Sirois, M.; Cao, J. R.; George, M.; 24. Traeger, J. C.; Mommers, A. A. Org. Mass Spectrom. 1987, 22, Holmes, J. L.; Ross, C. W., III J. Mass Spectrom. 1996, 31, 55–61. 592–596. (d) Nguyen, M. T.; Bouchoux, G. J. Phys. Chem. 1996, 100, 25. Holmes, J. L.; Terlouw, J. K. Org. Mass Spectrom. 1980, 15, 2089–2093. (e) Audier, H. E.; Koyanagi, G. K.; McMahon, T. B.; 383–396. Tholmann, D. J. Phys. Chem. 1996, 100, 8220–8223. 26. (a) Szulejko, J. E.; Fischer, J. J.; McMahon, T. B.; Wronka, J. Int. 20. (a) Terlouw, J. K.; Weiske, T.; Schwarz, H.; Holmes, J. L. Org. J. Mass Spectrom. Ion Processes 1988, 83, 147–161. (b) Audier, Mass Spectrom. 1986, 21, 665–671. (b) Audier, H. E.; Monteiro, H. E.; McMahon, T. B. J. Mass Spectrom. 1997, 32, 201–208. C.; Berthomieu, D.; Tortajada, J. Int. J. Mass Spectrom. Ion 27. Hammerum, S.; Hansen, M. M.; Audier, H. E. Int. J. Mass Processes 1991, 104, 145–161. (c) Kondrat, R. W.; Morton, T. H. Spectrom. Ion Processes 1997, 160, 183–192. Org. Mass Spectrom. 1991, 26, 410–415. (d) Kondrat, R. W.; 28. Hvistendahl, G.; Williams, D. H. J. Am. Chem. Soc. 1975, 97, Morton, T. H. J. Org. Chem. 1991, 56, 952–957. (e) Audier, H. E.; 3097–3101. Berthomieu, D.; Morton, T. H. J. Org. Chem. 1995, 60, 7198– 29. Hudson, C. E.; McAdoo, D. J. J. Am. Soc. Mass Spectrom. 1998, 7208. 21. (a) Schwarz, H.; Stahl, D. Int. J. Mass Spectrom. Ion Phys. 1980, 9, 130–137. 36, 285–289. (b) Dawson, P. H. Int. J. Mass Spectrom. Ion 30. Chalk, A. J.; Radom, L. J. Am. Chem. Soc. 1998, 120, Processes 1983, 50, 287–297. (c) Jarrold, M. F.; Illies, A. J.; 8430–8437. Kirchner, N. J.; Bowers, M. T. Org. Mass Spectrom. 1983, 18, 31. Heats of formation for related species are taken from Lias, 388–395. (d) Meot-Ner (Mautner), M.; Ross, M. M.; Campana, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; J. E. J. Am. Chem. Soc. 1985, 107, 4839–4845. (e) Harrison, A. G. Mallard, W. G. J. Phys. Chem. Ref. Data 1988, 17(Suppl. 1), Org. Mass Spectrom. 1987, 22, 637–641. 1–861. For those derived from PA, values are calibrated with 22. Tu, Y.-P.; Chen, Y.-Z.; Chen, S.-N.; Wang, M.-L.; Jing, Z.-Z. newly evaluated PA data taken from Hunter, E. P. L.; Lias, Org. Mass Spectrom. 1990, 25, 9–13. S. G. J. Phys. Chem. Ref. Data 1998, 27, 413–656.