An experimental and theoretical study of the long-lived radical cation of CH3OCH2CH2OH

Article abstract of DOI:10.1021/ja952145p

The long-lived radical cation of 2-methoxyethanol, CH₃OCH₂CH₂OH^.+, is stable toward isomerization in the gas phase. This radical cation reacts with neutral reagents via three main channels in a Fourier-transform ion cyclotron resonance mass spectrometer. CH₃OCH₂CH₂OH^.+ abstracts an electron from reagents with ionization energies less than that of CH₃OCH₂CH₂OH. Neutral reagents with higher ionization energies replace CH₃OCH₂^. or CH₂O in the ion. Both replacement reactions are probably driven by the formation of a hydrogen bond between the entering nucleophile and the hydroxyl group in CH₃OCH₂CH₂OH^.+, which breaks the C-C bond in CH₃OCH₂CH₂OH^.+. For strong nucleophiles, this process is so exothermic that the resulting ion/molecule complex dissociates by loss of CH₃OCH₂^.. However, weak nucleophiles yield a longer-lived ion-molecule complex. Within this complex, CH₃OCH₂^. can react with the initially produced ion by replacement of CH₂O (a weaker nucleophile) to yield the hydrogen-bridged complex of CH₃OCH₂^. and the nucleophile. This reaction provides a general approach for the gas-phase synthesis of different hydrogen-bridged radical cations. Ab initio molecular orbital calculations suggest that the most stable geometry of CH₃OCH₂CH₂OH^.+ is characterized by an unusually long C-C bond (1.775 ? at the UMP2/6-31G** level of theory). This finding is in agreement with the observed facile cleavage of the C-C bond in CH₃OCH₂CH₂OH^.+. The isomeric ions ^.CH₂O⁺(CH₃)CH₂OH, (CH₃)₂O⁺CH₂O^., ^.CH₂OCH₂CH₂OH₂⁺, ^.CH₂O(CH₃)-H⁺···O=CH₂, and CH₃OC(H)H···O(H)=CH₂^.+ were calculated to be less stable than CH₃OCH₂CH₂OH^.+ (at the UMP2/6-31G**//UHF/6-31G**+ZPE level of theory). Only the ion (CH₃)₂O-H···O=CH^.+ was found to lie lower in energy than CH₃OCH₂CH₂OH^.+ (by 7.2 kcal/mol at the UMP2/6-31G**+ZPVE level of theory). However, experimental evidence does not support the formation of this hydrogen-bridged ion upon ionization of 2-methoxyethanol.

Full text of DOI:10.1021/ja952145p

3914
J. Am. Chem. Soc. 1996, 118, 3914-3921
An Experimental and Theoretical Study of the Long-Lived
Radical Cation of CH3OCH2CH2OH
Jaana M. H. Pakarinen,† Rebecca L. Smith,‡ Pirjo Vainiotalo,*,†
Tapani A. Pakkanen,*,† and Hilkka I. Kentta1maa*,‡
Contribution from the Departments of Chemistry, UniVersity of Joensuu, FIN-80101 Joensuu,
Finland, and Purdue UniVersity, West Lafayette, Indiana 47907-1393
ReceiVed June 29, 1995X
Abstract: The long-lived radical cation of 2-methoxyethanol, CH3OCH2CH2OH•+, is stable toward isomerization in
the gas phase. This radical cation reacts with neutral reagents via three main channels in a Fourier-transform ion
cyclotron resonance mass spectrometer. CH3OCH2CH2OH•+ abstracts an electron from reagents with ionization
energies less than that of CH3OCH2CH2OH. Neutral reagents with higher ionization energies replace CH3OCH2• or
CH2O in the ion. Both replacement reactions are probably driven by the formation of a hydrogen bond between the
entering nucleophile and the hydroxyl group in CH3OCH2CH2OH•+, which breaks the C-C bond in CH3OCH2-
CH2OH•+. For strong nucleophiles, this process is so exothermic that the resulting ion/molecule complex dissociates
•
by loss of CH3OCH2 . However, weak nucleophiles yield a longer-lived ion-molecule complex. Within this complex,
•
CH3OCH2 can react with the initially produced ion by replacement of CH2O (a weaker nucleophile) to yield the
•
hydrogen-bridged complex of CH3OCH2 and the nucleophile. This reaction proVides a general approach for the
gas-phase synthesis of different hydrogen-bridged radical cations. Ab initio molecular orbital calculations suggest
that the most stable geometry of CH3OCH2CH2OH•+ is characterized by an unusually long C-C bond (1.775 Å at
the UMP2/6-31G** level of theory). This finding is in agreement with the observed facile cleavage of the C-C
•
•
+
bond in CH3OCH2CH2OH•+. The isomeric ions CH2O+(CH3)CH2OH, (CH3)2O+CH2O•, CH2OCH2CH2OH2
,
•CH2O(CH3)-H+‚‚‚OdCH2, and CH3OC(H)H‚‚‚O(H)dCH2•+ were calculated to be less stable than CH3OCH2CH2-
OH•+ (at the UMP2/6-31G**//UHF/6-31G**+ZPE level of theory). Only the ion (CH3)2O-H‚‚‚OdCH•+ was found
to lie lower in energy than CH3OCH2CH2OH•+ (by 7.2 kcal/mol at the UMP2/6-31G**+ZPVE level of theory).
However, experimental evidence does not support the formation of this hydrogen-bridged ion upon ionization of
2-methoxyethanol.
Introduction
Both types of ions have been established to represent stable
gas-phase species.1,4,6b,13
The removal of an electron from an organic molecule often
significantly affects its structure, relative thermodynamic stabil-
ity, and the barrier heights for isomerization and dissociation.1
Isomerization reactions are common for organic radical cations.
Indeed, the combined results obtained from mass spectrometry
experiments and high-level ab initio molecular orbital calcula-
tions have demonstrated that radical cations formed upon
removal of an electron from a stable neutral molecule can be
significantly less stable than their nonconventional isomers with
no stable neutral counterparts. Such nonconventional radical
cations include hydrogen-bridged radical cations and distonic
radical cations (i.e., ionized ylides, zwitterions, or biradicals).1-16
The 2-methoxyethanol radical cation CH3OCH2CH2OH•+ (1)
undergoes interesting unimolecular dissociation reactions,16-22
some of which have been proposed to involve hydrogen-bridged
(8) Postma, R.; van Helden, S. P.; van Lenthe, J. H.; Ruttink, P. J. A.;
Terlouw, J. K.; Holmes, J. L. Org. Mass Spectrom. 1988, 24, 503.
(9) van Driel, J. H.; Heerma, W.; Terlouw, J. K.; Halim, H.; Schwarz,
H. Org. Mass Spectrom. 1985, 20, 665.
(10) Cao, J. R.; George, M.; Holmes, J. L.; Sirois, M.; Terlouw, J. K.;
Burgers, P. C. J. Am. Chem. Soc. 1992, 114, 2017.
(11) Heinrich, N.; Schmidt, J.; Schwarz, H.; Apeloig, Y. J. Am. Chem.
Soc. 1987, 109, 1317.
(12) (a) Postma, R.; Ruttink, P. J. A.; van Baar, B.; Terlouw, J. K.;
Holmes, J. L.; Burgers, P. C. Chem. Phys. Lett. 1986, 123, 409. (b) Burgers,
P. C.; Lifshitz, C.; Ruttink, P. J. A.; Schaftenaar, G.; Terlouw, J. K. Org.
Mass Spectrom. 1989, 24, 579.
(13) For recent reviews, see: (a) Hammerum, S. Mass Spectrom. ReV.
1988, 7, 123. (b) Stirk, K. M.; Kiminkinen, M.; Kentta¨maa, H. I. Chem.
ReV. 1992, 92, 1649.
(14) Schaftenaar, G.; Postma, R.; Ruttink, P. J. A.; Burgers, P. C.;
McGibbon, G. A.; Terlouw, J. K. Int. J. Mass Spectrom. Ion Processes
1990, 100, 521.
(15) McGibbon, G. A.; Kingsmill, C. A.; Terlouw, J. K.; Burgers, P. C.
Org. Mass Spectrom. 1992, 27, 126.
(16) See, for example: Wesdemiotis, C.; Polce, M. J. J. Am. Soc. Mass
Spectrom. 1995, 6, 1030 and references therein.
† University of Joensuu.
‡ Purdue University.
X Abstract published in AdVance ACS Abstracts, April 15, 1996.
(1) Roth, H. D. In Topics in Current Chemistry; Springer-Verlag: Berlin,
1992; Vol. 163.
(2) (a) McAdoo, D. J. Mass Spectrom. ReV. 1988, 7, 363. (b) Longe-
vialle, P. Mass Spectrom. ReV. 1992, 11, 157.
(3) Heinrich, N.; Schwarz, H. Ion and Cluster Ion Spectroscopy and
Structure; Maier, J. P., Ed.; Elsevier: Amsterdam, 1989; pp 329-372.
(4) (a) Pakarinen, J. M. H.; Vainiotalo, P.; Pakkanen, T. A.; Kentta¨maa,
H. I. J. Am. Chem. Soc. 1993, 115, 12431. (b) George, M.; Kingsmill, A.
C.; Suh, D.; Terlouw, J. K.; Holmes, J. L. J. Am. Chem. Soc. 1994, 116,
7807.
(17) (a) Morton, T. H. Tetrahedron 1982, 38, 3195. (b) Biermann, H.
W.; Morton, T. H. J. Am. Chem. Soc. 1983, 105, 5025.
(18) Vazquez, S.; Mosquera, R. A.; Rios, M. A.; van Alsenoy, C. J.
Mol. Struct. (Theochem) 1989, 188, 95.
(5) Postma, R.; Ruttink, P. J. A.; van Duijneveldt, F. B.; Terlouw, J. K.;
Holmes, J. L. Can. J. Chem. 1985, 63, 2798.
(6) (a) Burgers, P. C.; Holmes, J. L.; Hop, C. E. C. A.; Postma, R.;
Ruttink, P. J. A.; Terlouw, J. K. J. Am. Chem. Soc. 1987, 109, 7315. (b)
Hoke, S. H., II; Yang, S. S.; Cooks, R. G.; Hrovat, D. A.; Borden, W. T.
J. Am. Chem. Soc. 1994, 116, 4888.
(19) Ruttink, P. J. A.; Burgers, P. C. Org. Mass Spectrom. 1993, 28,
1087.
(20) Audier, H. E.; Milliet, A.; Leblanc, D.; Morton, T. H. J. Am. Chem.
Soc. 1992, 114, 2020.
(7) van Baar, B. L. M.; Burgers, P. C.; Holmes, J. L.; Terlouw, J. K.
Org. Mass Spectrom. 1988, 23, 355.
(21) Cao, J. R.; George, J. L.; Holmes, J. L.; Sirois, M.; Terlouw, J. K.;
Burgers, P. C. J. Am. Chem. Soc. 1992, 114, 2017.
0002-7863/96/1518-3914$12.00/0 © 1996 American Chemical Society

Experimental and Theoretical Study of CH3OCH2CH2OH•+
J. Am. Chem. Soc., Vol. 118, No. 16, 1996 3915
Scheme 1
the long-lived radical cation generated upon ionization of CH3-
OCH2CH2OH is stable toward isomerization. Ab initio molec-
ular orbital calculations carried out at the UMP2/6-31G** level
of theory suggest that the most stable conformer of CH3OCH2-
CH2OH•+ is characterized by an unusually long C-C bond. This
bond is readily cleaved upon gas-phase reactions with neutral
reagents.
Experimental and Theoretical Methods
All experiments were carried out using an Extrel Fourier-transform
ion cyclotron resonance mass spectrometer (Model 2001 FT/MS system)
equipped with 3 T superconducting magnet (operated at 2.3 T),
described in detail previously.4a The cell was differentially pumped
with two turbomolecular pumps (Balzers TPU 330) to achieve a base
pressure of <1 × 10-9 Torr in each cell, as read by an ionization gauge.
Reagents were introduced by using either a pulsed valve inlet system
or one of two Extrel FT/MS heated batch inlet systems equipped with
variable leak valves.
and/or distonic radical cation intermediates. The strong in-
tramolecular hydrogen bond18 in the neutral CH3OCH2CH2OH
was initially thought to remain after ionization (1, Scheme 1)
and to lead to isomerization of the initially formed radical cation
to a hydrogen-bridged complex of formaldehyde and the
methoxymethyl radical, CH3O(CH2)‚‚‚H‚‚‚OdCH2•+ (2, Scheme
1).17 This ion was suggested to undergo a hydrogen atom shift
to yield another hydrogen-bridged intermediate, (CH3)2-
O‚‚‚H‚‚‚OdCH•+ (3), which decomposes by expulsion of CHO•.
A similar mechanism was proposed19 for the loss of CHO• from
the molecular ion of ethylene glycol, but this mechanism was
later proved to be incorrect through labeling experiments.20,21
“Chirp” excitation (2.7 MHz bandwidth, 124 Vp-p amplitude, 3.0-
3.1 kHz/µs sweep rate) was used to excite ions for detection. The
spectra were subjected to one zero fill prior to Fourier transformation,
and they were recorded as 32K data points at a digitizer rate of 5.3
MHz. The spectra shown are an average of g15 acquisitions.
The radical cation of CH3OCH2CH2OH•+ was generated from
methoxyethanol introduced into one side of the dual cell by electron
ionization or by electron abstraction using CS2•+ reagent ions (ionization
energies23a (IE) of CS2 and CH3OCH2CH2OH are 10.07 and 9.6 eV,
respectively). The electron energy (20-75 eV), emission current (2-
10 µA), and ionization time (5-50 ms) were optimized separately for
each experiment. Undesired ions generated by the electron beam in
the other side of the dual cell were removed by applying a negative
potential on the remote trapping plate of this cell. The ions remaining
in the other cell were then transferred into the cleaned cell by grounding
the conductance limit plate for 100 µs and collisionally cooled with
argon pulsed into the cell. The radical cation of CH3OCH2CH2OH
(m/z 76) was isolated through the use of the Stored Waveform Inverse
Fourier Transform (SWIFT)24 excitation method (Extrel FTMS SWIFT
Module). The molecular ion was then allowed to react with a selected
neutral reagent for a variable time period. Primary products were
identified based on their constant branching ratios at short reaction
times. The decay of the relative abundance of the reactant ion as a
function of time was used to deduce the reaction rate constants (k).
All the samples were obtained commercially and used as received.
The nominal sample pressures were 1.2 × 10-7 to 2.5 × 10-7 Torr.
The pressure readings were corrected for the sensitivity25a of the ion
gauges and for the instrument’s pressure gradient by using common
procedures.4a,23c,25b The precision of the reaction rate measurements is
better than 10%; the accuracy is estimated to be better than (50%.
For collision-activated dissociation (CAD) experiments, the ions were
cooled by allowing them to collide with a stationary pressure of argon
(nominal pressure 1.2 × 10-7 Torr) for a relatively long time period
(1-8 s). The isolated ions were accelerated to a preselected final kinetic
energy26 by using an on-resonance radio frequency pulse. Dissociation
products for different estimated laboratory kinetic energies (12-100
eV) were measured by varying the duration of the excitation pulse. A
time period of 100 ms was allowed for collisions of the accelerated
ions with argon.
The role of hydrogen-bridged intermediates in the dissociation
of CH3OCH2CH2OH•+ was reinvestigated20,21 by using differ-
ently deuterated analogues of CH3OCH2CH2OH. The results
convincingly demonstrate that the mechanism initially proposed
for the CHO• loss must be incorrect. Various findings, including
the observation of extensive intramolecular hydrogen scrambling
involving all the hydrogen atoms except the methylene hydro-
gens in CH2OH, were rationalized by the intermediacy of the
distonic ions •CH2O+(CH3)CH2OH (4) and (CH3)2O+CH2O• (5)
in the pathway leading to loss of CHO• (Scheme 2). The
mechanism proposed20 for this dissociation involves the cleavage
of the C-C bond, rotation of a methylene group, and C-O
bond formation to generate the distonic ion •CH2O+(CH3)CH2-
OH (4). Migration of the hydroxylic hydrogen to the terminal
methylene carbon yields the other distonic ion, (CH3)2O+CH2O•
(5), which dissociates to yield protonated dimethyl ether and
CHO•. Another dissociation pathway, loss of water, was
•
proposed to occur via formation of a third distonic ion, CH2-
OCH2CH2OH2+ (6, Scheme 2).20 Recent results16 suggest that
•
+
this reaction yields CH2CH2OCH2 and a minor amount of
CH2dCHOCH3•+ and not the oxetane radical cation proposed20
earlier.
The above investigations convincingly ruled out hydrogen-
bridged complexes as reactive intermediates during dissociation
of CH3OCH2CH2OH•+, while providing support for the equili-
bration of three different distonic ions with the internally excited
radical cation CH3OCH2CH2OH•+. These intriguing findings
raise a question concerning the structure of the nonfragmenting,
long-liVed CH3OCH2CH2OH•+. The present study shows that
Standard ab initio molecular orbital calculations were performed
using the GAUSSIAN9227 series of programs. Optimized geometries
Scheme 2

3916 J. Am. Chem. Soc., Vol. 118, No. 16, 1996
Pakarinen et al.
Table 1. Reaction Rate Constants, Reaction Efficiencies, and Product Ions Measured for Reactions of Ionized 2-Methoxyethanol (m/z 76, IE
) 9.6 eV, PA ) 182.6 kcal/mol23a) with Various Neutral Reagents
primary
product ions
secondary
d
neutral reagent MW PAa (kcal/mol) IEb (eV)
kc
kcollision
effe
0.4
m/z
%
reaction
product ionsf (m/z)
•
•
CH3CHO
44
186.6
10.2
1.1
2.6
75
90
70
30
replacement of CH3OCH2
replacement of CH2O
89
CH3CH2CHO
(CD3)2CO
cyclopentanone
CD3OD
CH3CH2OH
58
64
84
36
46
189.6
196.7g
198.8
181.9g
188.3
10.0
9.7g
9.3
1.5
1.9
2.4
0.6
1.4
2.3
2.3
2.5
1.9
1.8
0.7
0.8
1.0
0.3
0.8
89
100
100
100
70
replacement of CH3OCH2•
replacement of CH3OCH2•
replacement of CH3OCH2
replacement of CH2O
117
95
129
115
82
169
10.9g
10.5
72-74
•
77
40
replacement of CH3OCH2
93
92
60
replacement of CH2O
•
(CH3)2CHOH
CH3OCH3
CH3CH2I
60
46
191.2
192.1
10.1
10.0
9.3
1.4
0.9
0.8
1.8
1.5
1.6
0.8
0.6
0.5
91
60
replacement of CH3OCH2
121
93
106
77
40
replacement of CH2O
•
80
replacement of CH3OCH2
92
77
156
62
94
20
replacement of CH2O
H abstractionh
e- transfer
156
=176.0
20
185i
(trace)
none
none
80
CH3SCH3
CH3SSCH3
62
94
200.6
196.0
8.7
1.5
1.8
1.4
1.9
1.1
1.0
100
100
e- transfer
8.2j
e- transfer
a Proton affinity values are from ref 23a. b Adiabatic ionization energies are from ref 23a unless otherwise noted. c Rate constant; ×10-9 cm3 s-1
correspond to the proton-bound dimer of the neutral reagent. g Value for the corresponding undeuterated reagent. h H-atom abstraction by radical
cations from alkyl iodides has been reported earlier; see, for example, ref 32b. i CH3CH2I+CH2CH3. j Value from refs 23b and 23c.
.
d Collision rate (×10-9 cm3 s-1) calculated according to ref 25c. e Efficiency ) k/kcollision
.
f All the observed secondary product ions except one
(checked using Sybyl28 molecular modeling software), harmonic
frequencies, and zero-point vibrational energies (ZPVE) were calculated
by employing spin-unrestricted Hartree-Fock (UHF) procedures carried
out up to the split-valence polarization 6-31G** basis set. Single point
energies were calculated for all the 6-31G** geometries by including
electron correlation through the use of Møller-Plesset perturbation
theory carried out to the second order (UMP2/6-31G**//UHF/6-31G**).
Four different structures (1, 1′′, 2, and 3) were also optimized at the
UMP2/6-31G** level. The force constant matrices calculated for all
the stationary points were checked to have the correct number of
negative eigenvalues (none for equilibrium structures). Since Hartree-
Fock calculations are known29 to overestimate the vibrational frequen-
cies by 10-15%, the zero-point energies given in Table 2 were scaled
by a factor of 0.9. The basis set superposition errors (BSSE) were
evaluated for three structures at the UHF/6-31G** and UMP2/6-31G**
levels by using the counterpoise (CP) correction procedure.5,30a
Figure 1. Temporal variation of ion abundandes upon reaction of CH3-
•
OCH2CH2OH•+ with CH3CH2OH to yield the CH3OCH2 -replacement
product (m/z 77) and the CH2O-replacement product (m/z 92). Both
primary product ions react with CH3CH2OH to yield the secondary
product (CH3CH2OH)2H+ (m/z 93). The linear decay of the natural
logarithm of the relative abundance of the reactant ion as a function of
time indicates that the reaction follows pseudo-first-order kinetics.
Results and Discussion
Structure of the Radical Cation of CH3OCH2CH2OH.
Neutral CH3OCH2CH2OH was ionized by electron impact or
(22) Burgers, P. C.; Holmes, J. L.; Terlouw, J. K.; van Baar, B. Org.
Mass Spectrom. 1985, 20, 202.
(23) (a) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin,
R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1988, 17, Supp. 1. (b) Li,
W.-K.; Chiu, S.-W.; Ma, Z.-X.; Liao, C.-L.; Ng, C. Y. J. Chem. Phys. 1993,
99, 8440. (c) Leeck, D. T.; Kentta¨maa, H. I. Org. Mass Spectrom. 1994,
29, 106.
(24) Wang, T.-C. L.; Ricca, T. L.; Marshall, A. G. Anal. Chem. 1986,
58, 2938.
(25) (a) Bartmess, J. E.; Georgiadis, R. M. Vacuum 1983, 33, 149. (b)
Ikezoe, Y.; Matsuoka, S.; Takebe, M.; Viggiano, A. Gas Phase Ion-
Molecule Reaction Rate Constants Through 1986; the Mass Spectroscopy
Society of Japan, Tokyo, 1987. (c) Su, T.; Chesnavich, W. J. J. Chem.
Phys. 1982, 76, 5183.
(26) Grosshans, P. B.; Shields, P.; Marshall, A. G. J. Am. Chem. Soc.
1990, 112, 1275.
(27) GAUSSIAN 92: Frisch, M. J.; Trucks, G. W.; Head-Gordon, M.;
Gill, P. M. W.; Wong, M. W.; Foresman, J. B.; Johnson, B. G.; Schlegel,
H. B.; Robb, M. A.; Replogle, E. S.; Gomperts, R.; Andres, J. L.;
Raghavachari, K.; Binkley, J. S.; Gonzalez, C.; Martin, R. L.; Fox, D. J.;
Defrees, D. J.; Baker, J.; Stewart, J. J. P.; Pople, J. A.; Gaussian, Inc.:
Pittsburgh, PA, 1992.
(28) Tripos Associates, Inc., St. Louis, MO 63144, U.S.A.
(29) Pople, J. A.; Schlegel, H. B.; Krishnan, R.; Defrees, J. D.; Binkley,
J. S.; Frisch, M. J.; Whiteside, R. A.; Hout, R. F.; Hehre, W. J. Int. J.
Quantum Chem. 1981, 15, 269.
(30) (a) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553. (b) Turi,
L.; Dannenberg, J. J. J. Phys. Chem. 1993, 97, 2488. (c) Turi, L.;
Dannenberg, J. J. J. Phys. Chem. 1993, 97, 7899. (d) Schwenke, D. W.;
Truhlar, D. G. J. Chem. Phys. 1985, 82, 2418.
•+
by using CS2 (a more gentle method) in the dual-cell of a
Fourier-transform ion cyclotron resonance (FT-ICR) mass
spectrometer, transferred into the other side of the cell, isolated,
collisionally cooled, and allowed to react with different neutral
reagents as a function of time. The measured reaction rate
constants (k), calculated collision rate constants (kcollision), and
reaction efficiencies (eff ) k/kcollision) are summarized in Table
1, together with the observed primary product ions, their
branching ratios, and some relevant thermochemical data
obtained from the literature. All the reactions were found to
follow pseudo-first-order kinetics (Figure 1), suggesting that the
reactant ions are likely to have a single structure and not a
mixture of isomeric structures. Further support for this proposal
was obtained from the finding that the product distributions and
the reaction rates are independent of the method used to generate
the reactant ion.
Reaction of the radical cation with reagents with ionization
energies (IE) lower than that of CH3OCH2CH2OH (IE ) 9.6
eV23a) generally results in exclusive electron transfer (Table 1).
For example, electron transfer dominates the reaction with CH3-
SSCH3 (IE ) 8.0 eV), CH3SCH3 (IE ) 8.7 eV), and CH3CH2I

Experimental and Theoretical Study of CH3OCH2CH2OH•+
J. Am. Chem. Soc., Vol. 118, No. 16, 1996 3917
Figure 3. (a) Generation of the ion of m/z 95 by the reaction of CH3-
OCH2CH2OH•+ with (CD3)2CO (220 ms reaction time), (b) isolation
of the ion of m/z 95 after transfer into the other side of the dual-cell
reaction chamber, and (c) collision-activated dissociation of the ion of
m/z 95 by using argon as the target gas at a nominal pressure of 1.2 ×
10-7 Torr (100 eV estimated collision energy, 100 ms reaction time).
fact, this characteristic behavior has been used to identify many
distonic ions.33-35 Hence, the absence of CH3S• abstraction,
together with the occurrence of electron transfer, indicates that
the radical cation of CH3OCH2CH2OH does not have a distonic
structure. This finding, together with the other results discussed
above, strongly suggests that the radical cation of CH3OCH2-
CH2OH has retained the connectivity of its neutral precursor.
Replacement Reactions of CH3OCH2CH2OH•+. The ion
CH3OCH2CH2OH•+ shows unusual reactivity toward reagents
with ionization energies above that of CH3OCH2CH2OH: facile
Figure 2. A typical MS/MS/MS experiment illustrated for CH3CH2-
CHO. (a) Ionization of CH3OCH2CH2OH by electron transfer using
CS2•+, (b) isolation of CH3OCH2CH2OH•+ (m/z 76) after transfer into
the other side of the dual-cell reaction chamber (CS2•+ ions are among
the isolated ions since CS2•+ has the same nominal mass value as CH3-
OCH2CH2OH•+), (c) reaction of CH3OCH2CH2OH•+ with CH3CH2CHO
(400 ms reaction time) to yield a major primary product ion of m/z 89
and a secondary product ion of m/z 117 (the peaks at m/z 57-59 were
•
•+
replacement of CH2O or CH3OCH2 or both occurs (Table 1;
verified to arise from reactions of CS2 with CH3CH2CHO by
examining reactions of CH3OCH2CH2OH•+ generated in the absence
of CS2), (d) isolation of the primary product ion of m/z 89, and (e)
reaction of this product ion with CH3CH2CHO to yield the ion of m/z
117 (400 ms reaction time).
Figures 2 and 3). These reactions are discussed below
separately.
•
(a) CH3OCH2 Replacement. The reagents CH3CH2OH,
(CH3)2CHOH, CH3OCH3, CH3CHO, CH3CH2CHO, (CD3)2CO,
and cyclopentanone react with CH3OCH2CH2OH•+ by replace-
•
(IE ) 9.3 eV). This result is expected for a radical cation with
the same connectivity as in CH3OCH2CH2OH. If the connec-
tivities were different, the recombination energy of the ion would
be likely to be significantly lower than the ionization energy of
CH3OCH2CH2OH.31 For example, electron transfer is not
commonly observed for the radical cations of molecules that
after ionization isomerize to a distonic ion13b,32 or to a hydrogen-
bridged structure.4a
The observation of exclusive electron transfer upon reaction
of the radical cation of CH3OCH2CH2OH with CH3SSCH3 is
especially noteworthy. Most known distonic ions react with
CH3SSCH3 by abstracting CH3S• or are unreactive.13b,32-35 In
ment of CH3OCH2 (Table 1). This reaction is probably initiated
by formation of a strong hydrogen bond between the hydroxyl
group of the ion and the entering neutral reagent. The energy
released upon the hydrogen bond formation (approximately 30
kcal/mol36) likely induces the cleavage of the C-C bond in CH3-
OCH2CH2OH•+ (Scheme 3). A precedent to this type of
reactivity, cleavage of the C-C bond in CH3CH2OH•+ upon
hydrogen bond formation, has been observed recently by
others.37 In support of the above mechanism, the reaction
efficiency (k/kcollision) was found to be the highest for the neutral
reagents with the greatest proton affinities (Table 1). For
example, cyclopentanone (PA ) 198.8 kcal/mol23a) replaces
(31) For example, electron abstraction by 2 (∆Hf = 134 kcal/mol;
assumed to be the same as ∆Hf(1′′) as the calculations suggest; see reference
23a) would yield as the neutral products CH2dO (-26 kcal/mol; ref 23a),
CH3OH (-48.2 kcal/mol; ref 23a), and CH2: (93 kcal/mol; ref 23a), or
possibly CH2OH (-6.2 kcal/mol; ref 23a) and CH3OCH2 (-3 kcal/mol;
ref 23a). Both pathways are associated with an energy release of only 5-6
eV. This is much less than is required for electron abstraction from the
reagents studied here.
(32) (a) Leeck, D. T.; Stirk, K. M.; Zeller, L. C.; Kiminkinen, L. K. M.;
Castro, L. M.; Vainiotalo, P.; Kentta¨maa, H. I. J. Am. Chem. Soc. 1994,
116, 3028. (b) Stirk, K. M.; Smith, R. L.; Orlowski, J. C.; Kentta¨maa, H.
I. Rap. Commun. Mass Spectrom. 1993, 7, 392.
(33) Stirk, K. M.; Orlowski, J. C.; Leeck, D. T.; Kentta¨maa, H. I. J. Am.
Chem. Soc. 1992, 114, 8604.
(34) Smith, R. L.; Franklin, R. L.; Stirk, K. M.; Kentta¨maa, H. I. J. Am.
Chem. Soc. 1993, 115, 10348.
(35) Smith, R. L.; Kentta¨maa, H. I. J. Am. Chem. Soc. 1995, 117, 1393.
(36) Formation of a hydrogen bond between two simple, gaseous,
oxygen-containing molecules is typically associated with ∆H ≈ -30 kcal/
mol; e.g., for CH3+CH2OH2+ + CH3CH2OH, ∆H ) -32 kcal/mol; for CH3-
CH2CH2CH2OH2 + CH3CH2CH2OH, ∆H ) -32 kcal/mol; for (CH3-
CH2)2OH+ + (CH3)2CO, ∆H ) -29 kcal/mol. See: Keesee, R. G.;
Castleman, A. W., Jr. J. Phys. Chem. Ref. Data 1986, 15.
(37) McMahon, T. B. Personal communication.
•
•

3918 J. Am. Chem. Soc., Vol. 118, No. 16, 1996
Pakarinen et al.
Scheme 3
•
CH3OCH2 at collision rate (efficiency 1.0). The less basic
reagents (CD3)2CO (PA ) 196.7 kcal/mol) and CH3CH2CHO
(PA ) 189.6 kcal/mol) react at a lower efficiency (0.7-0.8),
and CH3CH2OH (PA ) 188.3 kcal/mol) and CH3CHO (186.6
kcal/mol) at an even lower efficiency (0.3-0.4). The least basic
reagents CD3OD (PA ) 182 kcal/mol) and CH3CH2I (PA )
176 kcal/mol) do not react with CH3OCH2CH2OH•+ by CH3-
•
OCH2 replacement (Table 1).
Figure 4. Spectra measured at the final stage of an MS/MS/MS
experiment carried out by using CH3CHO as the neutral reagent
(nominal pressure 1.2 × 10-7 Torr): CH3OCH2CH2OH was ionized
by electron transfer using CS2•+, CH3OCH2CH2OH•+ was isolated after
transfer into the other side of the dual cell and allowed to react with
Examination of the structure of the product ions provides
further support for the mechanism proposed above. According
to this mechanism (Scheme 3), the CH3OCH2 -replacement
product is a proton-bridged complex of the neutral reagent and
CH2O. An ion with this structure would be expected to react
with nucleophiles by exclusive and facile replacement of the
weaker base, CH2O. This is exactly what was observed: all
the CH3OCH2 -replacement products undergo a facile replace-
ment of CH2O, giving rise to the protonated dimer of the neutral
reagent.
•
•+
•+
CH3CHO (CS2 ions were among the isolated ions since CS2 has
the same nominal mass value as CH3OCH2CH2OH•+; however, this
does not affect the findings since CS2•+ is unreactive toward CH3CHO).
The two product ions formed (m/z 75 and 90) were isolated. Upon
reaction with CH3CHO, both ions yield an ion of m/z 89: the ion of
m/z 75 upon replacement of CH2O (Figure 4a; 200 ms reaction time)
•
•
and the ion of m/z 90 upon replacement of CH3OCH2 (Figure 4b;
reaction time 300 ms).
The reaction sequence leading to the protonated dimer of the
neutral reagent was examined by multistage MS/MS/MS experi-
ments. The experiment performed with CH3CH2CHO is illus-
trated in Figure 2. The ion CH3OCH2CH2OH•+ was generated
in one side of the dual cell (Figure 2a), transferred into the other
cell, isolated (Figure 2b), and allowed to react with CH3CH2-
95; Figure 3a), transferred into the other side of the cell, isolated
(Figure 3b), accelerated to different laboratory kinetic energies
(12-100 eV), and allowed to collide with argon. As expected38
for the hydrogen-bridged ion (CD3)2CdO‚‚‚H+‚‚‚OdCH2 that
is composed of two molecules with different proton affinities,
the weaker base (CH2O) is preferentially cleaved off upon
excitation (CH2O: PA ) 171.7 kcal/mol;23a (CD3)2CO: PA )
196.7 kcal/mol23a). Formation of (CD3)2COH+ dominates the
reaction at all collision energies (m/z 65; Figure 3c). This result
provides strong support for the proposed hydrogen-bridged
•
CHO (Figure 2c) to form the CH3OCH2 -replacement product
expected to have the structure CH3CH2CHdO‚‚‚H+‚‚‚OdCH2
(m/z 89). This product ion was isolated (Figure 2d) and allowed
to react with CH3CH2CHO (Figures 2d and 2e). Substitution
of CH2O (PA ) 171.7 kcal/mol;23a CH3CH2CHO: PA ) 189.6
kcal/mol23a) occurred to yield CH3CH2CHdO‚‚‚H+‚‚‚OdCHCH2-
CH3 as the only observed product (m/z 117). A similar reaction
sequence carried out for CH3CHO is illustrated in Figure 4.
•
structure (CD3)2CdO‚‚‚H+‚‚‚OdCH2 for this CH3OCH2
replacement product.
(b) CH2O Replacement. Neutral reagents with proton
affinities close to or less than 190 kcal/mol, i.e., CH3OH (PA
) 182 kcal/mol), CH3CHO (PA ) 186.6 kcal/mol), CH3CH2-
OH (PA ) 188.3 kcal/mol), (CH3)2CHOH (PA ) 191.2 kcal/
•
The CH3OCH2 -replacement product CH3CHdO‚‚‚H+‚‚‚OdCH2
(m/z 75, Figure 4a) was found to react with CH3CHO (PA )
186.6 kcal/mol23a) by exclusive CH2O replacement to yield CH3-
CHdO‚‚‚H+‚‚‚OdCHCH3 as the only primary product (m/z 89,
Figure 4b).
(38) See, for example: (a) Cooks, R. G.; Patrick, J. S.; Kotiaho, T.;
McLuckey, S. A. Mass Spectrom. ReV. In press. (b) McLuckey, S. A.;
Cameron, D.; Cooks, R. G. J. Am. Chem. Soc. 1981, 103, 1313. (c) Turecek,
F. In Supplement E: The chemistry of hydroxyl, ether and peroxide groups;
Patai, E. S., Ed.; John Wiley & Sons Ltd.: Chichester, 1993; Vol. 2, pp
395-399.
•
The structure of the CH3OCH2 -replacement product formed
upon reaction of CH3OCH2CH2OH•+ with (CD3)2CO was also
examined by collision-activated dissociation. The product ion
was generated in one side of the dual-cell reaction chamber (m/z

Experimental and Theoretical Study of CH3OCH2CH2OH•+
J. Am. Chem. Soc., Vol. 118, No. 16, 1996 3919
Table 2. Calculated Total Energies (hartrees), Relative Energies in Parentheses (kcal/mol), Zero-Point Vibrational Energies (ZPVE, in
kcal/mol), and Final Relative Energies (kcal/mol) of the [C3H8O2]•+ Structures
UMP2/6-31G**//
UMP2/6-31G**//
compd
UHF/3-21G UHF/6-31G UHF/6-31G* UHF/6-31G** UHF/6-31G** ZPVEa UHF/6-31G**+ZPVE
CH3OCH2CH2OH•+ (1)
-266.18356 -267.54036 -267.64794 -267.66625
(0) (0) (0) (0)
-266.17062 -267.52931 -267.63874 -267.65719
(8.1) (6.9) (5.8) (5.7)
-266.15514 -266.51113 -267.63737 -267.65603
(17.8) (18.3) (6.6) (6.4)
-268.41105
(0)
67.77
67.13
67.92
64.80
65.48
67.67
68.18
68.20
65.36
-268.30305
(0)
CH3OCH2CH2OH•+ (1′)
-268.39981
-268.29283
(7.1)
(6.4)
CH3OCH2‚‚‚CH2OH•+ (1′′)
-268.42891
(-11.2)
-268.32067
(-11.1)
CH3(CH2)O-H‚‚‚OdCH2•+ (2) -266.18686 -267.54109 -267.64543 -267.66491
-268.41952
(-5.3)
-268.31625
(-8.3)
(-2.1)
-266.19700 -267.54882 -267.65828 -267.67674
(-8.4) (-5.3) (-6.5) (-6.6)
-266.18262 -267.53047 -267.64073 -267.65961
(0.6) (6.2) (4.5) (4.2)
-266.19987 -267.54702 -267.65613 -267.66922
(-10.2) (-4.2) (-5.1) (-1.9)
-266.19223 -267.54449 -267.64260 -267.66596
(-5.4) (-2.6) (3.4) (0.2)
CH3OC(H)H‚‚‚O(H)dCH2•+ (7) -266.14807 -267.51759 -267.63415 -267.65340
(22.3) (14.3) (8.7) (8.1)
(-0.5)
(1.6)
(0.8)
(CH3)2O-H‚‚‚OdCH•+ (3)
CH3O(CH2)CH2OH•+ (4)
(CH3)2OCH2O•+ (5)
-268.43756
(-16.6)
-268.33321
(-18.9)
-268.41809
(-4.4)
-268.31025
(-4.5)
-268.41133
-268.30267
(0.2)
(0.2)
CH2OCH2CH2OH2•+ (6)
-268.42596
(-9.4)
-268.40414
(4.3)
-268.31727
(-8.9)
-268.29998
(1.9)
a Calculated at the UHF/6-31G** level of theory and scaled by a factor of 0.9.
Table 3. Counterpoise Corrections for the Basis Set Superposition
Errors (BSSE), e, Calculated with the UHF/6-31G** Basis Set
(values in kcal/mol)
mol), and CH3OCH3 (PA ) 192.1 kcal/mol), replace CH2O in
CH3OCH2CH2OH•+. This reaction is likely initiated by CH3-
OCH2• replacement within the collision complex, as discussed
above. If the CH3OCH2• replacement is highly exothermic, the
structure
e(A)a
e(B)b
e(sum)
CH3OCH2‚‚‚CH2OH•+ (1′′)
e(CH3OCH2
0.62
)
e(•CH2OH)
1.72
+
•
resulting complex dissociates rapidly to CH3OCH2 and the
2.34
1.03
0.79
hydrogen-bridged complex of CH2O and the neutral reagent (i.e.,
CH3(CH2)O-H‚‚‚OdCH2•+ (2) e(CH3(-CH2)O+H) e(CH2O)
•
CH3OCH2 replacement occurs; some of these reactions were
0.16
0.87
e(•CHO)
0.65
(CH3)2O-H‚‚‚OdCH•+ (3)
e((CH3)2O+H)
0.14
found to be so exothermic that the ion produced is metastable
and slowly dissociates by the loss of CH2O). However,
reactions of less basic reagents will yield a lower-energy and
longer-lived ion-molecule complex. Within this complex, CH3-
a e(A) is the correction obtained for the fragment A in the presence
of the atomic orbitals of the B fragment. b e(B) is the correction obtained
for the fragment B in the presence of the atomic orbitals of the A
fragment.
•
•
OCH2 can react with the CH3OCH2 -replacement product
(Scheme 3) to replace CH2O (the weaker base: PA ) 172 kcal/
•
mol for CH2O; PA ) 173 kcal/mol39 for CH3OCH2 ).
Table 4. The Energies (hartrees), Relative Energies in Parentheses
(kcal/mol), Zero-Point Vibrational Energies (ZPVE, in kcal/mol),
and Final Relative Energies (kcal/mol) of Four-Structures Calculated
at the Electron Correlation Level
According to the mechanism proposed above, the final ionic
•
product is a hydrogen-bridged complex of CH2OCH3 and the
neutral reagent molecule. Examination of the reactivity of the
product ion (m/z 90) formed upon the reaction of CH3CHO
suggests that this ion indeed has the expected structure •CH2O-
(CH3)‚‚‚H+‚‚‚OdCHCH3: this ion reacts with CH3CHO (PA
UMP2/
UMP2/
compd
6-31G** ZPVEa 6-31G**+ZPVE
CH3OCH2CH2OH•+ (1)
-268.41491 64.21
(0)
-268.43540 66.73
(-12.9)
-268.31257
(0)
•
) 186.6 kcal/mol23a) by a facile replacement of CH3OCH2 (PA
CH3OCH2‚‚‚CH2OH•+ (1′′)
-268.32906
) 173 kcal/mol39) to generate CH3CHdO‚‚‚H+‚‚‚OdCHCH3
(m/z 89, Figure 4c, 4f, and 4g).
(-10.4)
CH3(CH2)O-H‚‚‚OdCH2•+ (2) -268.42395 62.04
(-5.7)
-268.32508
(-7.9)
It is noteworthy that the reactions of CH3OCH2CH2OH•+ lead
to the formation of stable hydrogen-bridged complexes of the
(CH3)2O-H‚‚‚OdCH•+ (3)
-268.44110 63.07
(-16.4)
-268.34059
(-17.6)
•
neutral reagent and CH3OCH2 . The ion CH3OCH2CH2OH•+
a Calculated at the UMP2/6-31G** level of theory and scaled by a
factor of 0.9.
provides a tool for the strategic synthesis of a variety of
hydrogen-bridged complexes of the type •CH2O(CH3)‚‚‚H+‚‚‚N
wherein N is the neutral reagent used.
optimized at the UMP2/6-31G** level (Figure 5; Tables 4 and
5). The lowest energy structures obtained at this level of theory
differ in some respects from those obtained at the UHF level
(UHF/6-31G**). For example, the long bond lengths in the
UHF/6-31G** optimized structures 1′′ (C2-C3), 2 (O2-H6), and
3 (O2-H7) are shorter at the electron correlation level (Figure
5). These results are discussed in detail below. The extent of
spin contamination, as indicated by S2 , was found to be within
an acceptable range (0.756-0.767) and close to the value of
0.75 for the pure spin state in all the calculations.
Molecular Orbital Calculations. Ab initio molecular orbital
calculations using the GAUSSIAN9227 program were performed
•+
to study the relative stabilities of seven isomeric C3H8O2
structures. Several possible conformers were examined for each
isomer. Only the most stable conformers are discussed here
(see Tables 2-5), with the exception of the ion CH3OCH2CH2-
OH•+. For this ion, results obtained for three different
conformations are presented (1, 1′, and 1′′). All the structures
were optimized in a stepwise manner up to the UHF/6-31G**
level of theory. The structures 1, 1′′, 2, and 3 were also
Basis set superposition errors (BSSE) were evaluated for the
ions 1′′, 2, and 3 by using the counterpoise method, expected
to provide an upper limit for BSSE.30 For an ion A‚‚‚B, this
method involves calculating the sum (e(sum); see Tables 3 and
5) of the energy differences (e(A) and e(B)) between the energies
•
(39) Estimated from the heat of formation of CH3OCH2 ) -3 kcal/
mol (ref 23a) and from a calculated heat of formation of CH3O(H+)CH2
) 190 kcal/mol; see: (a) Bouma, W. J.; Nobes, R. H.; Radom, L. Org.
Mass Spectrom. 1982, 17, 315. (b) Bouma, W. J.; Nobes, R. H.; Radom,
L. J. Am. Chem. Soc. 1983, 105, 1746.
•

3920 J. Am. Chem. Soc., Vol. 118, No. 16, 1996
Pakarinen et al.
Table 5. Counterpoise Corrections for the Basis Set Superposition
Errors (BSSE), e, Calculated with the UMP2/6-31G** Basis Set (in
kcal/mol)
structure
e(A)a
e(B)b
e(sum)
CH3OCH2‚‚‚CH2OH•+ (1′′)
e(CH3OCH2 )
3.63
e(CH2OH+)
2.00
•
5.63
3.26
2.08
CH3(CH2)O-H‚‚‚OdCH2•+ (2) e(CH3(•CH2)O+H) e(CH2O)
0.62
2.64
e(•CHO)
1.64
(CH3)2O-H‚‚‚OdCH•+ (3)
e((CH3)2O+H)
0.44
a e(A) is the correction obtained for the fragment A in the presence
of the atomic orbitals of the B fragment. b e(B) is the correction obtained
for the fragment B in the presence of the atomic orbitals of the A
fragment.
of the isolated fragments (E(A) and E(B)) and the energy of
the corresponding fragment with the basis set of the whole ion
(e.g., for A, this involved calculation of EAB(A) by including
the ghost orbitals of B; e(sum) ) e(A) + e(B) ) [E(A) - EAB
-
(A)] + [E(B) - EAB(B)]).5,30 The counterpoise estimates were
found to be relatively small (1-2 kcal/mol) at the UHF level
but greater (2-6 kcal/mol) at the UMP2 level, as expected.5,30
The orbital space of the fragments •CH2OH, CH2O, and •CHO
+
is more affected by the ghost orbitals than that of CH3OCH2
,
CH3(CH2)OH•+, and (CH3)2OH+. Overall, the counterpoise
estimates suggest that BSSE has only a small effect on the
relative energies of the ions 1′′, 2, and 3. At the UMP2/6-
31G**+ZPVE level of theory, the relative energies of 1′′, 2,
and 3 were found to change from 0, 2.5, and -7.2 kcal/mol to
0, 0.1, and -10.8 kcal/mol, respectively.
The energies of 1′ and 2 (relative to that of 1) were found to
be fairly independent of the level of theory employed. The
relative stability of the structure 5 decreases with the increasing
size of the basis set but eventually converges to that of 1. The
opposite is true for the remaining structures (1′′, 3, 4, 6, and
7). The relative energies of the structures 1′′, 3, 4, and 6
decrease most dramatically when electron correlation was
included, e.g., for 3, UHF/6-31G** yields -6.6 kcal/mol Vs
UMP2/6-31G**//UHF/6-31G** yields -16.6 kcal/mol and
UMP2/6-31G** yields -16.4 kcal/mol. The relative energies
calculated at the UMP2/6-31G**+ZPVE level for the structures
optimized at UHF/6-31G** and UMP2/6-31G** levels were
found to be nearly equal.
The relative energies of a large number of conformations of
CH3OCH2CH2OH have been calculated earlier.18 The calcula-
tions of the covalently bonded CH3OCH2CH2OH•+ included the
most stable of these conformations as the initial input structures,
such as the hydrogen-bridged structure shown in Scheme 1. All
these conformers were allowed to optimize freely (no geometry
constraints were employed). None of the final optimized
structures contained stabilizing intramolecular hydrogen bond-
ing. For example, optimization of the hydrogen-bridged
structure shown in Scheme 1 yielded 1 as the optimized structure
(Figure 5).
The ion CH3OCH2CH2OH•+ with the linear geometry 1′ lies
6.4 and 17.5 kcal/mol higher in energy, respectively, than the
two bent structures 1 and 1′′ (UMP2/6-31G**//UHF/6-
31G**+ZPVE level of theory; Table 2; Figure 5). The most
stable structure 1′′ is characterized by a remarkably long C-C
bond (1.969 Å Vs 1.521 Å (1) and 1.518 Å (1′′); UHF/6-31G**).
The structure 1′′ also contains an unusually short CH3O-CH2
bond (1.285 Å Vs 1.466 Å (1) and 1.464 Å (1′)) and a slightly
shorter HO-CH2 bond than the other two structures (1.308 Å
Vs 1.391 Å (1) and 1.386 Å (1′)). Hence, the structure 1′′ is
best described as a loosely bonded complex of CH3OCH2• and
CH2OH+. At the UMP2/6-31G** level of theory (Figure 5),
the minimum energy geometry of 1′′ is similar to that obtained
Figure 5. UHF/6-31G** and UMP2/6-31G** (marked as such)
optimized geometries for different [C3H8O2]•+ isomers.
at the UHF/6-31G** level. However, the C-C bond in the
UMP2/6-31G** optimized structure is slightly shorter (1.775
Å). The presence of a long C-C bond in an organic radical
cation is not unprecedented: the ions CH3CH2OH•+ and HOCH2-
CH2OH•+ have been calculated to have structures with remark-

Experimental and Theoretical Study of CH3OCH2CH2OH•+
J. Am. Chem. Soc., Vol. 118, No. 16, 1996 3921
•
ably long one-electron C-C bonds (CH3CH2OH•+, 2.013 Å at
the UHF/4-31G level;40a,b HOCH2CH2OH•+, 2.01 Å at the RHF/
DZP level40c). The results obtained in the above calculations,
especially the long C-C bond in 1′′, as well as the location of
the greatest positive charge at the hydroxyl hydrogen in this
ion (Figure 5), support the proposal that the entering nucleophile
initially forms a hydrogen bond with the hydroxyl hydrogen,
as shown in Scheme 3.
examined, CH2OH+ and CH3OCH2 are connected chiefly by
weak electrostatic forces, as indicated their great separation
(2.071 Å). This ion was found to lie significantly higher in
energy (13 kcal/mol) than the ion 1′′.
Conclusions
Examination of the gas-phase reactions of the long-lived, low-
energy radical cation generated from CH3OCH2CH2OH suggests
that the connectivity of CH3OCH2CH2OH does not change upon
ionization. The ion CH3OCH2CH2OH•+ undergoes gas-phase
reactions quite unique for an organic radical cation: neutral
As pointed out above, the internally excited ion CH3OCH2-
CH2OH•+ has been suggested to equilibrate with the distonic
ions •CH2O+(CH3)CH2OH (4), (CH3)2O+CH2O• (5), and •CH2-
OCH2CH2OH2+ (6) prior to or during dissociation (Scheme 2).20
The most stable geometries found for these distonic ions lie
6.9, 11.3, and 2.2 kcal/mol higher in energy than 1′′, respectively
(at the UMP2/6-31G**//UHF/6-31G**+ZPVE level of theory).
Hence, it is not surprising that these distonic ions are not
generated from the nondissociating ion CH3OCH2CH2OH•+ in
•
reagents replace CH3OCH2 in the ion. The reaction is likely
initiated by the formation of a strong hydrogen bond between
the entering nucleophile and the hydroxyl group in CH3OCH2-
CH2OH•+. The exothermicity of the hydrogen bond formation
(about 30 kcal/mol) drives the cleavage of the C-C bond. A
precedent to this type of reactivity has been proposed earlier
for CH3CH2OH•+.37 Dissociation of the internally excited ions
CH3OCH2CH2OH•+ by the loss of CHO• (Scheme 2) is also
thought to be initiated by the cleavage of the C-C bond.20 All
these findings are in agreement with the finding of a remarkably
long C-C bond in the most stable structure calculated for CH3-
OCH2CH2OH•+ (UMP2/6-31G**).
Weak nucleophiles react with CH3OCH2CH2OH•+ through
an additional channel. Replacement of CH3OCH2 with weak
nucleophiles yields a long-lived ion-molecule complex that can
undergo further reactions. Within this complex, CH3OCH2
replaces CH2O (a weaker base) to form the hydrogen-bridged
complex of CH3OCH2 and the neutral reagent. This reaction
+
the experiments discussed above. The ion •CH2OCH2CH2OH2
(6) is more stable by 5.6 kcal/mol and 9.0 kcal/mol, respectively,
than the distonic ions •CH2O+(CH3)CH2OH (4) and (CH3)2O+-
CH2O• (5) (at the UMP2/6-31G**//UHF/6-31G**+ZPVE level
of theory). The most stable geometry of the ion 6 is stabilized
by an intramolecular hydrogen bond (1.908 Å) between the
C-O-C oxygen and one of the C-O-H hydrogens (note the
small O-CH2-CH2, CH2-CH2-OH2, and CH2-O-H bond
angles: 104.3°, 103.5°, and 108.4°, respectively). The length
of the CH2O-CH2 bond (1.378 Å) indicates that the bond has
considerable double bond character.
•
•
•
The most stable geometry of the ion CH2O(CH3)-H+‚‚‚
•
OdCH2 (2) was found to be slightly less stable than 1′′ (by 2.5
kcal/mol at the UMP2/6-31G**+ZPVE level of theory). In
contrast, the ion (CH3)2O-H+‚‚‚OdCH• (3) is more stable than
1′′ by 7.2 kcal/mol (at the UMP2/6-31G**+ZPVE level of
theory). Experimental evidence does not support the formation
of either one of these ions upon ionization of CH3OCH2CH2-
OH. Exclusive replacement of CH2O is expected for an ion
with structure 2 since CH2O is bonded less strongly than CH3-
provides a general synthetic route to stable hydrogen-bridged
radical cations in the gas phase.
Ab initio molecular orbital calculations (UMP2/6-31G**//
UHF/6-31G**+ZVPE) suggest that the ion CH3OCH2CH2OH•+
lies 6.6, 11.3, and 2.2 kcal/mol lower in energy than the distonic
ions •CH2O+(CH3)CH2OH (4), (CH3)2O+CH2O• (5), and •CH2-
+
OCH2CH2OH2 (6), respectively. Hence, although the inter-
•
•
nally excited radical cation CH3OCH2CH2OH•+ may equilibrate
with these distonic ions prior to or during dissociation (Scheme
2),20 it is not surprising that the long-lived, low-energy ions do
not undergo equilibration.
OCH2 in the ion (the -O‚‚‚H‚‚‚O bond length to CH3OCH2
is 0.999 Å while that to CH2O is 1.525 Å). In the ion 3, HCO•
is the less strongly bonded fragment (the hydrogen bond to CH3-
OCH3 is 0.973 Å while that to HCO• is 1.687 Å). Hence, this
ion would be identified by a facile replacement of HCO•, a
reaction that was not observed in any of the experiments.
Acknowledgment. The Academy of Finland, University of
Joensuu, and the Finnish Chemical Society, as well as the
National Science Foundation (CHE-9409644), are thanked for
partial financial support of this work. Chris L. Stumpf is
thanked for running some of the experiments.
•+
Finally, the complex CH3OC(H)H‚‚‚O(H)dCH2 (7) was
also examined computationally. In the most stable geometry
(40) (a) Bouma, W. J.; Dawes, J. M.; Radom, L. Org. Mass Spectrom.
1983, 18, 12. (b) Postma, R.; Ruttink, P. J. A.; Van Baar, B.; Terlouw, J.
K.; Holmes, J. L.; Burgers, P. C. Chem. Phys. Lett. 1986, 123, 409. (c)
Ruttink, P. J. A.; Burgers, P. C. Org. Mass Spectrom. 1993, 28, 1087.
JA952145P