Reaction of Distonic Aryl and Alkyl Radical Cations with Amines: The Role of Charge and Spin Revealed by Mass Spectrometry, Kinetic Studies, and DFT Calculations

Article abstract of DOI:10.1002/cplu.201900706

Gas-phase reaction of the aromatic distonic radical cations 4-Pyr^+. and 3-Pyr^+. with amines led to formation of the corresponding amino pyridinium ions 4-Pyr⁺NR₂ and 3-Pyr⁺NR₂ through amine addition at the site of the radical, followed by homolytic β-fragmentation. The reaction efficiencies range from 66–100 % for 4-Pyr^+. and 57–86 % for 3-Pyr^+., respectively, indicating practically collision-controlled processes in some cases. Computational studies revealed that the combination of positive charge and spin makes nucleophilic attack by the amine at the site of the radical barrierless and strongly exothermic by about 175±15 kJ mol^?1, thereby rendering ‘conventional’ radical pathways through hydrogen abstraction or addition onto π systems less important. Exemplary studies with 4-Pyr^+. showed that this reaction can be reproduced in solution. A similar addition/radical fragmentation sequence occurs also in the gas-phase reaction of amines with the aliphatic distonic radical cation Oxo⁺C^., showing that the observed charge-spin synergism is not limited to aromatic systems.

Full text of DOI:10.1002/cplu.201900706

Accepted Article
Title: Reaction of Distonic Aryl and Alkyl Radical Cations with Amines:
The Role of Charge and Spin Revealed by Mass Spectrometry,
Kinetic Studies and DFT Calculations
Authors: Benjamin Andrikopoulos, Parvinder Sidhu, Bethany Taggert,
Joses Nathanael, Richard O'Hair, and Uta Wille
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemPlusChem 10.1002/cplu.201900706
Link to VoR: http://dx.doi.org/10.1002/cplu.201900706
A Journal of
www.chempluschem.org

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
Reaction of Distonic Aryl and Alkyl Radical Cations with Amines:
The Role of Charge and Spin Revealed by Mass Spectrometry,
Kinetic Studies and DFT Calculations
Benjamin Andrikopoulos, Parvinder K. Sidhu, Bethany I. Taggert, Joses G. Nathanael, Richard A. J.
O’Hair,* Uta Wille*[a]
Dedicated to Prof. Frances Separovic on the occasion of her 65th birthday and her retirement.
Abstract: Gas-phase reaction of the aromatic distonic radical cations
4-Pyr+• and 3-Pyr+• with amines led to formation of the corresponding
amino pyridinium ions 4-Pyr+NR2 and 3-Pyr+NR2 through amine
addition at the site of the radical, followed by homolytic b-
fragmentation. The reaction efficiencies range from 66-100% for 4-
Pyr+• and 57-86% for 3-Pyr+•, respectively, indicating practically
collision-controlled processes in some cases. Computational studies
revealed that the combination of positive charge and spin makes
nucleophilic attack by the amine at the site of the radical barrierless
and strongly exothermic by about 175 ± 15 kJ mol-1, thereby rendering
‘conventional’ radical pathways through hydrogen abstraction or
addition onto p systems less important. Exemplary studies with 4-
followed by dissociation to form a s radical intermediate that is
subsequently attacked by an N-, O- or C-nucleophile, such as
sodium amide, an alkoxide or an enolate, respectively. During this
addition the unpaired electron transitions into a p* orbital, from
where it is transferred onto another substrate to propagate the
radical chain.
Removal of an electron from a p system, as shown in Scheme
1b for benzene as example, leads to a p radical cation, which can
be trapped by either another radical through radical
recombination (not shown) or by a nucleophile at the positive
charge site. [2] In the latter case the unpaired electron acts as a
spectator during the bond-forming process. In the case of
aromatic substrates, the sequence is usually terminated by a
second oxidation and deprotonation that restores the aromatic p
system.
•
Pyr+ showed that this reaction can be reproduced in solution. A
similar addition/radical fragmentation sequence occurs also in the
gas-phase reaction of amines with the aliphatic distonic radical cation
Oxo+C•, showing that the observed charge-spin synergism is not
limited to aromatic systems.
(a) Reductive SET:
Nu
Nu
X
e
Nu
e
X
σ radical
π* radical
(b) Oxidative SET:
Introduction
Nu
Nu
Nu-H
Free radicals have found numerous applications in synthetic
chemistry and are also involved in many biological systems, as
well as intermediates in combustion and environmental
transformation and degradation processes. Since the unpaired
electron is considered as the most reactive site, it comes as no
surprise that reactions involving radicals are commonly believed
to proceed through radical pathways.
However, that this picture cannot be generalised is illustrated
by several known reaction types where the unpaired electron
seems to act as an ‘observer’ of non-radical process occurring in
the same molecule. The Sandmeyer reaction is an example of a
radical-nucleophilic aromatic substitution (SRN1) that was
discovered by Bunnett in 1970 (Scheme 1a)[1] and involves a
nucleophile (Nu) replacing an existing substituent X (most
commonly a halogen atom) on the aromatic ring. The reaction is
initiated by reductive single electron transfer (SET), which is
e
e
H
π radical
π radical
CID
H
(c) This work:
Nu-R
Nu
I
(or hν)
Nu-R
N
N
N
N
I
R
σ radical
π radical
Scheme 1. Examples for reactions of nucleophiles with radicals and radical
ions.[1,2]
Here we present a variation of this concept, which we discovered
during an ongoing research project aimed at gaining a better
mechanistic understanding of environmental radical polymer
degradation. These processes are initiated at the gas/solid
interface, which can be experimentally simulated by performing
gas-phase studies in multi-stage mass spectrometers using the
capability of ion traps to generate distonic radical ions, and then
study their reactions with neutral compounds.[3] The chemistry of
reactant radical ions and charged reaction products can be
directly monitored in a time resolved fashion to identify reactive
intermediates associated with the formation of early degradation
products,[4] and to obtain kinetic data for individual reaction steps.
The distonic radical ion approach to study gas-phase radical
reactions by mass spectrometry commonly considers the charge
tag as innocent entity, which does not influence the mechanism
[a]
Mr B. Andrikopoulos, Ms P. K. Sidhu, Ms B. I. Taggert, Dr. J. G.
Nathanael, Prof. R. A. J. O’Hair, Prof. U. Wille
School of Chemistry, Bio21 Institute
The University of Melbourne
30 Flemington Road, Parkville, Victoria 3010, Australia
E-mail: rohair@unimelb.edu.au; uwille@unimelb.edu.au
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
by which the radical reacts, compared to the neutral counterpart.
Distonic radical ions have therefore been successfully employed
to investigate radical reactions in many different areas, for
example atmospheric chemistry,[5] organic synthesis,[6] and
biological chemistry.[7] Our previous studies, where we used
distonic radical cations to explore the gas-phase chemistry of
peroxyl, alkoxyl and alkyl radicals,[6,8] enabled identification of
several new reaction pathways that could contribute to polymer
degradation under environmental conditions.[9] However,
Kenttämaa et al. recently found that the radical reactivity of some
distonic radical cations can be challenged by a competing
nucleophilic reaction enabled by the positive charge.[10]
define the interplay between charge and spin in these
transformations.
Results and Discussion
1. Radical generation in the mass spectrometer
•
•
The distonic aryl and alkyl radical cations 4-Pyr+ , 3-Pyr+ and
Oxo+C• were generated in situ in the ion trap mass spectrometer
(see Experimental Methods for details of the experimental set-
up). The isomeric pyridinium radicals 3-Pyr+• and 4-Pyr+• (m/z 93)
were obtained from the respective N-methyl iodo pyridines 7a and
7b through collision-induced dissociation (CID), as described
previously (Scheme 2a,b).[8a]
Amines are prevalent structural moieties in polymers (Figure
1a), for example in melamine formaldehyde resins or melamine
crosslinked polyesters (i), in polyamines, such as polyvinyl amine
(ii), polyethylene imine (iii) and poly(aminostyrene) (iv).
(a)
O
O
O
I
ONO2
4-Pyr+
m/z 93
CID
- I
CID
CID
(a) Amines in polymers:
- CO
- NO2
- 2 H2CO
N
N
N
NH2
NH2
(ii)
RO
N
OR
7a
m/z 220
8
9
m/z 121
m/z 227
N
N
n
(c)
Ph
(b)
O
ONO2
I
RO
N
N
N
OR
OR
3-Pyr+
m/z 93
CID
- I
H
N
O
O
ONO2
O
CID
CID
- glycine
Oxo+C
m/z 148
OR
(i)
N
n
n
O
O
(iii)
(iv)
- NO2
- H2CO
7b
m/z 220
Ph
10
m/z 299
11
m/z 224
NH3
(b) Model amines studied in this work:
O
NH2
N
NH
NH2
N
Scheme 2. Generation of the distonic radical cations 4-Pyr+•, 3-Pyr+ and
Oxo+C• in the mass spectrometer.
•
N
H
MOR
TEA
DIPEA
DIPA
CHA
AA
(c) Distonic alkyl and aryl radical cations studied in this work:
4%
O
The pyridinium ester 9 was originally designed to access an O-
centred alkoxyl radical through CID of the labile O-NO2 bond but
rapid sequential loss of two H2CO molecules led to the acyl radical
5%
N 2
5%
92%
89%
7%
Ph
3
1
4
90%
N
3
5
6
4%
5
O
4%
•
8 at m/z 121, which upon a second CID produced 4-Pyr+
4-Pyr+
3-Pyr+
Oxo+C
(Scheme 2a). The synthetic procedure and CID mass spectra for
Figure 1. (a) Amine motifs in polymers and (b) model systems studied in this
work; (c) spin densities in 4-Pyr+•, 3-Pyr+• and Oxo+C• calculated with M062X/6-
31+G*.
compound 9 are provided in the Supporting Information (SI,
•
Figure S1). Comparison of the reactions of 4-Pyr+ derived from
precursors 7a or 9 confirmed that its reactivity was independent
of the generation method. We recently developed glycerol
derivative 10 as precursor for Oxo+C• (m/z 148), which is obtained
through sequential CID via the cyclic tertiary carboxonium ion
11,[7d] followed by homolytic scission of the O–NO2 bond and loss
of formaldehyde (Scheme 2c).[8b] According to DFT calculations
(see below for details) the spin density in these distonic radical
cations is predominantly (ca. 90%) located on the desired carbon
atom (Figure 1c).
While hydrolytic degradation of melamine-containing polymers
has been well studied,[11] direct involvement of amine moieties in
radical-mediated degradation has received considerably less
attention. We have therefore explored the gas-phase reactions of
•
•
the distonic aryl and alkyl radical cations 4-Pyr+ , 3-Pyr+ and
Oxo+C• with a series of primary, secondary and tertiary amines,
e.g., triethyl amine (TEA), diisopropylethyl amine (DIPEA),
diisopropyl amine (DIPA), cyclohexyl amine (CHA), allyl amine
(AA) and morpholine (MOR) (Figure 1b,c), which represent
simplified models for amine residues present in polymers with
sufficient volatility to enable gas-phase studies in the ion trap
mass spectrometer. Rapid nucleophilic addition of the amine to
the distonic radical cation that is directed by the unpaired electron
in conjunction with the charge is a key-finding (Scheme 1c), with
2. Gas-phase mass spectrometric studies
Reactions with Pyr+•: Figure 2 shows the mass spectra of the ion-
molecule reactions of 4-Pyr+• with TEA (a), DIPEA (b), DIPA (c),
and CHA (d), which were taken at reaction times where the signal
•
intensity of 4-Pyr+ had dropped to about 20% (120-240 ms).
Product assignment was accomplished using computational
studies (see section 3).
Interestingly, in the reaction with these amines the expected
hydrogen atom abstraction (HAT) to give Pyr+H at m/z 94 was
•
4-Pyr+ reacting in a similar fashion also in solution. The
experimental studies were augmented by DFT calculations to
gain mechanistic understanding of these reactions, in particular to
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
only a very minor pathway.[12] In the reaction with TEA the major
seconds, suggesting closed-shell species. Fragmentation by CID
only occurred only at very high energies, which further confirmed
their stability. It should be noted that the products at m/z 100 in
the reaction of TEA, at m/z 128 in the reaction of DIPEA, at m/z
100 and 102 in the reaction of DIPA, and at m/z 98 in the reaction
of CHA could be assigned to protonated and/or oxidized amines,
which were likely formed under the experimental conditions and
not further investigated.[13]
product at m/z 165 could be assigned as the 4-diethylamino
•
pyridinium ion 12a resulting from formal addition of NEt2 . The
reaction with DIPEA led to a major product at m/z 179, which
could be assigned as the ethyl isopropylamino pyridinium ion 13a,
suggesting a reaction through addition of the amine to the
aromatic ring and loss of an isopropyl moiety. A product resulting
from loss of an ethyl group (which would appear at m/z 193) was
not formed. Similarly, the product at m/z 151 in the reaction with
DIPA corresponds to the isopropyl amino pyridinium ion 14a,
resulting from formal addition of NHiPr•. In the reaction involving
CHA, the product at m/z 109 could be assigned as the 4-amino
pyridinium ion 15a, which could be formed through formal addition
•
The reaction of 3-Pyr+ with these four amines also led to
formation of the respective isomeric 3-substituted products 12b –
15b (the site of amine substitution was elucidated
computationally, see section 4). However, the HAT reaction to
give Pyr+H appeared to be more favourable than in the reactions
•
•
•
of NH2 . Products corresponding to addition of the amine to 4-
with 4-Pyr+ . The mass spectra for the reaction of 3-Pyr+ with
•
Pyr+ without elimination of an alkyl moiety were not detected in
TEA, DIPEA, DIPA and CHA are shown in Figures S2-S5.
any of these reactions. Isolation of 12a - 15a in the ion trap
showed no further reaction with the respective amine over several
NHiPr
(a)
(c)
NEt2
N
N
14a
12a
4-Pyr+
*
4-Pyr+
*
Et2N
iPrHN
Pyr+H
Pyr+H
H2NiPr2
(b)
(d)
NEtiPr
NH2
N
N
4-Pyr+
13a
EtiPrN
15a
Pyr+H
NH2
*
4-Pyr+
Pyr+H
*
Figure 2. Mass spectra of the ion-molecule reactions of 4-Pyr+• (m/z 93) with (a) TEA after 135 ms ([TEA] ca. 1.2 ´ 1010 molecules cm-3), (b) DIPEA after 240 ms
([DIPEA] ca. 1.1 ´ 1010 molecules cm-3), (c) DIPA after 120 ms ([DIPA] ca. 1.2 ´ 1010 molecules cm-3) and (d) CHA after 120 ms ([CHA] ca. 1.2 ´ 1010 molecules
cm-3). The mass-selected precursor ion is indicated by an asterisk (*).
To compare the chemical reactivity of the isomeric Pyr+•, we
analysed the mass spectrometric yields obtained for the reaction
with each amine at the same time point. Since mass
discrimination could be excluded the peak area of each ionic
species directly reflected their concentration. While the overall
reactivity of both Pyr+• with the amines is similar, compared with
4-Pyr+• the reaction of 3-Pyr+• gave a higher yield of the HAT
product Pyr+H (m/z 94) at the expense of amine adduct formation,
indicating a higher radical reactivity of the latter (Table S1).
To further explore this finding, we studied the reaction of AA,
which could support radical pathways through both HAT (allylic C-
H bonds) and addition (p-system). The mass spectra obtained
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
after 110 ms for the reaction of AA with 4-Pyr+• and 3-Pyr+•
revealed clear differences between these two radical cations
(Figure 3; the mass spectrometric yields at this time point are
included in Table S1).
occurred in the reaction with 4-Pyr+•, but several unidentifiable
products appeared at higher m/z values that suggest MOR
addition, followed by fragmentation (see Figure S6).
Reactions with Oxo+C•: Compared with both isomeric Pyr+• the
reaction of Oxo+C• with amines was considerably slower. Figure
4 shows the mass spectrum of the reaction with TEA after 500
ms, where the signal intensity of unreacted Oxo+C• was about
30%. The mass spectra of the reactions involving DIPEA and
DIPA showed a similar outcome and are provided in Figure S7.
While all reactions were characterised by significant amounts of
protonation and/or oxidation side-products of the respective
amines, none gave a product at m/z 149 corresponding to
Oxo+CH, clearly excluding a pathway through HAT.
(a)
109
NH2
N
4-Pyr+
15a
93
*
N
120
16a
Pyr+H
94
Et3NH
(b)
120
NEt2
NH2
O
Ph
O
N
19a
Oxo+C
O
15b
N
3-Pyr+
Et2N
93
*
*
Ph
NEt3
16b
18a
109
N
Pyr+H 94
134
17b
Figure 4. Mass spectrum of the reaction of OxoC+• (m/z 148) with TEA after 500
ms ([TEA] ca. 8.5 ´ 109 molecules cm-3). The mass-selected precursor ion is
indicated by an asterisk (*).
Figure 3. Mass spectra of the reaction of AA with (a) 4-Pyr+• (m/z 93) and (b) 3-
Pyr+• (m/z 93) after 110 ms ([AA] ca. 1.2 ´ 1010 molecules cm-3); the mass-
selected precursor ion is indicated by an asterisk (*).
The adduct at m/z 220 formed in the reaction with TEA formally
results from addition of NEt2 to Oxo+C•. Computational studies
•
revealed the likely structure of the product as the open-chain
iminium ion 19a (see section 4). CID of 19a showed fragmentation
into an ion at m/z 105, which can be assigned as the
phenylacylium ion PhCºO+ (21). In addition, a product at m/z 206
was also formed, which was tentatively assigned as the
acylammonium ion 18a.[15] We believe that the latter results from
ring-opening of OxoC+• to 20, followed by fragmentation into the
acylium ion 21,[8b] which is subsequently trapped by the amine
(Scheme 3). This assignment could be confirmed by CID on
isolated 18a in the ion trap, which revealed fragmentation into 21.
Thus, while the major product at m/z 109 in the reaction of 4-Pyr+•
could be assigned as the 4-amino pyridinium ion 15a resulting
from formal addition of the amine and loss of an allyl moiety
(Figure 3a), the isomeric 15b was obtained only as a minor
product in the reaction of 3-Pyr+•. Instead, the major product
appeared at m/z 120, which could be assigned as the vinyl
pyridinium ion 16b (Figure 3b). In addition, small amounts of a
product at m/z 134 also appeared, which could correspond to the
allyl pyridinium ion 17b. In the reaction of 4-Pyr+•, on the other
hand, the isomeric vinyl pyridinium 16a was a minor product and
only trace amounts of 17a at m/z 134 were formed. Both products
16 and 17 indicate radical addition of Pyr+• to the p-system in AA.
Products from oxidation and/or protonation of AA could not be
detected, because their masses were below the mass cut-off of
the ion trap. Overall, formation of Pyr+H at m/z 94 through HAT
was a minor reaction of both Pyr+• but slightly more important for
3-Pyr+• than for 4-Pyr+•.[14]
O
Ph
O
O
O
O
NR3
Ph
O
Ph
Ph
Ph
NR3
O
Oxo+C
O
O
21
m/z 105
20
18
On the other hand, in the reaction of MOR with 3-Pyr+• HAT
and formation of Pyr+H was found as the major pathway. HAT also
Scheme 3. Proposed mechanism for formation of the acyl ammonium ions 18
in the reaction of Oxo+C• with amines.
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
•
3. Kinetic studies of the gas-phase reaction of Pyr+ with
amines
reactions in some cases. Within error, the decay rate coefficients
for Pyr+• are similar to the sum of the rate coefficients for product
formation, e.g., Pyr+H, Pyr+-NR2 and amine oxidation (see Table
S2), confirming that no major pathway resulting in loss of the
charge tag occurred, which would render the product
undetectable by mass spectrometry.
Absolute rate coefficients for the consumption of Pyr+• through
reaction with TEA, DIPEA, DIPA, CHA and AA according to: Pyr+•
+
amine ® products, were obtained under quasi-thermal
conditions by monitoring the decay of the Pyr+• signal at m/z 93
([M]•+) as a function of reaction time under pseudo-first order
conditions.[8b] The rate coefficients for product formation, i.e.,
Pyr+H, the adducts Pyr+-NR2 as well as 16, 17 and amine
oxidation products were obtained from kinetic modelling of the
experimental concentration–time profiles using the program
Dynafit 4.[16] Details are given in Experimental Methods.[17] The
rate data are compiled in Table 1. An exemplary concentration-
time profile is shown in Figure S11, and the kinetic data for amine
oxidation are compiled in Table S2. It should be noted that the
concentration-time profiles for all products did not show any
induction period, suggesting that they result from either a one-
step reaction or from multistep processes, where intermediates
are transformed into the products on faster timescales than can
be resolved in the mass spectrometer. The mechanism will be
discussed in section 4.
Comparison of the rate coefficients kAdd for the formation of
Pyr+-NR2 with that for the ‘conventional’ radical pathways kRad
through HAT or addition to p-systems (formation of Pyr+H, as well
as 16a,b and 17b in the reaction involving AA) confirmed the
differences in the chemical behaviour of both Pyr+• observed in
the product studies (section 2). For the reaction with the tertiary
and secondary amines TEA, DIPEA and DIPA with 4-Pyr+• a value
for kAdd/kRad of 9.7 is obtained, whereas it is only 3.3 for 3-Pyr+•,
clearly showing the higher reactivity of 4-Pyr+• for amine addition
at the expense of conventional radical reactivity. This finding was
confirmed by comparing kAdd/kRad for the reaction involving AA
(where several sites are available to promote radical chemistry),
which revealed a value of 2.4 for 4-Pyr+• and only 0.4 for 3-Pyr+•.
For the reaction involving the less nucleophilic primary CHA, the
ratio kAdd/kRad is about 50% smaller for both Pyr+• compared with
TEA, DIPEA and DIPA, respectively, which further supports the
suggestion that formation of Pyr+-NR2 involves nucleophilic amine
addition to Pyr+•.
Table 1. Absolute second-order rate coefficients k for the reaction of 4-Pyr+• and
3-Pyr+• with amines at 298 K.[a,b,c]
k ´ 10-10 (cm3 molecule-1 s-1)
4. Computational mechanistic studies
kAdd
kRad
/
Amine
Efficiency[e]
Pyr+•
consumption
Pyr+H
formation
Pyr+-NR2
formation
[d]
In recent gas-phase studies of isomeric distonic pyridinium radical
cations with amino acids Kenttämaa et al. reported outcomes
similar to those shown in Figures 2-4. They proposed that the
reaction is initiated by nucleophilic addition of the amine to the
pyridinium radical, where the efficiency of the transfer of the
amino group was found to increase with the electron affinity of the
radical cation; however, an ‘active’ role of the unpaired electron
during the addition was not considered.[10a,d,g] Thus, in the reaction
with 4-Pyr+• addition of NR3 could occur at the radical site C-4,
which should also be activated for nucleophilic attack, to form
adduct 22, which is followed by homolytic b-fragmentation to give
the amino pyridinium ions 12-15 with release of R• (Scheme 4).[19]
In the reaction involving 3-Pyr+•, where the sites of positive
polarisation and unpaired electron do not coincide, NR3 addition
might occur at C-3 (radical site) as well as C-2 and C-4, which are
activated by positive polarisation (not shown).
8.9
0.8
9.1
11.4
3.2
70%
57%
TEA
7.3
2.1
6.8
9.3
0.9
7.3
7.9
3.0
--[f]
DIPEA
DIPA
CHA
AA[g]
8.0
1.4
4.2
10
1.0
9.8
9.8
3.7
66%
58%
8.1
2.3
8.4
15
2.7
15
5.6
2.0
100%
86%
12
5.0
10
9.3
0.6
8.3
2.4
0.4
69%
61%
8.2
0.7
3.3
[a] Data in dirst line for 4-Pyr+•, in second line (in italics) for 3-Pyr+•. [b] Rate
coefficients for Pyr+• consumption from pseudo-first order measurements; rate
data for product formation were fitted using the Dynafit 4 program (see
Experimental Methods). [c] Rate coefficients have an experimental error of 25%.
[d] kRad = all radical processes (see text). [e] Determined from Pyr+• decay rate
using trajectory collision rate kado (´ 10-9 cm3 molecule-1 s-1): TEA = 1.28, DIPA
= 1.40, CHA = 1.40, AA = 1.35. [f] Not available (see text). [g] Rate coefficient
(in cm3 molecule-1 s-1) for formation of 16a: (2.8 ± 0.7) ´ 10-10, 16b: (5.8 ± 1.5) ´
R
NR3
nucleophilic
addition
NR2
R2N
NR2
δ+
β-fragmentation
δ+
4-Pyr+
δ+
N
N
N
N
R
10-10,17b: (1.0 ± 0.3) ´ 10-10
.
I
II
22
12 - 15
Overall, the reactions of both Pyr+• with amines were very fast with
the rate coefficients for the consumption of 4-Pyr+• being about
20% higher than for 3-Pyr+•. The reaction efficiencies, which were
calculated from the trajectory collision rate, kado, derived from the
average dipole orientation (ADO) theory[18] for all amines (except
for DIPEA for which polarizability data were not available from
literature), range from 66-100% for 4-Pyr+• and 57-86% for 3-
Pyr+•, respectively, indicating practically collision-controlled
Scheme 4. Proposed mechanism for formation of the amino-substituted
pyridinium ions 12 – 15 through non-radical addition of amines to Pyr+•, followed
by homolytic b-fragmentation (shown exemplary for the reaction of 4-Pyr+•).
To better understand the interplay between charge and unpaired
electron in these reactions, DFT calculations were performed at
the M062X/6-31+G* level of theory,[20a,21] which has been
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
previously shown to provide sufficiently accurate energies to
identify the most likely reaction pathways.[8b,22] The energies of the
gas-phase calculations are given as enthalpies, as they best
describe chemical reactivity under low-pressure bimolecular
conditions,[23] whereas for the reactions in solution (see section 5)
free energies are considered. The Gaussian archive entries for all
optimized stationary points, energies and imaginary frequencies
for transition states are given in the SI.
Table 2 compiles the energies for the different reaction
pathways of Pyr+• with TEA, DIPEA and MOR. The reaction
proceeds through formation of an initial ion-molecule association
complex (reactant complex A) that is 40 - 60 kJ mol-1 lower in
energy than the free reactants, or ‘entrance channel’, which is
defined as reference point and set to 0 kJ mol-1. Pathways with
energies above the entrance channel are too slow to occur on the
timescale of the experiment.[24]
Table 2. Gas-phase reaction of Pyr+• with amines. M062X/6-31+G* enthalpies in kJ mol-1, relative to the free reactants.
Pyr+
Pyr+
TS2
TS1
Pyr+
NR2
+
Pyr
N+R3
N
N
N
NR3
NR3
A
R
model for
charge site
model for
radical site
free reactants
0.0 kJ mol-1
B
C
Pyr+H
4-Pyr+
4-Pyr
(reactant complex)
(product complex)
TS3
model for
electric field
Pyr+H
R2N(R-H)
N
D
(product complex)
4-Pyr
Entry
Pyr+•
amine
Reaction site
in Pyr+•
A
TS1
--[b]
B
TS2
C
TS3[a]
D
1
2
4-Pyr+•
4-Pyr+•
NEt3 (TEA)
-56.3
-51.9
C-4
C-4
-188.4
-104.7[c]
-183.0[c]
-28.9
-191.0
-95.6[c]
-100.2[e]
-193.2[c]
-208.0[e]
--[d]
-37.4[f]
--
NEtiPr2 (DIPEA)
--[b]
-157.9
-215.1[f]
morpholine
(MOR)
--[b,g]
-157.6[g]
-45.3[h]
--
--
--
--
-31.3
-154.0
3
4
4-Pyr+•
3-Pyr+•
-46.8
-44.7
C-4
-26.3[h]
-19.5[j]
-123.1[j]
C-3
C-4
--[b]
-41.1
-179.8
-50.4
-80.1[c]
--
-154.5[c]
--
-36.9
-198.5
NEt3
5
6
7
Pyr+H
NEt3
NEt3
NEt3
-43.3
-14.6
C-4
C-4
C-4
-25.7
32.7
--[b]
-28.1
16.4
--
--
--
--
--
--
--
--
--
--
--
--
4-Pyr•
Å···4-Pyr•
-187.2[k]
-194.3
[a] HAT from N-CH2. [b] No barrier (could not be located). [c] R• = Et•. [d] Not determined. [e] R• = iPr•. [f] (R-H)• = iPr•. [g] N addition. [h] O addition. [j] HAT from O-
CH2. [k] Geometry differed considerably from that of the other reactant complexes A.[25]
A transition state TS1 for the addition of the amine to the radical
centre in 4-Pyr+• (C-4) and in 3-Pyr+• (C-3) could not be located,[26]
suggesting a barrierless step. Amine addition at C-4 in 3-Pyr+• will
be discussed below. Formation of the Pyr+•-amine adducts B is
strongly exothermic with 158 – 189 kJ mol-1 (entries 1-4). The
subsequent homolytic b-fragmentation via TS2 is associated with
a barrier of about 60-100 kJ mol-1 and leads to the product
association complex C in an essentially energy-neutral or in the
case of 3-Pyr+• slightly endothermic reaction. However, since B,
TS2 and C are energetically well below the entrance channel, the
b-fragmentation should be very fast. These findings support the
mass spectrometric data that adducts B could not be detected
even at very short reaction times. Relief of steric strain in the case
of DIPEA is likely responsible for the higher exothermicity of step
B ® C (entry 2), where the calculations confirm the
experimentally observed exclusive release of the more stable
secondary radical iPr• over the primary Et• (see Figure 2b).
Although HAT from the a-carbon atom in amines is
thermodynamically highly favourable, this pathway is associated
with a barrier (TS3) of 15-27 kJ mol-1 for the reaction of 4-Pyr+•
with TEA and DIPEA, which cannot compete with the barrier-less
formation of the amine adduct B. On the other hand, TS3 is
considerably lower for the reaction involving 3-Pyr+• (27 vs 8 kJ
mol-1, entries 1 vs 4), which is in line with the experimentally
observed higher radical reactivity of 3-Pyr+•. In MOR the N-CH2
moiety is the kinetically most favourable site for HAT, with TS3
being about 15 kJ mol-1 (entry 3). The computations predict also
a barrierless addition of MOR to 4-Pyr+• via the N atom. As
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
mentioned above and shown in the mass spectrum in Figure S6a,
the signals at higher m/z values suggest that addition of MOR to
4-Pyr+• occurred, but it was not possible to assign the various
products that result from subsequent fragmentation of the 4-Pyr+•-
MOR adduct.
reaction at C-3 and C-4. The calculations reveal that amine
addition at C-3 is not only barrier-less but also thermodynamically
clearly more favourable than at C-4, despite being the less
electropositive site (Table 2, entry 4).
(a)
To evaluate the role of the radical site in 4-Pyr+•, we calculated
the addition of TEA to C-4 in the closed-shell Pyr+H (similar to a
nucleophilic aromatic substitution, SNAr). The data in entry 5 show
that this addition is associated with a barrier of some 20 kJ mol-1
and is endothermic. On the other hand, for the addition of TEA to
C-4 in the neutral 4-Pyr• (similar to an SRN1 mechanism; see
Scheme 1a) both TS1 and adduct B are located above the
entrance channel, making this reaction too slow to occur under
our experimental conditions (entry 6). These findings show that
the unpaired electron in Pyr+• in conjunction with the positive
charge facilitate amine attack on the aromatic ring at the site of
the unpaired electron. A similar accelerating effect on the amine
addition was found, when the charge-causing methyl group in 4-
Pyr+• was replaced by a positive point charge (+1.0), i.e., Å···4-
Pyr•, to simulate an electric field (entry 8 vs 9 and 1).[27]
25 (HOMO)
22 (SOMO)
LUMO+3
LUMO+4
(b)
(c)
1.496
1.529
2%
1.528
3%
1.495
2.088
52%
1.516
33%
1.488
4%
15%
22%
1.383
SOMO
(HOMO)
10%
SOMO
(HOMO)
SOMO
(HOMO-1)
12%
9%
21%
NEt3
NEt3
11%
N
N
B
TS2
III
IV
Natural bond orbital (NBO) population analyses[20b,28] revealed
that in both 3- and 4-Pyr+• the SOMO is submerged below the
MOs of the aromatic p-electrons to become the fourth-highest
energy molecular orbital (Figure 5a with 4-Pyr+• as example, only
the HOMO is shown, HOMO-1 and HOMO-2 on the remaining
centres of the aromatic system are similar). The electron spin
density plot shown exemplary for 4-Pyr+• confirms the sp2
character of the SOMO on C-4, which is unrelated to the three
higher lying occupied orbitals of the p system. On the other hand,
the SOMO interacts with the s* orbitals of neighbouring C-C
bonds (LUMO+3 and LUMO+4), enabling delocalisation so that
about 10% of the spin density in 4-Pyr+• can be found on C-2 and
C-6, and in 3-Pyr+• on N and C-5, respectively (see also Figure
1).[29,30]
Figure 5b shows the optimised geometry of adduct B in the
reaction of 4-Pyr+• with TEA. Only 5% of the spin density is located
on the amine moiety, which is characterised by all four C-N bonds
being slightly elongated compared with typical C–N single bonds
(ca. 1.479 Å).[31] The remaining 95% spin density is distributed
over the pyridinium ring involving two p-shaped SOMOs at both
C-4 and N (which are now also the two highest occupied MOs)
and the p* orbitals of the connecting carbon framework (not
shown), suggesting that adduct B could be described through the
resonance forms III and IV. In the transition state (TS2) of the
subsequent homolytic b-fragmentation, the SOMO (=HOMO)
interacts with the s* orbital (LUMO) of the breaking N–C bond.
TS2 occurs comparatively late on the potential energy surface
with about 50% of the spin density on the leaving Et• fragment.
The bond scission is accompanied by considerable contraction of
the remaining N–Et bonds to practically single bond length, as
well as of the N–Pyr bond, which at 1.383 Å is close to that of a
partial C–N double bond (1.352 Å)[31] and already resembles the
product resonance structure I in Scheme 4.
LUMO
Figure 5. Reaction of 4-Pyr+• with TEA. Geometries optimised with M062X/6-
31+G*. (a) Electronic arrangement and MO configuration in 4-Pyr+•; NBO
population analysis with wB97x/6-31+G*. (b) Geometry and MO configuration in
adduct B; NBO population analysis with wB97x/6-31+G*. (c) Geometry and MO
configuration of transition state TS2; NBO analysis with M062X/6-31+G*. Bond
lengths in Å, the percentage numbers are spin densities. The α electron spin
density, highest doubly occupied molecular orbital (HOMO), singly occupied
molecular orbital (SOMO) and selected lowest unoccupied molecular orbitals
(LUMO) are indicated. The numbering of molecular orbitals is assigned in the
order of increasing energy.
Figure 6a shows that the optimised geometries for B and TS2
after initial amine addition at C-3 closely resemble those found for
the addition at C-4 in 4-Pyr+• with regards to both geometry and
spin density. On the other hand, nucleophilic amine addition at C-
4 in 3-Pyr+• is associated with a small barrier (TS1) of 3 kJ mol-1.
Analysis of the orbitals revealed that the bond forming process
involves interaction of the lone pair at N (in HOMO-1) with the
LUMO at C-4 (MO not shown), without any contribution by the
unpaired electron that is located in HOMO-4 (Figure 6b; HOMO,
HOMO-2 and HOMO-3 are occupied by the six p electrons of the
Pyr system). The adduct B is only slightly lower in energy than the
reactant complex A and characterised by a long C-N bond of
1.619 Å and 90% spin density on C-3 (as in 3-Pyr+•) and only 5%
on the TEA moiety, clearly indicating that the unpaired electron is
not involved in the amine addition at C-4.
The calculated energy profile for the reaction of Pyr+• with AA
revealed that the radical processes for 3-Pyr+• through HAT and
addition to the p system are lower in energy than for 4-Pyr+•,
confirming the experimentally found higher preference for 3-Pyr+•
to undergo ‘conventional’ radical chemistry. The data are
presented in Scheme S1.
As mentioned above, the addition of amines to 3-Pyr+• could
principally occur at the radical site C-3, as well as at C-2 and C-
4, since both latter positions would be expected to be activated
for nucleophilic attack. We will focus our discussion on the
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
(a) Attack at C-3
Attempts to compute a direct pathway from amine adduct F to
1.528
1.531
the amino substituted dioxolanes
K through homolytic b-
1.496
43%
1.399
22%
2.112
1.496
fragmentation failed, but instead revealed the open-chain iminium
complex H, which could be formed through a concerted ring-
opening/b-fragmentation.[32] However, since the barrier TS7 lies
close to or even above the entrance channel, respectively, this
pathway should be too slow to occur under our experimental
conditions.
39%
5%
3%
6%
1.519
21%
17%
18%
1.482
22%
4%
B (C-3 adduct)
TS2 (C-3 adduct)
On the other hand, a stepwise mechanism consisting of
radical ring-opening via TS5 to yield the radical cation G, followed
by homolytic b-fragmentation via TS6 to yield H is highly feasible
with all computed ground and transition states lying well below the
entrance channel. It should be noted that for the reaction with
DIPEA the predicted preferred radical scission of an N–iPr bond
over an N–Et bond in the fragmentation G ® H is in agreement
with the experimental findings (see Figure S7a).
(b) Attack at C-4
1.480
1.515
1.511
1.619
3%
1.472
91%
1.479
3%
SOMO
(HOMO-4)
2.069
3%
90%
SOMO
(HOMO-3)
HOMO
2%
1.505
3%
3%
TS1 (C-4 adduct)
B (C-4 adduct)
HOMO-1
HOMO-1
HOMO-2
(a)
Oxo+C
NR3
NR3
O
O
G
-156.6
O
O
Oxo+C
NR3
TS5
TS4
3
1
5
1
+
NR3
Ph
Ph
-45.3
-20.9
-45.8
*
Figure 6. Reaction of 3-Pyr+• with TEA. Geometries optimised with M062X/6-
31+G*. (a) Adduct B and TS2 during reaction at C-4. (b) TS1 and adduct B
during reaction at C-3. NBO population analysis of MO configuration analysed
with M062X/6-31+G*. Bond lengths in Å, the percentage numbers are spin
densities. The α electron spin density of relevant MOs is shown.
free reactants
0.0 kJ mol-1
E
F
-125.4
-119.2
(reactant complex)
TS7
-R
-149.2
-57.9
-37.3
4.0
-3.1[b]
-67.9[a]
TS6 -89.6[b]
-63.7[a]
-11.0
TS8 -25.4[c]
-25.0
NR2
O
O
Ph
NHR2
O
O
OxoC+H
K
Ph
To conclude, upon amine addition to the radical site in either 3- or
4-Pyr+• the SOMO transitions from an sp2 orbital (s radical) to a p
orbital (p radical), enabling spin delocalisation over the Pyr ring
that leads to stabilisation of adduct B by about 150-200 kJ mol-1,
compared to the reactions with the closed-shell (Pyr+H) or neutral
(4-Pyr•) counterparts, respectively (Table 2, entry 1 vs 5 and 6).
The unpaired electron is ejected via an alkyl radical in the
subsequent homolytic b-fragmentation, leading to formation of the
amino substituted pyridinium [Pyr-NR2]+.
In contrast to Pyr+•, amine addition to Oxo+C• occurs at a
carbon centre that is not a part of an aromatic p system. Scheme
5a shows the calculated energies of the stationary points of the
reaction of Oxo+C• with TEA and DIPEA. Amine addition to form
the adduct F via TS4 is kinetically considerably more favourable
than HAT from the a-CH group in the amines via TS8, confirming
the experimental finding that Oxo+CH (m/z 149) was not formed
(see Figure 4). The geometry of TS4 is shown in Scheme 5b for
the reaction involving TEA, the geometries of all stationary points
for this reaction are presented in Figure S19. TS4 is characterised
by a long C–N distance of 2.308 Å consistent with an early
transition state. MO analysis revealed a strong interaction
between the SOMO (= HOMO-1) and the LUMO on the tertiary
carbon atom (C-3) of the dioxolane ring, enabling spin
delocalisation. In the C–N bond forming process the SOMO
(cannot be located)
R
R2N(R-H)
H
J
(product complex)
-160.3
(product complex)
-133.8[a]
O
Ph
-168.8[b]
-162.6[a]
H
O
-188.8[c]
OxoC
-172.2
(b)
1.474
1%
SOMO
(HOMO-1)
LPN
10%
2.308
3%
41%
19%
N
5%
LUMO
6%
7%
7%
O
Ph
O
LPN
SOMO
TS4
Scheme 5. (a) M062X/6-31+G* enthalpies (in kJ mol-1) of the stationary points
for the various pathways in the reaction of Oxo+C• with TEA (top number) and
DIPEA (bottom number in italics), relative to the free reactants. HAT pathway
calculated for abstraction from the amine a-carbon. [a] R• = Et•, [b] R• = iPr•, [c]
HAT from secondary C-H in iPr. *TS4 for the addition of DIPEA could not be
located, suggesting a barrier-less process. (b) Optimised geometry and MO
configuration of TS4 for the addition of TEA to Oxo+C•; NBO analysis with
wB97x/6-31+G*. Bond lengths in Å, the percentage number are spin densities.
The α electron spin density, and selected molecular orbitals are indicated.
5. Reaction of 4-Pyr+• with amines in solution
accepts one electron of the lone pair at nitrogen, i.e., LPN
®
Since Kenttämaa et al. showed in previous work that distonic
pyridine radical cations react with ethers, such as THF, in the
condensed phase through both radical and non-radical
pathways,[10c,f] we explored whether the results from our mass
spectrometric studies could also be reproduced in solution.[33]
The experiments were performed by irradiating a solution of
the radical precursor 7a and the amine in acetonitrile in a Rayonet
photoreactor with l = 350 nm[10c] at 35oC for 24 hours. The
SOMO, leaving the non-participating electron on N (11% spin
density). Formation of the adduct F, which has the SOMO centred
on C-3, is strongly exothermic by about 120 kJ mol-1. For
comparison, the addition of TEA to the neutral radical OxoC• (see
Scheme 5a) is predicted to be highly endothermic by ca 168 kJ
mol-1 (data not shown), clearly revealing the role of the positive
charge in Oxo+C• on the reaction energetics.
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
reaction mixture was analysed by direct injection ESI high-
resolution mass spectrometry (HRMS, see Table S3), which
provided confident assignment of the molecular formulae of the
products. Generation of 4-Pyr+• under these conditions was
confirmed by irradiating a solution of 7a in acetonitrile in the
presence of two equivalents of the radical trap TEMPO. A product
at m/z 249 was formed, which was confirmed by HRMS as the
adduct 4-Pyr+-OTEMP (Figure S8).
The mass spectrum in Figure 7a shows that the reaction with
TEA indeed led to formation of the adduct 12a at m/z 165. No
reaction occurred between 7a and TEA in the absence of light
under otherwise identical conditions (Figure S9), clearly excluding
an SNAr mechanism.
reaction with amines in solution. Interestingly, the outcome of the
reaction of 4-Pyr+• with the primary amines AA and CHA in the
condensed phase was different from that in the gas phase. In the
reaction involving AA a product appeared at m/z 149 (Figure 7b),
which could be assigned as the allylamine-substituted pyridinium
24. Likewise, the product at m/z 191 formed in the reaction of 4-
Pyr+• with CHA corresponds to 4-cyclohexylamine pyridinium (25;
shown in Figure S10a). Using the reaction with AA as example,
the mechanism in Scheme 6 suggests that 24 result from the
initially formed amine adduct B through hydrogen loss.
H2N
HN
HN
O2
TS9
HO2
H
TS10 *
112.4
N
N
N
(a)
M
B
0.0
L
HNEt3
(product complex)
-135.1
(product complex)
63.5
OH
Scheme 6. Reaction of 4-Pyr+• with AA in solution. M062X/6-31+G* free
energies in acetonitrile (in kJ mol-1) of the stationary points relative to the free
reactants or relative to the adduct complex B – O2 for the reaction involving O2.
*TS could not be located.
NEt2
N
23
N
Pyr+H
12a
However, the calculated free energies for this homolytic scission
in acetonitrile using the Conductor-like Polarizable Continuum
Model (CPCM) model[34] predict a barrier TS9 of 112.4 kJ mol-1
and an endothermicity of 63.5 kJ mol-1, which seems to be unlikely
to occur under the experimental conditions. We therefore propose
that hydrogen abstraction proceeds with assistance of residual O2
present in the solution. While a transition state for this abstraction
could not be located, indicating a barrierless process, formation
of the product complex M is strongly exothermic.[35]
(b)
HN
N
To explore whether the reaction is limited to amines, we
performed the photolysis of 7a in neat methanol and found
formation of a product at m/z 124, which could be assigned as the
4-pyridinium methyl ether (26; shown in Figure S10b). When the
reaction was performed in deuterated methanol, the mass shifted
to m/z 127, confirming the proposed transfer of the OMe group
onto 4-Pyr+•. However, in contrast to the reactions with the
amines, a large amount of unconsumed radical precursor 7a was
still present after 24 hours. Furthermore, when the reaction was
performed with MeOH dissolved in acetonitrile, formation of 4-
pyridinium methyl ether did not occur. These findings suggest that
interception of the photogenerated tight radical pair [4-Pyr+• I•]
occurs more readily with the more nucleophilic amines allowing
the reaction to be performed under dilute conditions using a
solvent, whereas the lower nucleophilicity of MeOH requires
partial separation of the ion pair for a reaction to occur, which can
be achieved by performing the reaction in neat MeOH.[36]
24
OH
Pyr+H
NH3
N
23
Figure 7. Mass spectra of the reaction of 4-Pyr+I (7a, m/z 220, 0.086 mmol) in
the Rayonet photoreactor (l = 350 nm, 24 hrs, 35oC). (a) Reaction with TEA
(1.79 mmol in 15 mL MeCN), (b) Reaction with AA (3.34 mmol in 15 mL MeCN).
The HRMS data of the products are given in Table S3.
The major signal at m/z 102 could be assigned to protonated TEA
that was present in excess. We also observed formation of a new
product at m/z 110, which could be assigned as the hydroxy
pyridinium 23 that could possibly result from reaction of 4-Pyr+•
with residual water in the solvent, which acts as the nucleophile.
Product 12a was isolated and its structure confirmed by 1H NMR
spectroscopy (Figure S20), providing further support for the
proposed mechanism. Similar to the reactions in the gas phase,
HAT and formation of Pyr+H was only a very minor pathway in the
Conclusions
In this combined experimental and computational study, we have
shown that the presence of a positive charge in conjunction with
an unpaired electron considerably facilitates the nucleophilic
addition of amines to pyridine. Using the isomeric distonic aryl
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
radical cations 4-Pyr+• and 3-Pyr+•, the reaction was found to
occur through amine attack exclusively at the site of the unpaired
electron, which is followed by homolytic b-fragmentation, where
the unpaired electron is ejected via an alkyl radical to give the
amino substituted pyridinium [Pyr-NR2]+. This reaction is the
dominant pathway, rendering ‘conventional’ radical reactions
through HAT or addition to p systems of minor importance.
However, in 3-Pyr+•, where the site of spin and positive
polarisation in the aromatic ring do not coincide, radical behaviour
becomes slightly more important. The barrierless formation and
particular stability of the amine adduct [Pyr-NR3]+•, compared with
the endothermic formation of the closed-shell cationic adduct
[Pyr-NR3]+ through an SNAr-type reaction, or the neutral radical
adduct [Pyr-NR3]• through an SRN1-type reaction, respectively,
can be rationalised by the gain in energy resulting from
transformation of the largely localised sp2-type SOMO in Pyr+• into
a p-type SOMO in [Pyr-NR3]+ that enables delocalisation of the
unpaired electron over the pyridine ring. Exemplary studies for 4-
Pyr+• revealed similar outcomes for the reaction with amines and
even the weaker nucleophile methanol in solution.
conditions included needle potentials between 2.5-4.7 kV, a
capillary temperature of 250°C, capillary voltages between 0.0-
34.0, and tube lens voltages between 0.0-55.0 V. Ions of interest
for MSn experiments were isolated with a 2-5 m/z window. The
CID parameter was selected so the parent ion being fragmented
was reduced to ~15% abundance. Data were collected using
three micro-scans and taking between 10-100 duplicate spectra.
Kinetic Studies. Absolute rate coefficients were determined from
the peak areas in the mass spectra, which directly reflect the
concentration of the respective ionic species in the ion trap. The
rate coefficients for the consumption of Pyr+• through reaction with
the amines were obtained at 298 K by monitoring the decay of the
Pyr+• signal at m/z 93 ([M]•+) as a function of reaction time under
pseudo-first order conditions. Four to five different excess
concentrations[39] of the amine were used on time scales that
allowed the peak intensity of Pyr+• to decrease to ca. 20% and to
build up time-resolved decay profiles (8-13 reaction times). The
pseudo-first order rate coefficient kobs was obtained from the slope
of the plot ln[Pyr+•] vs. reaction time. The second-order rate
coefficient k was determined by plotting kobs vs. [amine], see
Figures S12 and S13. The rate coefficients for product formation
were obtained from kinetic modelling of the experimental
concentration–time profiles using the program Dynafit 4.[16] The
A related amine addition/radical fragmentation sequence was
also found in the reaction with the aliphatic distonic radical cation
Oxo+C• in the gas phase. These findings show that the synergistic
effects caused by the combination of charge and spin are not
limited to aromatic systems.
pseudo-first order rate coefficients for product formation, k’obs
,
were fitted for the HAT pathway: Pyr+• + amine ® Pyr+H + [amine-
H]•, at m/z 94, for the amine addition to Pyr+•: Pyr+• + amine ®
Pyr+-NR2, at m/z 165 (12a/b), m/z 179 (13a/b), m/z 151 (14a/b),
m/z 109 (15a/b), m/z 120 (16a/b) and m/z 134 (17b) and for the
amine oxidation Pyr+• + amine ® amine-H+ at m/z 100 (TEA and
DIPA), m/z 128 (DIPEA) and m/z 98 (CHA). The second-order
rate coefficient k for formation of each individual product was
obtained from plotting k’obs vs. [amine], which is shown in Figures
S14 – S17 in the SI.
Experimental Methods
Mass Spectrometric Experiments. Gas-phase studies were
conducted on a Thermo Scientific (Bremen, Germany) LTQ ESI
mass spectrometer, modified to enable the introduction of volatile
neutral reagents into the linear ion trap to perform ion-molecule
reactions (IMRs). Briefly, a suitably volatile liquid reagent was
injected at a rate of 5 – 20 µL hour-1 directly into the helium line
supplying helium bath gas to the linear ion trap. The helium line
was heated and maintained at a temperature above the boiling
point of the reagent, volatilising the liquid before it reached the ion
trap. The pressure regulator controlling the flow of helium into the
ion trap under normal operating conditions was bypassed, with
the helium pressure controlled manually to maintain an ion gauge
pressure that was the same as under normal operating conditions
(0.69 × 10–5 Torr), which was confirmed by testing against the
relatively slow ion-molecule reaction of Br- with CH3I, as reported
previously.[37] Further details are given in refs.[6,8,38] Ions
undergoing ion-molecule reactions (IMRs) in the ion trap of the
mass spectrometer are quasi thermalized to the temperature of
the helium bath gas (298 K).[38a] When performing IMRs, an ion of
interest was isolated inside the ion trap, where it was allowed to
react with the neutral. Each neutral compound used in the IMRs
was purified prior to use via distillation, with purity checked by
NMR.
Acknowledgements
Support by the Australian Research Council (DP170100035 and
DP180101187) and the Australian Government through provision
of "Australian Government Research Training Program
Scholarships“ is gratefully acknowledged. The DFT calculations
were performed on the Spartan High-Performance Computing
(HPC) System hosted by Research Platform Services at the
University of Melbourne.[40] We thank the Bio21 Mass
Spectrometry and Proteomics Facility for access to the Thermo
OrbiTrap Fusion Lumos mass spectrometer and Lars Goerigk for
helpful discussions.
Keywords: DFT calculations
• distonic ions • gas-phase
chemistry • mass spectrometry • reaction mechanisms
Compound samples were prepared to 75-100 µM in methanol
and were injected into the ESI source at a flow rate of 5 µL min-1.
The instrument was operated in positive ion mode, with the
settings tuned to optimise the signal of the initial parent ion peak
at the beginning of each data collection session. ESI source
[1]
[2]
J. F. Bunnett, J. K. Kim, J. Am. Chem. Soc. 1970, 92, 25, 7463-7464.
Selected examples: a) N. Holmberg-Douglas, D. A. Nicewic, Org. Lett.
2019, 21, 7114; b) Photochemistry (Vol. 23), eds. D. Bryce-Smith, A.
Gilbert, The Royal Society of Chemistry, 1992; c) J. G. Nathanael, L. F.
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
Gamon, M. Cordes, P. R. Rablen, T. Bally, K. F. Fromm, B. Giese, U.
[13] Oxidation could be initiated by reaction of the amine, for example TEA,
•
Wille, ChemBioChem 2018, 19, 922.
with the radical cation through electron transfer (Pyr+ + NEt3 ® Pyr +
[3]
[4]
K. M. Stirk, L. K. M. Kiminkinen, H. I. Kenttämaa, Chem. Rev. 1992, 92,
1649-1665.
N+•Et3) or hydride transfer (Pyr+• + NEt3 ® PyrH• + Et2N+=CHCH3).
[14] Isolation of the products 15-17 in the ion trap revealed no further reaction
with the amine, and CID did not lead to any fragmentation, even at high
collision energies, confirming their stability.
Examples for the identification of early polymer degradation products: a)
J. R. Lindsay Smith, E. Nagatomi, A. Stead, D. J. Waddington, S. D.
Bévière, J. Chem. Soc., Perkin Trans. 2 2000, 1193-1198; b) J. R.
Lindsay Smith, E. Nagatomi, D. J. Waddington, J. Chem. Soc., Perkin
Trans. 2 2000, 2248-2258; c) J. R. Lindsay Smith, E. Nagatomi, A. Stead,
D. J. Waddington, J. Chem. Soc., Perkin Trans. 2 2001, 1527-1533.
Selected examples: a) A. K. Y. Lam, C. Li, G. N. Khairallah, B. B. Kirk,
S. J. Blanksby, A J. Trevitt, U. Wille, R. A. J. O’Hair, G. da Silva, Phys.
Chem. Chem. Phys. 2012, 14, 2417-2426; b) C. Li, A. K. Lam, G. N.
Khairallah, J. M. White, R. A. J. O’Hair, G. da Silva, J. Am. Soc. Mass
Spectrom. 2013, 24, 493-501; c) A. T. Maccarone, B. B. Kirk, C. S.
Hansen, T. M. Griffiths, S. Olsen, A. J. Trevitt, S. J. Blanksby, J. Am.
Chem. Soc. 2013, 135, 9010-9014; d) G. da Silva, B. B. Kirk, C. Lloyd,
A. J. Trevitt, S. J. Blanksby, J. Phys. Chem. Lett. 2012, 3, 805-811; e) B.
B. Kirk, D. G. Harman, H. I. Kenttämaa, A. J. Trevitt, S. J. Blanksby, Phys.
Chem. Chem. Phys. 2012, 14, 16719-16730.
[15] The gas-phase chemistry of related N-acylpyridinium ions has been
examined: R. A. J. O'Hair, N. K. Androutsopoulos, Org. Lett. 2000, 2,
2567-2570.
[16] P. Kuzmic, Anal. Biochem. 1996, 237, 260-273.
[5]
[17] Because of the significant side reactions of the amines, kinetic studies
could not be performed for Oxo+C•.
[18] a) K. F. Lim, Quantum Chem. Program Exch. Bull. 1994, 14, 3; b) T. Su,
M. T. Bowers, Int. J. Mass Spectrom. Ion Phys. 1973, 12, 347-356.
Dipole moments from: CRC Handbook of Chemistry and Physics, 82nd
ed. [electronic resource], editor-in-chief D. R. Lide, CRC Press Boca
Raton, Fla. 2001. Polarizabilities from: R. Bosque, J. Sales, J. Chem. Inf.
Comput. Sci. 2002, 42, 1154-1163.
[19] An alternative route to formation of the amine adducts 12 – 15 through
an unprecedented homolytic substitution on the N atom of the amine by
Pyr+• can be excluded, since hypervalent ammonium radicals of type
Me(x=1-3)NH(4-x)• are unstable and only metastable as deuterated species:
F. Turecek, Top. Curr. Chem. 2003, 225, 77-129 and cited references.
[20] a) Gaussian 09, Revision B.01, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg,
M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery,
Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N.
Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A.
Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M.
Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G.
Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A.
D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J.
Fox, Gaussian, Inc., Wallingford CT, 2009; b) Gaussian 16, Revision
B.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H.
Nakatsuji, X. Li, M. Caricato, A. V. Marenich, J. Bloino, B. G. Janesko, R.
Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L.
Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J.
Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G.
Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, T. Vreven, K. Throssell, J. A. Montgomery, Jr., J. E.
Peralta, F. Ogliaro, M. J. Bearpark, J. J. Heyd, E. N. Brothers, K. N.
Kudin, V. N. Staroverov, T. A. Keith, R. Kobayashi, J. Normand, K.
Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M.
Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R.
L. Martin, K. Morokuma, O. Farkas, J. B. Foresman, and D. J. Fox,
Gaussian, Inc., Wallingford CT, 2016.
[6]
[7]
a) C. H. Li, G. N. Khairallah, A. K. Y. Lam, R. A. J. O’Hair, B. B. Kirk, S.
J. Blanksby, G. da Silva, U. Wille, Chem. Asian J. 2013, 8, 450-464; b)
G. N. Khairallah, R. A. J. O’Hair, U. Wille, J. Phys. Chem. A 2014, 118,
3295-3306.
a) S. Osburn, B. Chan, V. Ryzhov, L. Radom, R. A. J. O’Hair, J. Phys.
Chem. A 2016, 120, 8184-8189; b) S. Wee, A. Mortimer, D. Moran, A.
Wright, C. K. Barlow, R. A. J. O’Hair, L. Radom, C. J. Easton, Chem.
Commun. 2006, 4233-4235; c) C. K. Barlow, A. Wright, C. J. Easton, R.
A. J. O’Hair, Org. Biomol. Chem. 2011, 9, 3733-3745; d) R. A. J. O'Hair,
J. Mass Spectrom. 2000, 35, 1377-1381.
[8]
[9]
a) B. Gervasoni, G. N. Khairallah, R. A. J. O’Hair, U. Wille, Phys. Chem.
Chem. Phys. 2015, 17, 9212-9221; b) B. I. Taggert, R. A. J. O’Hair, U.
Wille, J. Phys. Chem. A 2017, 121, 5290-5300.
The autoxidation mechanism used to rationalise radical degradation of
all polymers, regardless of composition, has originally been proposed for
the degradation of rubber, see: a) J. L. Bolland, Proc. R. Soc. London
Ser. A 1946, 186, 218-236; b) J. L. Bolland, G. Gee, Trans. Faraday Soc.
1946, 42, 236-243; c) J. L. Bolland, P. ten Have, Discuss. Faraday Soc.
1947, 2, 252-260; d) L. Bateman, Q. Rev. Chem. Soc. 1954, 8, 147-167.
[10] a) Y. Huang, L. Guler, J. Heidbrink, H. I. Kenttämaa, J. Am. Chem. Soc.
2005, 127, 3973-3978; b) A. Adeuya, J. M. Price, B. J. Jankiewicz, J. J.
Nash, H. I. Kenttämaa, J. Phys. Chem. A 2009, 113, 13663-13674; c) F.
Widaja, Z. Jin, J. J. Nash, H. I. Kenttämaa, J. Am. Chem. Soc. 2012, 134,
2085-2093; d) G. O. Pates, L. Guler, J. J. Nash, H. I. Kenttämaa, J. Am.
Chem. Soc. 2011, 133, 9331-9342; e) B. J. Jankiewicz, J. Gao, J. N.
Reece, N. R. Vinueza, P. Narra, J. J. Nash, H. I. Kenttämaa, J. Phys.
Chem. A 2012, 116, 3089-3093; f) F. Widaja, Z. Jin, J. J. Nash, H. I.
Kenttämaa, J. Am. Soc. Mass Spectrom. 2013, 24, 469-480; g) P. E.
Williams, B. J. Jankiewicz, L. Yang, H. I. Kenttämaa, Chem. Rev. 2013,
113, 6949-6985.
[11] a) D. R. Bauer, Appl. Polym. Sci. 1982, 27, 3651-3662; b) C. Gao, S.
Moya, H. Lichtenfeld, A. Casoli, H. Fiedler, E. Donath, H. Möhwald,
Macromol. Mater. Eng. 2001, 286, 355-361.
[12] In the gas phase, canonical radical cations of nucleobases and
nucleosides react with related amines via a range of reactions including
proton transfer, electron transfer, HAT, addition reactions and addition-
fragmentation reactions: a) L. Feketeová, G. N. Khairallah, B. Chan, V.
Steinmetz, P. Maître, L. Radom, R. A. J. O’Hair, Chem. Commun. 2013,
49, 7343–7345; b) L. Feketeová, B. Chan, G. N. Khairallah, V. Steinmetz,
P. Maitre, L. Radom, R. A. J. O’Hair, Phys. Chem. Chem. Phys. 2015,
17, 25837-25844; c) L. Feketeová, B. Chan, G. N. Khairallah, V.
Steinmetz, P. Maitre, L. Radom, R. A. J. O’Hair, J. Phys. Chem. Lett.
2017, 8, 3159–3165.
[21] Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215-241.
[22] All ground and transition states were verified by vibrational frequency
analysis at the same level of theory, and all identified transition states
showed only one imaginary frequency. The spin expectation value, <s2>,
was very close to 0.75 after spin annihilation, except for the adduct
complex B – O2 shown in Scheme 6, for which <s2> was 0.8.
[23] P. D. Dau, P. B. Armentrout, M. C. Michelinic, J. K. Gibson, Phys. Chem.
Chem. Phys. 2016, 18, 7334-7340.
[24] Because of the low pressure in the ion trap of the mass spectrometer,
energy exchange through collisions with the surroundings can be
excluded. Thus, the reactant complex has excess energy resulting from
the internal and kinetic energy of the free reactants and the electrostatic
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
energy that is released upon complex formation. Any subsequent
studied in this work. Since the focus in this work is to obtain a qualitative
understanding of the interactions between charge and spin in these
reactions, single reference calculations provide a useful approximation.
[31] CRC Handbook of Chemistry and Physics, 65th ed., (Eds. R. C. Weast,
M. J. Astle, W. H. Beyer, CRC Press, Boca Raton, 1984.
reaction of the reactant complex is therefore only possible, if the
activation barrier for this process is smaller or equal to the energy
difference between the entrance channel and the association complex;
see: a) C. H. DePuy, J. Org. Chem. 2002, 67, 2393-2401; b) M.
Speranza, Mass Spectrom. Rev. 1992, 11, 73-117.
[32] The ground state of a cyclic amino substituted dioxolane K could not be
located with UHF or DFT methods, clearly indicating that this framework
is not stable.
[25] The optimised geometries of selected reactant complexes A are given in
Figure S18. In complex [Å···4-Pyr•-TEA] the N atom of TEA points
towards the positive point charge, leading to significant stabilisation. It
was not possible to locate a ground state geometry for [Å···4-Pyr•-TEA]
that was similar to the other reactant complexes.
[33] For recent mechanism-based studies where gas-phase reactivity has
been translated into new solution methods for organic synthesis, see: a)
A. Noor, J. Li, G. N. Khairallah, Z. Li, H. Ghari, A. J. Canty, A. Ariafard,
P. S. Donnelly, R. A. J O’Hair, Chem. Commun. 2017, 53, 3854-3857; b)
Y. Yang, A. Noor, A. J. Canty, A. Ariafard, P. S. Donnelly, R. A. J O’Hair,
Organometallics 2019,38, 424-435.
[26] The transition state for the addition of amines via N could only be located
with UHF but not with DFT methods, such as BHandHLYP or M062X.
[27] Selected examples: a) M. Klinska, L. M. Smith, G. Gryn’ova, M. G.
Banwell, M. L. Coote, Chem. Sci. 2015, 6, 5623-5627; b) A. C. Aragonès,
N. L. Haworth, N. Darwish, S. Ciampi, N. J. Bloomfield, G. G. Wallace, I.
Diez-Perez, M. L. Coote, Nature 2016, 531, 88−91; c) R. Meir, H. Chen,
W. Lai, S. Shaik, ChemPhysChem 2010, 11, 301−310; d) W. Lai, H.
Chen, K.-B. Cho, S. Shaik, J. Phys. Chem. Lett. 2010, 1, 2082−2087; e)
H. Hirao, H. Chen, M. A. Carvajal, Y. Wang, S. Shaik, J. Am. Chem. Soc.
2008, 130, 3319−3327; f) S. Shaik, S. P. de Visser, D. Kumar, J. Am.
Chem. Soc. 2004, 126, 11746−11749; g) C. Geng, J. Li, T. Weiske, M.
Schlangen, S. Shaik, H. Schwarz, J. Am. Chem. Soc. 2017, 139, 1684-
1689; h) H. Schwarz, S. Shaik, J. Li, J. Am. Chem. Soc. 2017, 139,
17201-17212; i) L. Yue, N. Wang, S. Zhou, X. Sun, M. Schlangen, H.
Schwarz, Angew. Chem. Int. Ed. 2018, 57, 14635-14639; and references
therein.
[34] Y. Takano, K. N. Houk, J. Chem. Theory Comput. 2005, 1, 70-77.
[35] It should be noted that the same outcome was obtained when the
reaction was performed in a degassed solution under Ar or in an O2
stream, suggesting that very small amounts of O2 are sufficient for the
reaction to proceed. Further mechanistic studies on the termination step
are clearly required.
[36] Performing the reaction of 4-Pyr+I (7a) with amines or methanol in an
Accelerated Weathering Tester (Q-LAB QUV-se, using UVA-340
fluorescent lamps) at 50oC for 6 hours led to increased amounts of
substitution products, indicating that higher temperatures are beneficial
for the reaction outcome.
[37] a) S. Gronert, C. H. DePuy, v. M. Bierbaum, J. Am. Chem. Soc.1991,113,
4009; b) S. C. Brydon, Z. Ren, G. da Silva, S. F. Lim, G. N. Khairallah,
M. J. Rathjen, J. M. White, R. A. J. O’Hair, J. Phys. Chem. A 2019, 123,
8200.
[28] NBO analyses performed with M062X/6-31+G* (large degree of Fock-
exchange) and wB97x/6-31+G* (range-separated hybrid; see: J.-D.
Chai, M. Head-Gordon, J. Chem. Phys. 2008, 128, 084106) gave
essentially identical results.
[38] a) W. A. Donald, G. N. Khairallah, R. A. J. O’Hair, J. Am. Soc. Mass
Spectrom. 2013, 24, 811-815; b) W. A. Donald, C. J. McKenzie, R. A. J.
O’Hair Angew. Chem. Int. Ed. 2011, 50, 8379-8383.
[29] Population analyses confirmed that the SOMO is energetically below the
MOs of the aromatic p system also in the s-radicals Pyr• and the phenyl
radical Phe•.
[39] T. Waters, R. A. J. O’Hair, A. G. Wedd, J. Am. Chem. Soc. 2003, 125,
3384.
[30] Analysis of Fractional Occupancy Density (FOD) plots (see: S. Grimme,
A. Hansen, Angew. Chem. Int. Ed. 2015, 12, 12308) suggest that some
intermediates where mixing of orbitals with excited states occurs, might
require treatment by multireference methods. Unfortunately, such
computations are too expensive for the comparably large systems
[40] L. Lafayette, G. Sauter, L. Vu, B. Meade, Spartan Performance and
Flexibility: An HPC-Cloud Chimera. OpenStack Summit, Barcelona,
October 27, 2016.
This article is protected by copyright. All rights reserved.

10.1002/cplu.201900706
ChemPlusChem
FULL PAPER
Entry for the Table of Contents
FULL PAPER
SRN1-type
Working
together!
Nucleophilic
NEt3
N
Benjamin Andrikopoulos, Parvinder K.
Sidhu, Bethany I. Taggert, Joses G.
Nathanael, Richard A. J. O’Hair*, Uta
Wille*
addition of amines to aromatic
distonic radical cations Pyr+ is
Et3N
•
NEt3
N
N
TS1
reactant
complex
facilitated an interplay between spin
and positive charge, which direct
amine attack to the site of the
unpaired electron. This synergism
makes the reaction energetically
more favourable than the related
reactions in the absence of spin
(SNAr-type reaction) or charge
(SRN1-type reaction).
32.7
0.0
16.4
-14.6
TS1
-25.7
-28.1
Et3N
Page No. – Page No.
SNAr-type
-43.3
Et3N
-56.3
N
Reaction of Distonic Aryl and Alkyl
Radical Cations with Amines: The
Role of Charge and Spin Revealed
by Mass Spectrometry, Kinetic
Studies and DFT Calculations
reactant
complex
N
reactant
complex
No barrier!
Strongly exothermic!
-188.4
NEt3
N
This article is protected by copyright. All rights reserved.