Hydrogen atom adducts to the amide bond. Generation and energetics of amide radicals in the gas phase

Article abstract of DOI:10.1021/jp021402r

The 1-hydroxy-1-(N-methyl)aminoethyl radical (1) represents a simple model system for hydrogen atom adducts to the amide bond in gas-phase peptide and protein ions relevant to electron capture dissociation (ECD). Radical 1 was generated in the rarefied gas phase by femtosecond electron transfer to the stable cation prepared by selective O-protonation of N-methylacetamide. The main dissociations of 1 were loss of the hydroxyl hydrogen atom and the N-methyl group in a 1.7:1 ratio, as deduced from product analysis and deuterium labeling. The dissociations that occur on the 4.1 microsecond time scale are driven by large Franck-Condon effects on collisional electron transfer that deposit 93-103 kJ mol^-1 in the nascent radicals. Detailed analysis of the potential energy surface for dissociations of 1 revealed several conformers and isomeric transition states for dissociations of the O-H and N-CH₃ bonds. The experimental branching ratio is in quantitative agreement with RRKM calculations within the accuracy of the G2 potential energy surface and favors cleavage of the O-H bond in 1 and loss of H. This finding contrasts previously reported results, as discussed in the paper.

Full text of DOI:10.1021/jp021402r

J. Phys. Chem. A 2003, 107, 115-126
115
Hydrogen Atom Adducts to the Amide Bond. Generation and Energetics of Amide Radicals
in the Gas Phase
Erik A. Syrstad, Daniel D. Stephens, and Frantisˇek Turecˇek*
Department of Chemistry, Bagley Hall, Box 351700, UniVersity of Washington,
Seattle, Washington 98195-1700
ReceiVed: June 12, 2002
The 1-hydroxy-1-(N-methyl)aminoethyl radical (1) represents a simple model system for hydrogen atom adducts
to the amide bond in gas-phase peptide and protein ions relevant to electron capture dissociation (ECD).
Radical 1 was generated in the rarefied gas phase by femtosecond electron transfer to the stable cation prepared
by selective O-protonation of N-methylacetamide. The main dissociations of 1 were loss of the hydroxyl
hydrogen atom and the N-methyl group in a 1.7:1 ratio, as deduced from product analysis and deuterium
labeling. The dissociations that occur on the 4.1 microsecond time scale are driven by large Franck-Condon
effects on collisional electron transfer that deposit 93-103 kJ mol-1 in the nascent radicals. Detailed analysis
of the potential energy surface for dissociations of 1 revealed several conformers and isomeric transition
states for dissociations of the O-H and N-CH3 bonds. The experimental branching ratio is in quantitative
agreement with RRKM calculations within the accuracy of the G2 potential energy surface and favors cleavage
of the O;H bond in 1 and loss of H. This finding contrasts previously reported results, as discussed in the
paper.
Introduction
SCHEME 1
Reactions of amino acids, peptides, and proteins with radicals
generated in solution typically involve H atom abstraction
followed by fragmentations, rearrangements, dimerization, and
disproportionation of the transient reactive intermediates.1 These
chemical processes are of importance to areas ranging from
medicine (aging, radiation damage, and oxidative stress)2 to
nuclear waste storage.3 Analogous unimolecular dissociations
have been observed for gas-phase peptide radical cations that
undergo backbone cleavages and side-chain fragmentations upon
collisional activation.4 A different and unique type of reactivity
has been reported for multiply protonated gas-phase peptide and
protein cations that were partially reduced by capturing a thermal
electron. This process, called electron capture dissociation
(ECD),5 results inter alia in backbone cleavage of the peptide
ion that is of potential use for determining the amino acid
sequence. A mechanism has been suggested (Scheme 1)
consisting of electron capture by a charged group in the peptide
cation, e.g., the ꢀ-ammonium group of a protonated lysine
residue or the protonated guanidine group of an arginine residue,
to form a transient radical. This primary intermediate is expected
to dissociate rapidly and exothermically by loss of a hydrogen
atom that is captured by an adjacent amide bond to form an
isomeric CR′-C•(OH)-NH-CR radical. A radical-induced
N-CR bond cleavage in the latter results in fragmentation that
yields structurally significant products belonging to the c and z
series of sequence ions6 (Scheme 1).
and the ensuing multitude of dissociation pathways. In addition,
ECD studies of peptide ions sometimes fail to account for
complementary dissociation fragments of the c and z ion series,
so that the postulated mechanisms cannot be verified by product
analysis. Thus, it is advantageous to study simpler models that
are amenable to high-level ab initio calculations to provide
dissociation and transition state energies and that can be studied
by experimental methods that allow complete analysis of radical
dissociation products. In their early ECD study,5b Zubarev et
al. reported density functional theory calculations of dissocia-
tions of a hydrogen atom adduct to N-methylacetamide (1) which
was used as a model system. They concluded that although
dissociation of the O-H bond in 1 should predominate at low
excitation energies, at higher vibrational excitations the radical
Although Scheme 1 provides a plausible explanation of some
of the dissociations observed in ECD, the nature of the amide
radical intermediates, their dissociation energetics, and kinetics
are virtually impossible to investigate with large peptide or
protein ions because of the multitude of ion structures involved
* To whom correspondence should be addressed. Phone: (206) 685-
2041. Fax: (206) 685-3478. E-mail turecek@chem.washington.edu.
10.1021/jp021402r CCC: $25.00 © 2003 American Chemical Society
Published on Web 12/04/2002

116 J. Phys. Chem. A, Vol. 107, No. 1, 2003
Syrstad et al.
SCHEME 2
using ammonia and acetone as the CI reagent gas. Similarly,
selective protonation or deuteronation of 2 and 2a with acetone
and acetone-d6 yielded cations 1a+-1c+. Cations 1d+ and 1e+
were generated via protonation of 2b and 2c, respectively, with
ammonia. In all cases, CI reagent gas pressure was maintained
between 0.5 and 2.6 × 10-4 Torr as read on an ionization gauge
near the diffusion pump intake. Typical ionization conditions
were as follows: electron energy, 87-110 eV; emission current,
1 mA; and temperature, 188-227 °C. Cation-radical 2+• was
generated in a standard electron ionization (EI) source. Typical
ionization conditions were as follows: electron energy, 70 eV;
emission current, 500 µA; and temperature, 175-265 °C. Stable
precursor ions were passed through a quadrupole mass filter
operated in the radio frequency-only mode, accelerated to a total
kinetic energy of 8250 eV, and neutralized in the collision cell
floated at -8170 V. Dimethyl disulfide (DMDS) was introduced
into the differentially pumped collision cell at a pressure that
resulted in 70% transmittance of the precursor ion beam. The
ions and neutrals were allowed to drift to a four segment
conduit,9 where the ions were reflected by the first segment
floated at +250 V. Neutralization-collisional activation-reion-
ization (NCR) experiments were performed by introducing
helium as a collision gas into the conduit, at a pressure resulting
in 50% transmittance of the precursor ion beam. The neutral
flight times in standard neutralization-reionization mass spec-
trometry (NRMS) measurements were 4.1 µs. The fast neutral
species were reionized in the second collision cell with oxygen
at a pressure that resulted in 70% transmittance of the precursor
ion beam. The ions formed in the second collision cell were
decelerated, energy filtered, and analyzed by a second quadru-
dissociation should be dominated by N-CH3 bond cleavage
analogous to dissociation leading to c and z ion series in larger
peptide ions (Scheme 2).6
We now report a combined experimental and computational
study of 1, several of its isomers, and dissociation products.
Radical 1 and its deuterium-labeled analogues 1a-1e are
generated specifically by femtosecond electron transfer in the
gas phase, and their unimolecular dissociations are investigated
by neutralization-reionization mass spectrometry7 that provides
detailed product analysis. We also report dissociation and
transition state energies, obtained at high levels of ab initio
theory, that are further used for Rice-Ramsperger-Kassel-
Marcus (RRKM) calculations of unimolecular dissociation rate
constants. We wish to point out that, despite its structural
simplicity, 1 represents an intrinsically interesting system that
shows a complex potential energy surface for unimolecular
dissociations. In particular, we wish to compare the results of
our high-level study with those of Zubarev et al.5b regarding
radical 1 reactivity.
pole mass filter operated at unit mass resolution. The instrument
+•
was tuned daily to maximize the ion current of reionized CS2
.
Typical spectra consisted of 35-65 accumulated repetitive
scans.
Experimental Section
Collisionally activated dissociation (CAD) and metastable ion
(MI) spectra were measured on a JEOL HX-110 double-focusing
mass spectrometer of forward geometry (the electrostatic sector
E precedes the magnet B). Collisions with air were monitored
in the first field-free region at pressures resulting in 70% and
50% transmittance of the ion beam at 10 keV. The spectra were
obtained by scanning E and B simultaneously while maintaining
a constant B/E ratio (B/E-linked scan).
Materials. N-methylacetamide (2) (Aldrich), acetone (Fisher,
99.8%), dimethyl disulfide (DMDS, Aldrich), ammonia (Scott
Specialty Gases), and oxygen (Air Products) were used as
received. Deuterium-labeled reagents D2O and acetone-d6 (both
Cambridge Isotope Laboratories, 99.9% D) were also used as
received.
N-Methylacetamide-N-d (2a). N-methylacetamide (1.27 mL,
16.7 mmol) was stirred in 37.0 mL of D2O at room temperature
for 61 h, and the solvent was evaporated in vacuo.
N-Methylacetamide-2,2,2-d3 (2b). Neat (CD3CO)2O (5 g,
0.046 mol, Aldrich, 99% D) was added dropwise under stirring
to a 40% aqueous solution of methylamine (25 mL, 0.322 mol)
that was cooled to 0 °C. After addition, the solution was allowed
to warm to room temperature and stirred for 2 h. The solution
was saturated with ammonium sulfate and the product was
extracted in chloroform (5 × 20 mL). The chloroform extract
was dried over anhydrous MgSO4, the solvent was distilled off
through a 20 cm Vigreux column, and CD3CONHCH3 was
distilled, bp 110-115 °C/15 Torr. Yield: 2.7 g, 77%.
N-(Methyl-d3)acetamide (2c). Methyl-d3 ammonium hydro-
chloride (5 g, 67 mmol, Aldrich, 99% D) was added slowly at
0 °C to 35 mL of 4 M NaOH aqueous solution. After 10 min,
7 g (69 mmol) of neat acetic anhydride was added dropwise at
0 °C. After 1 h, the reaction mixture was worked up as above
and distilled at 15 Torr to give 3.6 g (71%) of CH3CONHCD3.
Methods. Measurements were performed on a tandem
quadrupole acceleration-deceleration mass spectrometer de-
scribed previously.8 Cation 1+ was generated in a tight chemical
ionization (CI) source via selective protonation at oxygen of 2
Calculations. Standard ab initio and density functional theory
calculations were performed using the Gaussian 98 suite of
programs.10 Geometries were initially optimized using Becke’s
hybrid functional (B3LYP)11 and the 6-31+G(d,p) basis set and
reoptimized at 6-311+G(2d,p). Spin-unrestricted calculations
(UB3LYP) were used for open-shell systems. In the UB3LYP
calculations, S2 operator expectation values ranged from 0.753
to 0.757 and 0.753 to 0.770 for local minima and transition
state structures, respectively, indicating negligible spin con-
tamination. Optimized structures were characterized by harmonic
frequency analysis as local minima (all frequencies real) or first-
order saddle points (one imaginary frequency). Zero-point
vibrational energies (ZPVE) were calculated from B3LYP/6-
311+G(2d,p) frequencies, which were scaled by 0.963 (ref 12).
The rigid-rotor harmonic oscillator approximation was used in
all thermochemical calculations. Single-point energies were
calculated at two levels of theory. Composite G2(MP2) ener-
gies13 were determined for all structures and transition states
from MP2/6-311+G(3df,2p) and quadratic configuration inter-
action calculations, QCISD(T)/6-311G(d,p).14 MP4(SDTQ)
energies were calculated using an expanding basis set (6-311G-
(d,p), 6-311+G(d,p), and 6-311G(2df,p)) and used to determine

Hydrogen Atom Adducts to the Amide Bond
J. Phys. Chem. A, Vol. 107, No. 1, 2003 117
SCHEME 3
G2 energies15 for several radicals, cations, and transition states
of interest. Spin contamination in the UMP2, UMP4(SDTQ),
and UQCISD(T) calculations was small for local minima, as
shown by spin expectation values S2 that ranged from 0.759
to 0.796. However, S2 values ranged from 0.798 to 0.967 for
transition states. Spin annihilation using Schlegel’s projection
method16 (PMP2)3 reduced the S2 values to e0.76 for all
structures.
Franck-Condon energies in vertical neutralization and reion-
ization were taken as absolute differences between the total G2-
(PMP2) energies of fully optimized ion or neutral structures
and those in which an electron has been added to an optimized
cation structure or subtracted from an optimized neutral
structure. No zero-point corrections were applied to the calcu-
lated Franck-Condon energies.
The labeled cations and radicals were generated by combining
selective protonation and deuteronation under H/D exchanging
or nonexchanging conditions. Thus, cation 1a+ was prepared
from 2 by selective deuteronation with (CD3)2COD+. This
reagent ion was generated by self-deuteronation of acetone-d6
which provides an essentially nonprototropic medium in the gas
phase, such that the N-proton in 2 does not undergo exchange
for D. Collisional reduction of 1a+ with DMDS produced radical
1a. Likewise, selective protonation of the carbonyl oxygen in
the N-D labeled compound 2b to form 1b+ was achieved by
reaction with (CH3)2COH+,22 followed by collisional reduction
of the cation to produce radical 1b. Radical 1c was generated
from amide 2b by deuteronation with (CD3)2COD+ forming 1c+,
followed by collisional reduction of the cation. Radicals 1d and
1e containing the label in nonexchangeable positions were
generated by NH4+/NH3 protonation of methyl-labeled N-
methylacetamides 2d and 2e, respectively, followed by colli-
sional electron transfer.
RRKM calculations used Hase’s program17 that was recom-
piled for MS-DOS and run under Windows NT.18 Vibrational
state densities were obtained by direct count of quantum states
in 2 kJ mol-1 steps for internal energies up to 205-310 kJ mol-1
above the threshold. The rotational states were treated adiabati-
cally,19 and the microscopic rate constants [k(E,J,K)] were
Boltzmann-averaged over the thermal distribution of rotational
J and K states pertaining to the ion source temperature.
+
In addition to selective protonation with NH4 and (CH3)2-
COH+, we also used stronger gas-phase acids, CH3OH2
+
(PA(CH3OH) ) 754 kJ mol-1),20,21 H3O+ (PA(H2O) ) 691 kJ
mol-1),20,21 and CH5+ (PA(CH4) ) 544 kJ mol-1 20,21
that should
)
be able to protonate 2 at N and O nonselectively and produce
mixtures of cation isomers. Collisional neutralization of cations
produced by nonselective protonation was also studied, as
discussed below.
Results and Discussion
Formation of 1-1e. Radicals 1-1e were generated in two
steps, as shown for 1 in Scheme 3. In the first step, N-
methylacetamide (2) was selectively protonated at the oxygen
atom with a gas-phase acid (BH+) to yield cation 1+. The
protonation selectivity was ensured by the choice of the gas-
phase acid and relied on the substantial difference in the topical
Precursor Ion Structures and Dissociations. Because the
structures of the incipient radicals 1-1e are determined by those
of the corresponding cations, we deemed it worthwhile to also
investigate the properties of the precursor cations. Moreover,
collisional reionization of neutral intermediates produces cations
whose dissociations need to be deconvoluted from those of the
radicals and molecules. Thus, some basic understanding of ion
structures, energies, and dissociation mechanisms is necessary.
proton affinities of the carbonyl oxygen (PA ) 893 kJ mol-1
)
and the nitrogen atom (PA ) 833 kJ mol-1) in 2 that were
obtained by G2 calculations. Thus, only the more basic carbonyl
oxygen can be protonated exothermically with a mild gas-phase
Four local minima were found for 1+, viz trans-anti-1+, cis-
anti-1+, trans-syn-1+, and cis-syn-1+ where cis and trans refer
to the configuration of the amide group and syn and anti refer
to the conformation of the OH group (Figure 1).23 Of the ion
isomers, trans-anti-1+ was the most stable conformer, followed
by cis-anti-1+, trans-syn-1+, and cis-syn-1+ (Table 2). However,
the small energy differences between trans-anti-1+ and cis-anti-
acid such as NH4 (PA(NH3) ) 853 kJ mol-1),20,21 whereas
+
endothermic protonation of the N-atom could not compete. In
the next step, ion 1+ was accelerated to 146 400 m s-1
corresponding to 8228 eV kinetic energy and discharged by
collisional electron transfer from gaseous dimethyl disulfide.
Because of the short duration of the electron-transfer collision,
estimated at <7 × 10-15 s from the collision kinematics, radical
1 is formed by vertical electron transfer with the bond
connectivity and geometry of the precursor cation.
1+ (∆H0 ) 9 kJ mol-1) and trans-syn-1+ (∆H0 ) 10 kJ mol-1
)
indicate that these two less stable conformers are also populated
in the gas phase at 473-523 K, corresponding to the range of

118 J. Phys. Chem. A, Vol. 107, No. 1, 2003
Syrstad et al.
Figure 1. B3LYP/6-311+G(2d,p)-optimized structures of trans-anti-, cis-anti-, trans-syn-, and cis-syn-1 and -1+. Bond lengths are in angstroms,
and bond and dihedral angles are in degrees.
TABLE 1: Collisionally Activated Dissociation Spectra of Ions 1+, 1a+-1e+ Relative Intensitya
m/z
1+
1a+
1b+
1c+
1d+
1e+
m/z
1+
1a+
1b+
1c+
1d+
1e+
76
75
74
73
72
62
61
60
59
58
57
56
55
54
53
52
51
50.5
48
47
46
45
44
43.5
43
42
41
4.3b,3
2.0
4.8b,3
2.3
40
39
38.5
38
37.5
37
36.5
36
35.5
35
34
33
32
31
30
29
28
27
26
25
1.2
1.6
0.8
1.6
1.4
1.3
0.9
1.9
1.2
1.7
1.3
2.9
23.7b,1
2.1b,3
0.7
0.6
27.0b,1
1.1
1.0
6.1b,1
1.3
0.0
1.5
5.6
5.9b,3
0.4
0.9
1.0
3.1
0.8
1.2
1.7
4.2
1.4
1.5
3.6
0.7
1.5
2.8
1.2
3.3b,4
1.2
1.1
1.3
1.2
5.7
0.7
0.5
0.8
1.6
1.3
1.7
1.0
1.4
1.2
2.6
0.8b,4
1.0
0.7
1.6
2.9
1.7
1.2
2.1b,4
1.4
1.5
1.2
2.8
3.2b,4
3.1
0.9
4.1b,2
2.2
2.8
5.9
2.0
2.6
4.3
2.6
1.4
1.3
0.7
3.6b,4
2.9b,6
4.4
3.6
2.1
3.9
2.3
1.5
0.9
5.1b,2
3.8
5.3
4.0
2.7
4.9
1.1
0.7
2.6b,4
0.7
2.7b,6
1.1
0.9
3.1b,3
3.0
4.3
4.1
2.0
4.9
1.4
1.2
4.8b,3
3.6
7.1
2.5
4.4
2.9
2.2
6.8b,2
4.3
9.4
1.5
5.5
1.5
0.8
0.5
1.1
1.1
0.6
0.5
0.8
0.7
1.3
-
1.2
0.7
1.2
1.2
1.4
1.3
1.0
1.2
3.8
2.4
2.7
1.0
2.7
3.4
2.0
3.8
2.7
25.1b,1
4.9
19
18
17
16
15
14
1.5
2.0
1.4
5.4
1.5
1.8
0.9
0.8
1.1
6.5
5.7b,5
1.3
27.5b,1
4.2
1.8
14.4b,2
3.2
1.9
20.2b,2
4.3
18.7b,2
3.1
1.5
1.8
1.1
0.8
18.0b,1
4.3
2.3
1.8
0.7
1.8
2.5
1.7
2.0
1.8
a Relative to the sum of CAD ion intensities. b Dominant products of metastable ion dissociations, relative abundance 1 > 2 > ...
ion source temperatures used in our experiments, and can
participate in ion dissociations. By comparison, cis-trans
isomers of 2 differ by 10-12 kJ mol-1 in favor of the trans
isomer.24
Metastable-ion (MI) and collisionally activated dissociations
(CAD) of 1+ and its labeled isotopomers 1a+-1e+ are sum-
marized in Table 1. Elimination of methylamine is the dominant
MI and CAD process yielding CH3CO+ at m/z 43 from 1+ and
CD3CO+ at m/z 46 from 1d+. The other important dissociations
are loss of H (m/z 73), CH4 (m/z 58), H2O (m/z 56), and ketene
+
(m/z 32) and formation of CH2dNH2 (m/z 30). Deuterium
labeling in 1c+ reveals that the loss of H involves mainly (92%)
hydrogen atoms from the methyl groups (Table 1). Hence, the
product cation-radicals correspond to a mixture of an N-

Hydrogen Atom Adducts to the Amide Bond
J. Phys. Chem. A, Vol. 107, No. 1, 2003 119
TABLE 2: Ion Relative Energies
relative energya
species/reaction
G2(MP2)b
G2b
exp. (298 K)c
trans-anti-1+
0
9
11
20
137
140
127
141
0
cis-anti-1+
9
10
20
trans-syn-1+
cis-syn-1+
trans-anti-1+ (VI)d
cis-anti-1+ (VI)d
trans-syn-1+ (VI)d
cis-syn-1+ (VI)d
trans-anti-1+ f trans-2 + H+
cis-anti-1+ f cis-2 + H+
trans-syn-1+ f trans-2 + H+
cis-syn-1+ f cis-2 + H+
CH3C)(O)NH2CH3+ f trans-2 + H+
trans-anti-1+ f CH3-NdC-CH3+ + H2O
trans-anti-1+ f CH3NH3- - -CH2dCdO+
887 (893)e
888 (893)e
876 (883)e
877 (883)e
827 (833)e
113
115
155
228
259
886 (893)e
887 (892)e
876 (883)e
876 (882)e
826 (832)e
889
trans-anti-1+ f CH2dCdO + CH3NH3
160f
240g
+
trans-anti-1+ f CH3NH2 + CH3-CdO+
trans-anti-1+ f HOdCdNH+ + C2H6
trans-anti-1+ f CH3C)(O)NH2+• + CH3
439
485
627
450
•
trans-anti-1+ f CH3C(OH)dNH+• + CH3
•
trans-anti-1+ f HOdCdNH+ + 2 CH3
•
a In units of kilojoules per mole at 0 K. b From spin-projected MP2 energies wherever it applies. c Experimental reaction enthalpies are based on
thermochemical data from ref 21 unless stated otherwise. d Energy difference in kilojoules per mole between the vertically ionized cation and the
optimized cation structure. e Topical proton affinities of 2 at 298 K. f Based on ∆Hf,298 (CH2dCdO) ) -54 kJ mol-1, see ref 31. g Based on
∆Hf,298 (CH3CdO+) ) 657 kJ mol-1, see ref 32.
methylacetamide enol, CH2dC(OH)sNHCH3+•, and the dis-
•
tonic ion, CH3C+(OH)-NH-CH2 . Elimination of methane
selectively involves the acetyl methyl group, but the hydrogen
atom is transferred nonselectively from the NH and OH groups.
Elimination of water proceeds by specific transfer of the NH
proton onto the hydroxyl group in 1+.
The energetics of the dissociations observed in the MI and
CAD spectra were obtained by G2(MP2) calculations as
summarized in Table 2. Elimination of water is the lowest-
energy dissociation which is 113 kJ mol-1 endothermic at the
0 K thermochemical threshold. Elimination of ketene forming
CH3NH3+ was the second most favorable dissociation that was
calculated to require a threshold energy of ∆Hrxn,0 ) 155 kJ
mol-1 (Table 2). Note that the elimination of ketene necessitates
a rearrangement by transfer of two protons and may involve a
hydrogen-bonded ion-molecule complex, OdCdCH2‚‚‚
H+‚‚‚H2N-CH3, as an intermediate which is 115 kJ mol-1
higher in energy than trans-anti-1+ (Table 2). Elimination of
methylamine was another low-energy dissociation of 228 kJ
mol-1 threshold energy. In contrast, losses of methyl by
dissociations of the C-C or N-C bonds are calculated to have
>400 kJ mol-1 threshold energies (Table 2). It should be noted
that the low-energy dissociations of 1+ involve rearrangements
by hydrogen transfer that are likely to have activation barriers
that may exceed the threshold energies.
Radical Dissociations. The stability and dissociations of
radicals 1-1e were studied by NRMS. Collisional neutralization
of 1+ produced a small fraction of stable radicals (0.5% of the
sum of NR ion intensities), which were detected following
reionization as survivor ions at m/z 74 in the NR mass spectrum
(Figure 2a). The primary unimolecular dissociations observed
upon NR were loss of H• (m/z 73) and loss of CH3 (m/z 59).
•
Other dissociations proceeded in a consecutive fashion through
the [C3,N,O,H7] and [C2,H4,N,O] molecules formed by the
primary dissociations, as discussed below.
The loss of H from neutral 1 was distinguished from the
analogous dissociation of reionized 1+ by comparing the CAD
and NR mass spectra of the isotopomers. NR of the OD-labeled
Figure 2. Neutralization (CH3SSCH3, 70% transmittance)-reionization
(O2, 70% transmittance) mass spectra of (a) 1+ and (b) 2+•. (c)
Neutralization (CH3SSCH3, 70% transmittance)-collisional activation
(He, 50% transmittance)-reionization (O2, 70% transmittance) mass
spectrum of 2+. Inset shows the difference spectrum from subtracting
from (a) a weighed average of (b) and (c).

120 J. Phys. Chem. A, Vol. 107, No. 1, 2003
Syrstad et al.
Figure 4. Neutralization (CH3SSCH3, 70% transmittance)-reionization
(O2, 70% transmittance) mass spectra of (a) 1d+ and (b) 1e+.
70-eV electron-ionization mass spectra that show identical
major fragments at similar relative intensities.
A product in the NR spectrum of 1 that cannot be accounted
for by consecutive 2 dissociations appears at m/z 44 (Figure
2a). Deuterium labeling shows unambiguously that this fragment
is formed by loss of both methyl groups in 1 and corresponds
to HOsCdNH+. Comparison of the NR and CAD spectra of
1+-1e+ further shows that the combined loss of two methyl
groups is not due to contributions from dissociations of reionized
1+-1e+. For example, the peak at m/z 44 shifts almost
completely to m/z 45 in the NR spectra of 1a+ and 1b+, whereas
in the corresponding CAD spectra there are substantial peaks
remaining at m/z 44 (Table 1). An intermediate for the loss of
the first methyl appears as a very weak peak at m/z 59 that
shows the expected mass shift to m/z 60 in 1a and 1b because
of the presence of OD and ND, respectively, and to m/z 61 in
1c. Deuterium labeling of the methyl groups shows loss of the
N-CH3 methyl (m/z 62 in 1d) but negligible loss of the acetyl
methyl from 1e, where the peak at m/z 62 is within the noise
level. This points to an acetimino enol structure for the primary
dissociation product at m/z 59, CH3C(OH)dNH (6).
O-H/N-CH3 Branching Ratio. To quantify the amounts
of 2 and 6 formed by dissociation of 1, we used the reference
NR and NCR spectra of 2 that were weighted, mixed, and
subtracted from the NR spectrum of 1 such as to annihilate the
survivor peak of 2 at m/z 73. The difference spectrum (Figure
2c, inset) showed virtually no negative peaks and was assigned
to represent reionized 6. Note also that the difference spectrum
shows no illogical losses that would be incompatible with
structure 6. The total NR ion currents in the weighted spectra
of 2 and the difference spectrum for 6 were further scaled by
the corresponding ratio of ionization cross sections σ(6)/σ(2)
) 0.78 that was estimated from the Fitch-Sauter additivity
scheme.25 This yielded a 2/6 branching ratio of 1.7.
Figure 3. Neutralization (CH3SSCH3, 70% transmittance)-reionization
(O2, 70% transmittance) mass spectra of (a) 1a+, (b) 1b+, and (c) 1c+.
ion 1a+ showed 45% loss of H and 55% loss of D (Figure 3a,
inset), whereas CAD of 1a+ showed predominant (96%) loss
of H (Table 1). Similarly, NRMS of N-CD3 labeled 1e+ showed
no measurable loss of D (Figure 4b), whereas CAD of 1e+
showed 33% loss of D and 67% loss of H. The combined
evidence from all our labeling experiments pointed to loss of
the H atom primarily from the OH group and to a lesser extent
from the acetyl CH3 group in dissociations of radical 1. Hence,
the main product of 1 dissociations is N-methylacetamide (2).
The origin of the other fragments in the NR mass spectrum
of 1 was investigated in light of the formation of 2 as the primary
dissociation product. To this end, we obtained the NR mass
spectrum of 2+• (Figure 2b) that contained most of the
dissociation products also present in the NR spectrum of 1. The
similarity becomes even greater upon collisional activation of
neutral 2, followed by collisional reionization (NCR spectrum,
Figure 2c), that yielded products in the m/z 15-73 mass region
that had very similar relative intensities to those in the NR
spectrum of 1 (Figure 2a). Because 2 is a stable molecule in
the gas-phase, it is safe to conclude that the dissociations
observed in its NR mass spectrum occur following collisional
reionization and not in the stage of the neutral molecule. This
conclusion is supported by the similarity of the NR and standard

Hydrogen Atom Adducts to the Amide Bond
J. Phys. Chem. A, Vol. 107, No. 1, 2003 121
In summarizing NR of 1, loss of H from the OH group and
loss of the N-methyl are the principal dissociations of radical 1
that occur in a 1.7:1 ratio.
Radical Structures and Energetics. Four local energy
minima for radical 1 conformers were initially found by high-
level ab initio calculations (Figure 1), e.g., trans-anti-1, cis-
anti-1, trans-syn-1, and cis-syn-1 that were analogous to the
corresponding ion structures. The energy differences among the
radical conformers were extremely small, as cis-anti-1, trans-
syn-1, and cis-syn-1 were only 1.9, 3.1, and 3.4 kJ mol-1 less
stable than trans-anti-1, respectively. Interconversion of the
radical isomers occurred through a combination of C2-O3 and
C2-N4 bond rotations. The B3LYP/6-311+G(2d,p) potential
energy surface (PES) was investigated for C2-O3 rotation in
the most favored trans-anti-1 radical, which had a dihedral angle
N4-C2-O3-H11 ) 166.1°. Through C2-O3 rotation, two
additional local minima were found. These rotamers had G2-
(MP2) energies of 1.5 and 3.2 kJ mol-1, corresponding to
dihedral angles N4-C2-O3-H11 ) -26.3 and +60.0, respec-
tively. Energy barriers of 3.0, 2.8, and 3.7 kJ mol-1, at dihedral
angles N4-C2-O3-H11 ) -127.1, +25.0, and +98.2°, respec-
tively, connected these radical rotamers to each other and to
trans-anti-1. Interestingly, the PES for C2-O3 rotation did not
include the trans-syn-1 rotamer, as the pyramidal arrangement
of C1, O3, and N4 around C2 was inverted between this radical
and trans-anti-1, and this conformation was retained throughout
the bond rotation.
Similar analysis of the PES for C2-N4 rotation from trans-
anti-1 found yet two more local minima, of dihedral angles C1-
C2-N4-C5 ) -43.6 and +55.9°, at 9.0 and 2.0 kJ mol-1
relative to trans-anti-1, respectively. Transition states connecting
these rotamers, at dihedral angles C1-C2-N4-C5 ) -113.3
and +88.7°, had relative energies of 17.8 and 23.5 kJ mol-1
,
respectively. A true transition state (one imaginary frequency)
was not found to connect the two above-mentioned local
minima, because of a discontinuity in the B3LYP potential
energy surface between a dihedral angle of 15 and 25°.
However, this transition state was estimated to be, at most, 10-
12 kJ mol-1 relative to trans-anti-1, so that rotamer intercon-
version would not be slowed at this energy barrier relative to
the other two. Potential energy surfaces for C2-O3 and C2-N4
rotations starting from the other three known local minima were
not determined, but it was clear that additional local minima
likely existed. Franck-Condon energies were calculated by G2
for the vertical neutralization of 1+ to 1 as 93 kJ mol-1 for
Figure 5. Neutralization (CH3SSCH3, 70% transmittance)-reionization
(O2, 70% transmittance) mass spectra of 1+ following protonation with
(a, top) CH3OH2+, (b, middle) H3O+, and (c, bottom) CH5
.
+
Note that, because 6 is an unstable tautomer of acetamide,
the corresponding cation-radical cannot be generated by direct
ionization. Attempts were made to generate the reference ion
6+• by alkene elimination from N-alkyl acetamide precursors,
but dissociative ionization produced only low yields of 2+•.
Radicals from High-Energy Protonation of 2. Radicals
formed by neutralization of cations produced by highly exo-
thermic protonation with CH3OH2+, H3O+, and CH5+ gave NR
spectra that were similar to those shown in Figure 2. In
particular, the spectra showed low-intensity survivor ions at m/z
74 and fragment ion relative intensities that pointed to dominant
loss of the hydroxyl hydrogen and formation of 2 as a neutral
intermediate (Figure 5). Interestingly, the ion relative intensities
resembled those in the NR spectrum of 2+• without collisional
activation and indicated that the intermediate 2 was less excited.
This may be due to a more efficient cooling of 1+ in the ion
source by collisions with CH3OH, H2O, and CH4 than in the
case of the acetone gas. The NR spectra in Figure 5 did not
show an increased relative abundance of CH3NH2 (m/z 31) and
its dissociation products that would indicate formation of the
trans-anti-1+ f trans-anti-1, 98 kJ mol-1 for cis-anti-1+
f
cis-anti-1, 99 kJ mol-1 for trans-syn-1+ f trans-syn-1, and 103
kJ mol-1 for cis-syn-1+ f cis-syn-1. This pointed to substantial
vibrational excitation in the ground electronic state (2X) of 1
formed by vertical electron transfer.
Radical Dissociation Energetics. Transition state geometries
for the loss of H from both O and N, as well as for loss of CH3
from N, in 1 were determined at the UB3LYP/6-311+G(2d,p)
level of theory (Figure 7). Potential energy surfaces along the
dissociation coordinates were investigated for reaction pathways
for O-H, N-H, and N-CH3 cleavages by a PES-mapping
method described previously.26 Briefly, the appropriate bonds
were stretched by 0.02 Å increments, whereas the remaining
internal degrees of freedom were fully optimized, and each point
near the transition state was treated with G2(MP2) single-point
calculations, providing corrected PES profiles. Corrected saddle
points were determined from polynomial fitting of these profiles
and were used for G2 single-point energy calculations. These
showed shortening of the transition state bond lengths from the
•
unstable CH3CO-NH2 -CH3 radical by neutralization of N-
•
protonated acetamide. CH3CO-NH2 -CH3 was calculated to
dissociate spontaneously to CH3CO• and methylamine.

122 J. Phys. Chem. A, Vol. 107, No. 1, 2003
Syrstad et al.
Figure 6. B3LYP/6-311+G(2d,p)-optimized structures of TS1-TS9.
UB3LYP-optimized saddle points, especially for N-H bond
dissociation (Table 4). Dissociations of the O;H bonds
encountered energy barriers that were identical for pairs of
starting radicals having the same conformation of methyl groups
around the amide bond. Starting from the syn- or anti- methyl
conformations (TS1 and TS2; 84 and 90 kJ mol-1, respectively),
these mirror-image, nearly planar saddle point geometries
differed only in the orientation of the O-H bond being out of
plane in either direction. Thus, trans-anti- and trans-syn-1, and
cis-anti- and cis-syn-1, converged to energetically, but not
structurally, identical transition states. Accordingly, the four (or
more) starting isomers of 1 produced just two unique dissocia-
tion products (cis- and trans-2 and H) from dissociation of O-H.
However, N-H cleavage resulted in four distinct dissociation
products, requiring a separate transition state connecting to each
summarized in the potential energy diagram shown in Figure
7.
Five other [C3,H8,N,O] isomers were found to be local energy
minima (Figure 8). The thermodynamically favored 2-hydroxy-
propane-2-aminyl isomers syn- and anti-3 represented potential
intermediates for methyl scrambling in radical 1, but this process
was associated with a substantial energy barrier of 214 kJ mol-1
(TS9, Figure 6). This indicated that methyl loss from 1 should
proceed unambiguously, with no methyl scrambling. The
1-hydroxyethane-1-(N-methyl)aminyl isomers syn- and anti-4
and the 1-(N-methyl)aminoethoxyl isomer 5 were found by G2
to be 7, 14, and 53 kJ mol-1 less stable than 1, respectively
(Table 3). As in the simpler amino(hydroxy)methyl system,26
these radical stabilities follow the relative order of typical bond
dissociation energies, BDE(O-H) > BDE(N-H) > BDE(Csp3
-
H).27
starting rotamer (TS5-TS8; 118, 132, 116, and 109 kJ mol-1
,
respectively). In addition, transition states were found for
N-CH3 dissociation leading to the two (presumably) most
favored products: trans-syn-1 f TS4 (84 kJ mol-1), and cis-
anti-1 f TS3 (93 kJ mol-1). All pertinent relative energies are
The PES calculations pointed to the following conclusions.
The conformers of 1 are nearly isoenergetic and can interconvert
at internal energies that are way below the lowest dissociation
threshold. Thus, any of the conformers can be available as a

Hydrogen Atom Adducts to the Amide Bond
J. Phys. Chem. A, Vol. 107, No. 1, 2003 123
Figure 7. Potential energy diagram for dissociations of 1.
reactant for a particular dissociation pathway. The transition
states for loss of the O-H hydrogen atom and the N-CH3
methyl group have very similar energies when starting from
the same syn- or anti-O-H conformers. However, dissociations
starting from the trans-amide conformers show slightly but
significantly lower TS energies for loss of H and CH3 than those
starting from the cis conformers.
result is to be expected from RRKM theory for competitive
dissociations through nearly isoenergetic but nonequivalent
transition states. In this special case, the ratio of rate constants
is given by eq 1
k(O‚‚‚H)/k(N‚‚‚CH3) ≈ W‡(O‚‚‚H)/W‡(N‚‚‚CH3) (1)
Dissociation Kinetics. To quantify the product formation by
radical dissociations, we used the G2 energy barriers and
UB3LYP/6-311+G(2d,p) harmonic frequencies and moments
of inertia to calculate RRKM rate constants. The results of these
calculations are shown in Figure 9. This figure provides a direct
comparison of rate constants for competitive O-H, N-H, and
N-CH3 dissociations, with each calculated from the starting
radical corresponding to the lowest energy barrier (see Figure
7). As discussed above, this comparison relied on the assumption
that rotamer interconversion occurred much faster than any of
the bond dissociations, producing an equilibrium mixture of
starting radicals. The largest calculated barrier to isomerization
was ∼24 kJ mol-1, substantially lower than the most favored
bond cleavage. The calculations indicated that, for the range of
internal energies investigated (dissociation threshold to 390 kJ
mol-1), loss of H from 1 should occur predominately from the
OH group. Unimolecular rate constants for O-H and N-H bond
dissociations differed from each other by 2 orders of magnitude
at about 150 kJ mol-1 internal energy and converged to 1 order
of magnitude at 265 kJ mol-1. However, O-H and N-CH3
bond dissociations were predicted to be very competitive at all
energies, as expected by their identical energy barriers. In
contrast to the previous report of Zubarev et al.,5b our calcula-
tions indicate that the dissociation of the O-H bond becomes
faster at very high excitations of the radical (Figure 10). This
where W‡(O‚‚‚H) and W‡(N‚‚‚CH3) are the quantum state counts
in the transition states for the corresponding bond dissociations.28
Hence, at sufficiently high internal energies where the nuances
in quantum state combinations and small differences in the
potential energy barriers become irrelevant, the different
vibrational frequencies for the stretching modes, ν(O;H) >
ν(N;CH3), cause that the transition state for the O‚‚‚H bond
cleavage that preserves the softer N-CH3 stretching vibration
can be realized in a greater number of quantum states and will
allow faster dissociation, as shown in Figure 10.
Dissociation Mechanisms Relevant to NRMS of 1. The
competitive dissociations of 1 by loss of H and CH3 that are
triggered by femtosecond electron transfer depend on the internal
energy imparted in the radical. The internal energy of vertically
formed 1 contains contributions from the internal energy of the
precursor ion (Eion), the Franck-Condon energy gained upon
electron transfer (EFC), and the excitation energy of the particular
electronic state (∆Eexc). We have shown previously29 that the
most probable internal energy Eint of radicals produced by NR
in the ground electronic state (∆Eexc ) 0) can be expressed as
a simple sum of two terms, e.g., Eint ) Eion + EFC. Of these,
EFC in 1 was obtained by G2 calculations as 93 kJ mol-1. Eion
was estimated as 95-99 kJ mol-1 from the thermal energy of
2 at the typical ion source temperature (30-34 kJ mol-1 at 460-
500 K) and an ≈80% fraction of protonation exothermicity (65

124 J. Phys. Chem. A, Vol. 107, No. 1, 2003
Syrstad et al.
TABLE 3: Radical Relative Energies
relative
energya
species/reaction
G2(MP2)b G2b
trans-anti-1
cis-anti-1
0
2
0
2
trans-syn-1
3
3
93
96
3
3
cis-syn-1
trans-anti-1 (VN)c
93
96
96
100
-16
-9
7
cis-anti-1 (VN)c
trans-syn-1 (VN)c
96
cis-syn-1 (VN)c
100
-16
-11
7
syn-3
anti-3
syn-4
anti-4
14
54
31
39
28
37
95
111
88
14
53
31
39
28
37
95
111
88
84
146
135
-
-
22
7
172
160
-
70
84
90
93
84
118
132
116
109
214
5
trans-anti-1 f trans-2 + H•
cis-anti-1 f cis-2 + H•
trans-syn-1 f trans-2 + H•
cis-syn-1 f cis-2 + H•
trans-anti-1 f trans-anti-CH3C(OH)dNCH3 + H•
cis-anti-1 f cis-anti-CH3C(OH)dNCH3 + H•
trans-syn-1 f trans-syn-CH3C(OH)dNCH3 + H•
cis-syn-1 f cis-syn-CH3C(OH)dNCH3 + H•
trans-anti-1 f trans-anti-CH2dC(OH)NHCH3 + H•
cis-anti-1 f cis-anti-CH2dC(OH)NHCH3 + H•
trans-anti-1 f trans-anti-CH3C(OH)NHCH2 + H•
cis-anti-1 f cis-anti-CH3C(OH)NHCH2 + H•
84
146
135
285
284
23
•
cis-anti-1 f anti-CH3C(OH)dNH + CH3
•
trans-syn-1 f syn-CH3C(OH)dNH + CH3
7
•
trans-anti-1 f C(OH)NHCH3 + CH3
172
160
439
71
86
93
95
85
119
133
117
109
214
•
cis-anti-1 f C(OH)NHCH3 + CH3
•
trans-anti-1 f C(OH)dNH•+ 2 CH3
trans-anti-1 f C(OH)dNH•+ C2H6
TS1
TS2
TS3
TS4
TS5
TS6
TS7
TS8
TS9
a In units of kilojoules per mole at 0 K. b From spin-projected MP2
energies wherever it applies. c Energy difference in kilojoules per mole
between the vertically neutralized radical and the optimized radical
structures.
Figure 8. B3LYP/6-311+G(2d,p)-optimized structures of syn- and
anti-3, -4, and 5.
and N-CH3 bonds, which all require lower energies in the
corresponding transition states.
TABLE 4: Transition State Bond Lengths
d(X-Y)a,b
Regarding competitive dissociations of O-H and N-H bonds
in 1, RRKM rate constants obtained for Eint ) 190 kJ mol-1
predict branching ratios to be k(O;H)/k(N;H) ≈ 25 (Figure
9), in qualitative agreement with experimental observations. In
addition, branching ratios k(X-D)/k(X-H) (X ) O and N) were
determined experimentally and computationally for loss of (H,D)
from the most favored starting isomer in O-D labeled 1a+ and
N-D labeled 1b+. NR of 1a+ shows a branching ratio
k(O-D)/k(N-H) ) 1.21, compared to a value of 13.0 predicted
from RRKM data (Figure 9). In 1b+, the NR spectrum indicates
a branching ratio k(N-D)/k(O-H) ) 0, compared to a value
of 0.013 predicted by RRKM calculations. Clearly, however,
the peaks for O-H and N-H dissociation products in the NR
spectra contained possible contributions from ionic C-H bond
cleavage. In addition, consecutive dissociations, mostly follow-
ing reionization, likely depleted the [M-D] and [M-H] NR
signals disproportionately, somewhat compromising quantitative
comparisons.
species
B3LYPc
G2(MP2)d
TS1
TS2
TS3
TS4
TS5
TS6
TS7
TS8
1.424
1.442
2.018
1.993
1.700
1.780
1.680
1.670
1.383
1.392
1.990
1.980
1.580
1.629
1.560
1.587
a In units of angstroms. b X ) O, N; Y ) H, CH3. c UB3LYP/6-
311+G(2d,p) saddle point. d G2(MP2) corrected saddle point; from
spin-projected MP2 energies.
kJ mol-1).30 These energy contributions were calculated for
trans-anti-1, as the precursor cation equilibrium should favor
this isomer, similar to the preference that has been determined
for protonated formamide cations.26 From the energy terms for
Eion and EFC one obtains Eint ) 188-192 kJ mol-1 in trans-
anti-1. This is sufficient to drive dissociation of the O-H, N-H,
Regarding competitive dissociations of the O-H and N-CH3
bonds, RRKM calculations predict the energy-dependent branch-

Hydrogen Atom Adducts to the Amide Bond
J. Phys. Chem. A, Vol. 107, No. 1, 2003 125
for neutral dissociation (4.1 µs). P(E) is given by eq 3,29 where
the parameters E0 and W were set to 73 and 117 kJ mol-1 to
give the mean internal energy E ) 190 kJ mol-1 (vide supra)
4(E - E0)
e-2(E - E )/W
(3)
0
P(E) )
W2
while allowing for 0.5% relative abundance of undissociated 1
to reproduce the NR spectrum (cf. Figure 2a). The experimental
ratio [2]/[6] ) 1.7 is matched by lowering the G2 value for
Ea,H from 81.2 to 78.7 kJ mol-1. Note that a 2.5 kJ mol-1
adjustment is well within the accuracy of G2 energy calculations
(4 kJ mol-1).13,15 Hence, theory and experiment are in near
perfect quantitative agreement.
Conclusions
The 1-hydroxy-1-(N-methyl)aminoethyl radical is both ther-
modynamically and kinetically stable in the gas phase. When
formed by fast electron transfer to O-protonated acetamide, the
radical acquires ca. 100 kJ mol-1 internal energy through
Franck-Condon effects that drive dissociations of the O-H
and N-CH3 bonds. The experimental branching ratio of 1.7:1
that prefers the O-H bond dissociation is reproduced by RRKM
calculations of the unimolecular rate constants within the
accuracy of the G2 potential energy surface. This study shows
that radicals derived from N-methylacetamide do not represent
the best model for electron capture dissociation, where N-C
bond dissociations leading to backbone fragmentation predomi-
nate. More sophisticated models that mimic the electronic
properties of peptide radicals and dissociation products are
needed for elucidating ECD mechanisms.
Figure 9. RRKM unimolecular rate constants (log, s-1) for dissocia-
tions of 1.
Acknowledgment. Support of this work by NSF (CHE-009-
0930) is gratefully acknowledged. Support for the Computational
Chemistry Facility was provided jointly by NSF-CRIF (CHE-
9808182) and the University of Washington. We thank Dr.
Martin Sad´ılek for assistance with CAD spectra measurements.
References and Notes
(1) Hawkins, C. L.; Davies, M. J. Biochim. Biophys. Acta 2001, 1504,
196-219.
(2) (a) Berlett, B.; Stadtman, E. R. J. Biol. Chem. 1997, 272, 20313-
20316. (b) Schoneich, C. Exp. Gerontol. 1999, 34, 19-34.
(3) (a) Wisniowski, P.; Carmichael, I.; Fessenden, R. W.; Hug, G. L.
J. Phys. Chem. A 2002, 106, 4573-4580. (b) Meisel, D. M.; Camaioni,
D.; Orlando, T. M. In ACS Symposium Series; American Chemical
Society: Washington, DC, 2001; Vol. 778, pp 342-361.
(4) (a) Gatlin, C. L.; Turecˇek, F.; Vaisar, T. J. Am. Chem. Soc. 1995,
117, 3637-3638. (b) Vaisar, T.; Gatlin, C. L.; Turecˇek, F. J. Am. Chem.
Soc. 1996, 118, 5314-5315. (c) Chu, I. K.; Rodriquez, C. F.; Lau, T.-C.;
Hopkinson, A. C.; Siu, K. W. M. J. Phys. Chem. B 2000, 104, 3393-
3397. (d) Wee, S.; O’Hair, R. A. J.; McFadyen, W. D. Rapid Commun.
Mass Spectrom. 2002, 16, 884-890.
(5) (a) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem.
Soc. 1998, 120, 3265-3266. (b) Zubarev, R. A.; Kruger, N. A.; Fridriksson,
E. K.; Lewis, M. A.; Horn, D. M.; Carpenter, B. K.; McLafferty, F. W. J.
Am. Chem. Soc. 1999, 121, 2857-2862. (c) Cooper, H. J.; Hudgins, R. R.;
Hakansson, K.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 2002, 13,
241-249.
(6) For peptide fragment nomenclature see (a) Roepstorff, P.; Fohlman,
J. Biomed. Mass Spectrom. 1984, 11, 601. (b) Biemann, K. Biomed.
EnVironm. Mass Spectrom. 1988, 16, 99.
(7) For comprehensive reviews of the technique, see: (a) Wesdemiotis,
C.; McLafferty, F. W. Chem. ReV. 1987, 87, 485-500. (b) Holmes, J. L.
Mass Spectrom. ReV. 1989, 8, 513-539. (c) Schalley, C. A.; Hornung, G.;
Schroder, D.; Schwarz, H. Chem. Soc. ReV. 1998, 27, 91-104. (d)
Zagorevskii, D. V.; Holmes, J. L. Mass Spectrom. ReV. 1999, 18, 87-118.
(8) Turecˇek, F.; Gu, M.; Shaffer, S. A. J. Am. Soc. Mass Spectrom.
1992, 3, 493-501.
(9) Turecˇek, F. Org. Mass Spectrom. 1992, 27, 1087-1097.
(10) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
Figure 10. Energy dependence of the branching ratio for the
dissociations of the O-H and N-CH3 bonds in 1 forming 2 and 6,
respectively.
ing ratios that are depicted in Figure 10 for the G2-calculated
TS energies. At near-threshold energies (84 kJ mol-1), loss of
CH3 is about twice as fast as loss of H. However, the curves
cross at Eint ) 97 kJ mol-1 and the k(O-H)/k(N-CH3)
branching ratio increases to 1.2 at Eint ) 190 kJ mol-1 and
converges to 1.5 at Eint g 400 kJ mol-1. Full quantitative
agreement with the branching ratio from NR dissociations can
be achieved by using eq 2 where P(E) is an energy distribution
function, Ea,H and Ea,CH3 are the pertinent TS energies, kH and
kCH3 are the RRKM unimolecular rate constants, and τ is the
time
∞
τ dt
kH(E)P(E)e-(k +k )t dE
H
CH3
∫ ∫
[2]
[6]
0
Ea,H
)
(2)
τ dt
kCH3(E)P(E)e-(k +k )t dE
∞
H
CH3
∫ ∫
0
Ea,CH3

126 J. Phys. Chem. A, Vol. 107, No. 1, 2003
Syrstad et al.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.6; Gaussian,
Inc.: Pittsburgh, PA, 1998.
(21) http://webbook.nist.gov/chemistry
(22) Although the proton affinity of acetone (812 kJ mol-1) makes
(CH3)2COH+ sufficiently acidic to protonate the amide nitrogen in 2 of PA
) 833 kJ mol-1, this protonation cannot compete kinetically with that of
the CdO group where the thermochemistry predicts it to occur at the
collision rate limit. For the relationship between proton-transfer kinetics
and thermochemistry, see Bouchoux, G.; Salpin, J. Y.; Leblanc, D. Int. J.
Mass Spectrom. Ion Processes 1996, 153, 37-48.
(23) Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic
Compounds; John Wiley and Sons: New York; 1994; p 23.
(24) Ma´rtinez, A. G.; Vilar, E. T.; Fraile, A. G.; Martinez-Ruiz, P. J.
Phys. Chem. A 2002, 106, 4942-4950 and references therein.
(25) Fitch, W. L.; Sauter, A. D. Anal. Chem. 1983, 55, 832-835.
(26) Syrstad, E. A.; Turecˇek, F. J. Phys. Chem. A, 2001, 105, 11144-
11155.
(27) McMillen, D. F.; Golden, D. M. Annu. ReV. Phys. Chem. 1982,
33, 493-532.
(28) Gilbert, R. G.; Smith, S. C. Theory of Unimolecular and Recom-
bination Reactions; Blackwell: Oxford, 1990; p 54.
(29) (a) Wolken, J. K.; Turecˇek, F. J. Phys. Chem. A, 1999, 103, 6268-
6281. (b) Wolken, J. K.; Turecˇek, F. J. Am. Chem. Soc. 1999. 121, 6010-
6018. (c) Gerbaux, P.; Turecˇek, F. J. Phys. Chem. A, 2002, 106, 5938-
5950.
(30) Uggerud, E. AdV. Mass Spectrom. 1995, 13, 53-70.
(31) Traeger, J. C. Int. J. Mass Spectrom. 2000, 194, 261-267.
(32) Traeger, J. C.; McLoughlin, R. G.; Nicholson, A. J. C. J. Am. Chem.
Soc. 1982, 104, 5318-5322.
(11) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 1372-1377. (b) Stephens,
P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994,
98, 11623-11627.
(12) Rauhut, G.; Pulay, P. J. Phys. Chem. 1995, 99, 3093-3100.
(13) Curtiss, L. A.; Raghavachari, K.; Pople, J. A. J. Chem. Phys. 1993,
98, 1293-1298.
(14) Pople, J. A.; Head-Gordon, M.; Raghavachari, K. J. Chem. Phys.
1987, 87, 5968-5975.
(15) Curtiss, L. A.; Raghavachari, K.; Trucks, G. W.; Pople, J. A. J.
Chem. Phys. 1991, 94, 7221-7230.
(16) (a) Schlegel, H. B. J. Chem. Phys. 1986, 84, 4530-4534. (b) Mayer,
I. AdV. Quantum Chem. 1980, 12, 189-262.
(17) Zhu, L.; Hase, W. L. Quantum Chemistry Program Exchange;
Indiana University: Bloomington, IN, 1994; Program No. QCPE 644.
(18) Frank, A. J.; Sadilek, M.; Ferrier, J. G.; Turecˇek, F. J. Am. Chem.
Soc. 1997, 119, 12343-12353.
(19) Zhu, L.; Hase, W. L. Chem. Phys. Lett. 1990, 175, 117-124.
(20) Hunter, E. P.; Lias, S. G. J. Phys. Chem. Ref. Data, 1998, 27,
413-656.