Environmental Polymer Degradation: Using the Distonic Radical Ion Approach to Study the Gas-Phase Reactions of Model Polyester Radicals

Article abstract of DOI:10.1021/acs.jpca.7b04217

A novel precursor to the distonic O- and C-centered radical cations Oxo⁺O^? and Oxo⁺C^? was designed and synthesized, which represents model systems for radicals produced during polyester degradation. The precursor is equipped with a nitrate functional group, which serves as a masked site for these alkoxyl and carbon radicals that are unleashed through collision-induced dissociation (CID). Oxo⁺O^? and Oxo⁺C^? feature a cyclic carboxonium ion as permanent charge tag to enable monitoring their ion-molecule reactions on the millisecond to second time scale in the ion trap of the mass spectrometer. The reactions of Oxo⁺O^? and Oxo⁺C^? with cyclohexene, cyclohexadiene, ethyl acetate, 1,1-dimethoxyethane, and 1,2-dimethoxyethane, which exhibit structural features present in both intact and defective polyesters, were explored through product and kinetic studies to identify "hot spots" for radical-induced damage in polyesters. All reactions with Oxo⁺O^? were extremely fast and proceeded predominantly through HAT. Oxo⁺C^? was about two orders of magnitude less reactive and did not noticeably damage aliphatic ester moieties through hydrogen abstraction on the time scale of our experiments. Radical addition to alkene π systems was identified as an important pathway for C-radicals, which needs to be included in polymer degradation mechanisms.

Full text of DOI:10.1021/acs.jpca.7b04217

Subscriber access provided by University of Newcastle, Australia
Article
Environmental Polymer Degradation: Using the Distonic Radical Ion
Approach to Study the Gas-Phase Reactions of Model Polyester Radicals
Bethany I. Taggert, Richard A. J. O'Hair, and Uta Wille
J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b04217 • Publication Date (Web): 23 Jun 2017
Downloaded from http://pubs.acs.org on June 26, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Physical Chemistry A is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 41
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
Environmental Polymer Degradation: Using the
Distonic Radical Ion Approach to Study the Gasꢀ
Phase Reactions of Model Polyester Radicals
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bethany I. Taggert, Richard A. J. O’Hair, Uta Wille*
School of Chemistry, Bio21 Institute, The University of Melbourne, 30 Flemington Road,
Parkville, VIC 3010, Australia.
Abstract
A novel precursor to the distonic Oꢀ and Cꢀcentred radical cations Oxo+O and Oxo+C
• • was
designed and synthesized, which represent model systems for radicals produced during polyester
degradation. The precursor is equipped with a nitrate functional group, which serves as a masked
site for these alkoxyl and carbon radicals that are unleashed through collisionꢀinduced
dissociation (CID). Oxo+O and Oxo+C
• • feature a cyclic carboxonium ion as permanent charge
tag to enable monitoring their ionꢀmolecule reactions on the millisecond to second timescale in
the ion trap of the mass spectrometer. The reactions of Oxo+O and Oxo+C
• with cyclohexene,
•
ACS Paragon Plus Environment
1

The Journal of Physical Chemistry
Page 2 of 41
1
2
3
4
5
6
7
8
9
cyclohexadiene, ethyl acetate, 1,1ꢀdimethoxyethane and 1,2ꢀdimethoxyethane, which exhibit
structural features present in both intact and defective polyesters, were explored through product
and kinetic studies to identify ‘hot spots’ for radicalꢀinduced damage in polyesters. All reactions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
with Oxo+O were extremely fast and proceeded predominantly through HAT. Oxo+C
• • was
about two orders of magnitude less reactive and did not noticeably damage aliphatic ester
moieties through hydrogen abstraction on the timescale of our experiments. Radical addition to
alkene π systems was identified as an important pathway for Cꢀradicals, which needs to be
included in polymer degradation mechanisms.
Introduction
Synthetic polymers are amongst the most widely used materials due to their extremely diverse
properties, which are controlled by their composition, the synthetic and processing techniques as
well as their chain length, in conjunction with any secondary bonding interactions between
polymer molecules. Environmental exposure of polymers leads to dramatic changes of their
properties resulting in decreased service life and limited usage.
Environmental polymer degradation can occur hydrolytically, thermally, and by UVꢀlight.1
Both thermal and photochemical degradation can proceed through radical pathways. Key
intermediates are polymerꢀderived peroxyl radicals, ROO•, resulting from reaction of a carbonꢀ
centred polymer radical, R
•
(which can be formed, for example, through mechanical stress),2
with oxygen, O2. ROO
damage through abstraction of a hydrogen atom from a neighbouring polymer strand to form a
hydroperoxide, ROOH, and a new polymer radical R . Although this ‘autoxidation’ mechanism
• are believed to act as carrier in a radical chain process that propagates
•
ACS Paragon Plus Environment
2

Page 3 of 41
The Journal of Physical Chemistry
1
2
was originally only proposed for the degradation of rubber,3ꢀ6 it has been generally adapted to
rationalize radical degradation in any polymer, irrespective of their structure. However, a recent
highꢀlevel computational study clearly showed that this radical propagation is energetically
unfavourable in polymers without activated C–H bonds,7 which suggests that important
pathways in environmental polymer degradation remain unrecognized. In fact, early degradation
products, which could shed light onto the mechanism(s), have been rarely identified.8ꢀ10
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Since reactions between O2 and polymer radicals R• occur at the interface between gasꢀphase
and polymer surface, gasꢀphase physical organic chemistry studies should provide valuable
insights into reactive intermediates associated with the formation of early degradation products.
A general way of investigating gasꢀphase radical reactions is by taking advantage of the
multistage mass spectrometry capability of ion traps to generate distonic radical ions and then
study their reactivity.11 The chemistry of any charged species in the reaction system can be
directly monitored in a time resolved fashion to obtain kinetic data for individual reaction steps.
This approach has been used to investigate radical reactions of relevance to numerous fields, for
example organic synthesis,12ꢀ13 atmospheric chemistry,14ꢀ18 and biological chemistry.19,20 We
have recently employed this methodology to study the reaction of distonic Nꢀmethylpyridinium
peroxyl radical cations (Pyr+OO
various substrates that exhibit typical structural motifs in intact or defective polyesters.21 Scheme
1 outlines the reaction of Pyr+OO
• •, with
),12,13 which were used as a simplified model for ROO
•
with cyclohexene, which shows that allylic hydrogen atom
transfer (HAT; pathway (a)) is only a minor pathway.
ACS Paragon Plus Environment
3

The Journal of Physical Chemistry
Page 4 of 41
1
2
Scheme 1. Reaction of Pyr+OO with cyclohexene.21
•
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The major reaction channel involves addition of Pyr+OO
rapid homolytic O–O bond fragmentation to yield an epoxide and Pyr+O
•
to the alkene π system, followed by
•
(pathway (b)), which is
a reactive oxygenꢀcentred radical species (RO
addition.21 Formation of highly reactive RO
•
) that reacts rapidly through HAT and radical
•
through reaction of ROO with π systems in
•
alkenes has not been considered as pathway in radical polymer degradation so far. This not only
highlights the deficiencies of the current mechanistic rationale, but also demonstrates the power
of the distonic radical ion approach, which allows the discovery of new, previously unidentified
degradation pathways.
In this work, we have designed and synthesised the novel distonic radical cation precursor (1)
to perform a mass spectrometric study of the reaction kinetics and products of two aliphatic
alkoxyl and alkyl radicals, which mimic important radical intermediates formed during polyester
degradation. This precursor contains glycerol monobenzoate as simplified model for the
phthalate structures typically seen in polyesters, and an Nꢀprotonated glycine moiety, which acts
as a temporary charge tag to enable its isolation in the ion trap of the mass spectrometer (Scheme
2).
ACS Paragon Plus Environment
4

Page 5 of 41
The Journal of Physical Chemistry
1
2
Scheme 2. Generation of the distonic radical cations Oxo+O and Oxo+C
• • in the mass
3
4
5
6
spectrometer and the ionꢀmolecule reactions studied in this work.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The nitrate group in 1 serves as a masked radical site that is unleashed through collisionꢀ
induced dissociation (CID) of the labile O–NO2 bond.22,23 Upon CID, first loss of glycine occurs
through a neighbouring group reaction24,25 to produce nitrate 2 at m/z 224, which features a
cyclic tertiary carboxonium ion26ꢀ28 as the fixed charge tag. A second stage of CID generates the
desired alkoxyl radical cation Oxo+O
of formaldehyde produces the Cꢀcentred radical cation Oxo+C
The chemistry of Oxo+O and Oxo+C
• was explored through their ionꢀmolecule reactions with
•
at m/z 178. Subsequent βꢀfragmentation with elimination
•
at m/z 148.
•
cyclohexene (CH), cyclohexadiene (CHD), ethyl acetate (EA), 1,1ꢀdimethoxyethane (1,1ꢀDME)
and 1,2ꢀdimethoxyethane (1,2ꢀDME), which exhibit important structural features present in both
intact and defective polyesters and are sufficiently volatile model systems to be studied in the gas
phase. In particular, CH and CHD provide detailed insight into the reactivity of alkene π systems
ACS Paragon Plus Environment
5

The Journal of Physical Chemistry
Page 6 of 41
1
2
3
4
5
6
7
8
9
that could be formed through thermal decomposition of polyesters via ester pyrolysis during the
manufacturing process.29 EA enables exploration of the reactivity of aliphatic ester linkages
towards radical damage, whereas 1,1ꢀ and 1,2ꢀDME represent different glycol motifs. For
selected reactions, the experiments were augmented by density functional theory (DFT)
calculations of the potential energy surface to identify the regioselectivity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Methods
Synthesis of starting materials. Reactions were monitored using thin layer chromatography
(tlc) on commercial silica gel 60 aluminiumꢀbacked plates coated with fluorescent indicator F254
(Merck). Plates were visualised using UV light (254 nm) or in conjunction with ninhydrinꢀ,
potassium permanganateꢀ or phosphomolybdic acidꢀbased stains. Purification by silica gel
chromatography was performed using Davisil Chromatographic Silica Media LC60A 40ꢀ63
micron. NMR spectra were recorded on a Varian INOVA 400 instrument, with operating
1
frequencies of 400 MHz for H NMR and 101 MHz for 13C NMR. Chemical shifts (δ) are in
ppm, with residual undeuterated solvent peaks used as an internal reference. High resolution
mass spectra (HRMS) were collected via electrospray ionisation (ESI) mass spectrometry, using
either a Thermo Scientific Exactive Plus Orbitrap mass spectrometer or a Thermo hybrid LTQꢀ
FTICR mass spectrometer (Thermo Scientific, Bremen, Germany).
The precursor 1 was synthesized as trifluoroacetate (TFA) salt according to Scheme 3. The
NMR spectra are shown in the Supporting Information (SI).
ACS Paragon Plus Environment
6

Page 7 of 41
The Journal of Physical Chemistry
1
2
3
4
Scheme 3. Synthesis of the distonic radical cation precursor 1.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1-Benzoyl-3-bromo-rac-glycerol (6). The procedure reported by Dawe et al.30 was followed to
synthesise Li2NiBr4 as a 0.4 M solution in tetrahydrofuran. Epoxide 531 (24 mmol) was dissolved
in tetrahydrofuran (71 mL) under nitrogen and cooled to 0°C, followed by dropwise addition of
the Li2NiBr4 solution (76 mL, ~30 mmol). Consumption of the epoxide was monitored by tlc and
complete after 35 mins. The reaction mixture was then treated with phosphate buffer (110 mL,
pH 7) and extracted with dichloromethane, dried and concentrated in vacuo. The crude product
was purified by column chromatography (petroleum spirits:ethyl acetate 3:1, Rf = 0.34) to
1
provide 6 as a white powder (82%). H NMR (400 MHz, CDCl3): δ = 8.08 – 8.02 (m, 2H,
3
aromatic), 7.62 – 7.56 (m, 1H), 7.49 – 7.42 (m, 2H), 4.53 – 4.43 (m, 2H), 4.21 (h, J = 5.2 Hz,
ACS Paragon Plus Environment
7

The Journal of Physical Chemistry
Page 8 of 41
1
2
3
3
13
3
4
5
6
7
8
9
1H), 3.61 (dd, J = 10.5, 4.9 Hz, 1H), 3.55 (dd, J = 10.6, 5.9 Hz, 1H), 2.66 (s, 1H); C NMR
(101 MHz, CDCl3): δ = 166.58, 133.40, 129.72, 129.43, 128.49, 69.41, 66.32, 35.08; HRMS
(ESI, m/z): [M+H]+ calcd for C10H12O3Br, 258.99643 and 260.99448; found 258.99639 and
260.99436.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2-Hydroxy-3-nitrooxypropyl benzoate (7). 6 (6.5 mmol) was suspended in dry acetonitrile (12
mL) under nitrogen along with silver nitrate (12 mmol). The mixture was kept at 50°C in the
dark for 7 d, with regular filtration to remove the silver bromide precipitate. Once complete, the
mixture was cooled to room temperature and filtered into saturated brine (40 mL). The resulting
silver chloride precipitate was separated and the filtrate extracted with ethyl acetate, dried
(magnesium sulfate), and concentrated in vacuo. The crude material was purified by column
chromatography (petroleum spirits:ethyl acetate 1:1, Rf = 0.65) to provide 7 as a waxy solid
1
(29%). H NMR (400 MHz, CDCl3): δ = 8.07 – 7.96 (m, 2H), 7.63 – 7.53 (m, 1H), 7.49 – 7.40
(m, 2H), 5.42 – 5.36 (m, 1H), 4.87 (ddd, 3J = 12.3, 3.7, 0.6 Hz, 1H), 4.78 (ddd, 3J = 12.4, 6.4, 0.6
Hz, 1H), 3.99 – 3.93 (m, 1H), 3.93 – 3.87 (m, 1H), 2.26 (s, 1H); 13C NMR (101 MHz, CDCl3): δ
= 165.87, 133.57, 129.79, 129.15, 128.50, 71.27, 70.66, 61.20. HRMS (ESI, m/z): [M+H]+ calcd
for C10H12NO6, 242.06591; found 242.06586; [M+Na]+ calcd for C10H11NO6Na, 264.04786;
found 264.04802.
2-(2-(tert-Butoxycarbonylamino)acetoxy)-3-nitrooxypropyl benzoate (8). 8 was synthesised
from 7 (1 equiv.), NꢀBocꢀprotected glycine (1.5 equiv.), dimethylaminopyridine (DMAP; 0.1
equiv.) and N,Nꢀdicyclohexylcarbodiimide (DCC; 1.5 equiv.), following the “General
Esterification Procedure” (see Supporting Information). 8 was obtained as a white solid (95%),
which was used without further purification. 1H NMR (400 MHz, CDCl3): δ = 8.05 – 7.98 (m,
2H), 7.63 – 7.56 (m, 1H), 7.51 – 7.42 (m, 2H), 5.52 (p, 3J = 4.9 Hz, 1H), 4.98 (s, 1H), 4.82 (dd,
ACS Paragon Plus Environment
8

Page 9 of 41
The Journal of Physical Chemistry
1
2
3
3J = 12.5, 3.8 Hz, 1H), 4.69 – 4.62 (m, 1H), 4.62 – 4.56 (m, 1H), 4.46 (dd, J = 12.2, 5.3 Hz,
3
4
5
6
7
8
3
13
1H), 3.97 (d, J = 5.8 Hz, 2H), 1.44 (s, 9H); C NMR (101 MHz, CDCl3): δ = 165.86, 157.38
155.55, 133.55, 129.73, 129.04, 128.57, 79.66, 70.12, 68.74, 62.08, 42.33, 28.25. HRMS (ESI,
m/z): [M+H]+ calcd for C17H23N2O9, 399.13981; found 399.13973; [M+Na]+ calcd for
C17H22N2O9Na, 421.12175; found 421.12177.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2-(1-Benzoyloxy-3-nitrooxy)propan-2-yloxy-2-oxoethanaminium (1). 8 was dissolved in
dichloromethane at room temperature, and cooled to 0°C. With stirring, trifluoroacetic acid
(TFA; 6 equiv.) was added dropwise, and the mixture was stirred at room temperature and
monitored by tlc. When consumption of 8 was complete, the solvent was evaporated under
vacuum, using toluene as an azeotrope to remove any remaining acid. The resulting TFA salt
1
was directly used in the mass spectrometric experiments without further purification. H NMR
(400 MHz, DMSOꢀd6): δ = 8.23 (s, 3H), 8.01 – 7.94 (m, 2H), 7.73 – 7.64 (m, 1H), 7.58 – 7.49
3
3
(m, 2H), 5.59 (ddd, J = 9.7, 6.5, 3.2 Hz, 1H), 4.95 (dd, J = 12.4, 3.1 Hz, 1H), 4.83 (dd, J =
12.4, 7.0 Hz, 1H), 4.57 (dd, 3J = 12.2, 3.8 Hz, 1H), 4.46 (dd, J = 12.3, 5.8 Hz, 1H), 3.90 (s, 2H);
HRMS (ESI, m/z): [M+H]+ calcd for C12H15N2O7, 299.08738; found 299.08727.
Mass Spectrometric Experiments. Gasꢀphase 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 ionꢀmolecule reactions (IMRs). Briefly, a
suitably volatile liquid reagent was injected at a rate of 5 – 50 µL minꢀ1 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
ACS Paragon Plus Environment
9

The Journal of Physical Chemistry
Page 10 of 41
1
2
ion gauge pressure the same as under normal operating conditions (0.69 × 10–5 Torr). Further
details are given in refs.12,13,32,33 Ions undergoing ionꢀmolecule reactions in the ion trap of the
mass spectrometer are quasi thermalized to the temperature of the helium bath gas (298 K).32
When performing IMRs, an ion of interest was isolated inside the ion trap, where it could react
with the neutral. Each neutral compound used in the IMRs was purified prior to use via
distillation, with purity checked by NMR.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
General Mass Spectrometer Settings. Compound samples were prepared to 75ꢀ100 µM in
methanol, and were injected into the ESI source at a flow rate of 5 µL minꢀ1. 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 conditions included needle
potentials between 2.5ꢀ4.7 kV, a capillary temperature of 250°C, capillary voltages between 0.0ꢀ
34.0, and tube lens voltages between 0.0ꢀ55.0 V. The Helium bath gas was adjusted to maintain
an ion gauge pressure of 0.69 × 10ꢀ5 Torr. Ions of interest for MSn experiments were isolated
with a 2ꢀ5 m/z window. The CID parameter was selected so the parent ion being fragmented was
reduced to ~15% abundance. Data was collected using three microꢀscans and taking between 10ꢀ
100 duplicate spectra.
Kinetic Studies. Absolute rate coefficients for the consumption of Oxo+O
reaction with the neutrals CH, CHD, EA, 1,1ꢀDME and 1,2ꢀDME according to: Oxo+O
neutral → products, were obtained at 298 K by monitoring the decay of the Oxo+O
+) as a function of reaction time under pseudoꢀfirst order
•
/C
•
through
/C
signal at
•
•
+
•
/C•
m/z 178 and m/z 148, respectively ([M]
•
conditions. Four to seven different excess concentrations of the neutral were used on time scales
that allowed the peak intensity of Oxo+O
/C• to decrease to ca. 20% and to build up timeꢀ
•
resolved decay profiles (8ꢀ17 reaction times). The pseudoꢀfirst order rate coefficient kobs was
ACS Paragon Plus Environment
10

Page 11 of 41
The Journal of Physical Chemistry
1
2
obtained from the slope of the semilogarithmic plot ln[Oxo+O
•/C•] vs. reaction time. The
3
4
5
6
7
8
secondꢀorder rate coefficient k was determined by plotting kobs vs. [neutral], see Figure S1 in the
SI. Since the reactions of Oxo+O
•
with both CHD and 1,2ꢀDME were extremely fast, kobs could
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
only be obtained for a narrow substrate concentration range. The second order rate coefficients
were therefore determined for each concentrationꢀtime profile according to kobs/[CHD or 1,2ꢀ
DME], followed by averaging the resulting data.
The rate coefficients for product formation, i.e., Oxo+OH/CH and the radical adducts 11a,b
and 21a,b were obtained from kinetic modelling of the experimental concentration–time profiles
using the program Dynafit 4.34 The concentration of the various ionic species was directly
reflected by their peak areas, since mass discrimination could be excluded. The pseudoꢀfirst
order rate coefficients for product formation, k’obs, were fitted for the HAT pathway: Oxo+O
+ neutral → Oxo+OH/CH + [neutralꢀH]
Oxo+O addition pathway: Oxo+O
(11a) and m/z 258 (11b); and [Oxo+C–neutral adduct]
•/C•
•
, at m/z 179 and m/z 149, respectively, and for the
•/C
•
•
/C
•
+ neutral → [Oxo+O–neutral adduct]
•, at m/z 260
•
at m/z 230 (21a) and m/z 228 (21b). The
secondꢀorder rate coefficient k’ for formation of each individual product was obtained from
plotting k’obs vs. [neutral], except for the lowꢀyielding adducts 11b and 21b in the reaction of
CHD with Oxo+O and Oxo+C
• •, respectively, where k’ was determined from the average of the
individual concentrationꢀtime profiles, i.e., k’obs/[CHD]. It was necessary to extend the kinetic
scheme for simulating the rate coefficients for formation of the products Oxo+CH and [Oxo+C–
neutral adduct]
•
by the reaction: Oxo+C
•
→ PhC≡O+, in order to account for the competing
unimolecular decomposition of Oxo+C
•. The data are shown in Figures S2 – S5 in the SI.
Computational Studies. DFT calculations were carried out utilizing the Gaussian 09
program35 at the M062X/6ꢀ31+G* level of theory. All ground and transition states were verified
ACS Paragon Plus Environment
11

The Journal of Physical Chemistry
Page 12 of 41
1
2
3
4
5
6
7
8
9
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. All energies are given as enthalpies, which have been shown to best
describe chemical reactivity under lowꢀpressure bimolecular conditions.36 We also explored the
BHandHLYP/6ꢀ31+G* method, but found that it calculates higher energies for the stationary
points in these gasꢀphase reactions than M062X/6ꢀ31+G*. In some cases, the BHandHLYP
computed transition states even indicated slow reactions, in contradiction with the experimental
findings. The M062X/6ꢀ31+G* data were considered sufficiently precise to obtain qualitative
information about radical reactivity and selectivity for the different pathways. The Gaussian 09
archive entries for all M062X/6ꢀ31+G* optimized ground and transition states are given in the
SI.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and Discussion
1. Mass spectrometry experiments
Figure 1 shows the CID mass spectrum of the distonic radical cation precursor 1 (see
Experimental Methods for details). As outlined in Scheme 2, the initial CID process produces
nitrate 2 at m/z 224 as the major product, which results from loss of glycine (Figure 1a). DFT
calculations revealed that the positive charge in 2 is stabilized through formation of a cyclic
carboxonium ion,26ꢀ28 suggesting that the cleavage of the glycine moiety is facilitated by
neighbouring group effects.24,25 Attempts to locate a stable ground state structure with an
uncyclised carbocation by DFT were not successful.
ACS Paragon Plus Environment
12

Page 13 of 41
The Journal of Physical Chemistry
1
2
3
4
5
6
7
(a)
O
ONO2
Ph
O
2
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
*
O
Ph
(b)
O
Oxo+C
O
O
Ph
O
Oxo+O
Ph
O
*
Figure 1. Mass spectra resulting from (a) CID of precursor 1 (m/z 299) to produce cyclic nitrate
2 at m/z 224; (b) CID of 2 leads to the distonic radical cations Oxo+O• (m/z 178) and Oxo+C•
(m/z 148). The massꢀselected precursor ion is indicated by an asterisk (*).
Isolation of 2 in the ion trap and CID leads to cleavage of the O–NO2 bond to form the alkoxyl
radical Oxo+O
Oxo+O produces the carbonꢀcentred radical Oxo+C
to the benzoyl cation PhC≡O+ at m/z 105 also occurred, which likely results from a sequential
ringꢀopening/dissociation process in Oxo+C
• at m/z 178 (Figure 1b). Subsequent spontaneous or CID induced fragmentation of
•
• (m/z 148). In addition to this, fragmentation
•
via an intermediate radical cation 9 (Scheme 4).
ACS Paragon Plus Environment
13

The Journal of Physical Chemistry
Page 14 of 41
1
2
3
4
Scheme 4. Proposed mechanism of formation of PhC≡O+ at m/z 105 through ringꢀ
5
6
7
opening/fragmentation of Oxo+C
•.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The minor product at m/z 123 in Figure 1b could result from a secondary reaction of PhC≡O+
with residual water present in the ion trap, in particular, since CID of the isolated m/z 123 species
produced only PhC≡O+. However, isolation of PhC≡O+ in the ion trap revealed no reaction even
after an exposure time of ten seconds, which suggests that m/z 123 is a direct fragmentation
product.
The site of the unpaired electron in Oxo+O and Oxo+C
• • was confirmed by mass spectrometry
through their reaction with O2, using Helium bath gas that contained 0.001% O2. Reaction with
Oxo+O
from abstraction of a αꢀhydrogen atom (Scheme 5).37
•
led to formation of a product at m/z 177 in accordance with aldehyde 10, which results
ACS Paragon Plus Environment
14

Page 15 of 41
The Journal of Physical Chemistry
1
2
3
4
Scheme 5. Experimental and computational analysis of the location of the unpaired spin and
5
6
charge in Oxo+O and Oxo+C
• •. Spin densities and charges (in square brackets) calculated with
7
8
9
M062X/6ꢀ31+G*.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The reaction of Oxo+C
•
with O2 led to a new product with m/z 180, which can be assigned to
the peroxyl radical cation Oxo+OO
•.
In fact, in some systems studied in this work
12,13,16ꢀ18,21
secondary reactions involving Oxo+OO
•
also occurred. We will report on the chemistry of such
aliphatic distonic peroxyl radical cations in a separate paper.
According to DFT calculations (see below), the unpaired electron in Oxo+O
• is located
practically exclusively (94%) on the oxygen atom, confirming its nature as Oꢀcentred radical.
The spin density in Oxo+C
•
is also predominantly (89%) located on the expected carbon atom
with some delocalisation onto the benzylic carbon atom and the phenyl ring. Analysis of the
Mulliken charges further revealed that charge and spin are not fully separated and that the radical
centre in Oxo+C
positive charge.
•
and the carbon atom adjacent to the radical centre in Oxo+O
• exhibit a partial
ACS Paragon Plus Environment
15

The Journal of Physical Chemistry
Page 16 of 41
1
2
3
4
5
6
7
8
9
It should be noted that we have also designed and synthesized a series of other precursor
molecules structurally related to 1, but in the mass spectrometer none of these underwent CID to
produce distonic Oꢀ and Cꢀcentred radical cations (see SI for details). This shows that seemingly
small structural differences have a profound and difficultꢀtoꢀpredict impact on the molecule’s
behaviour upon CID, making the design of suitable radical precursors a challenge.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.1 Ion-molecule reactions with Oxo+O
The reaction of Oxo+O
•
•
with CH was found to be very fast with the peak intensity dropping to
~20% after a maximum of 400 ms. The major pathway occurs through HAT, likely from the
allylic CH2 moiety, to give Oxo+OH at m/z 179 (Figure 2a). The corresponding αꢀcyclohexenyl
radical (not shown) is not charged and cannot be detected.
ACS Paragon Plus Environment
16

Page 17 of 41
The Journal of Physical Chemistry
1
2
3
4
5
(a)
O
OH
Ph
O
6
Oxo+OH
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Oxo+O
Oxo+C
Ph
O
O
*
O
11a
(b)
Oxo+OH
Ph
Oxo+O
Oxo+C
O
O
O
11b
*
m/z 258
(traces)
Figure 2. Mass spectrum of the reaction of Oxo+O with (a) CH after 280 ms ([CH] ca. 7.0 × 109
•
molecules cmꢀ3) and with (b) CHD after 180 ms ([CHD] ca. 2.5 × 109 molecules cmꢀ3). The
massꢀselected precursor ion is indicated by an asterisk (*).
Minor formation of the addition product between Oxo+O
• and CH at m/z 260 (i.e., 11a) also
occurred. The finding that aliphatic alkoxyl radicals react with alkenes with high preference
through allylic HAT, rather than by addition to the π system, is consistent with literature data38,39
and verifies our experimental setꢀup.
Due to the presence of highly activated bisꢀallylic C–H bonds, the reaction of Oxo+O
• with
CHD was extremely fast and only low excess [CHD] was required for the reaction to be
ACS Paragon Plus Environment
17

The Journal of Physical Chemistry
Page 18 of 41
1
2
completed on a millisecond timescale. Oxo+OH was formed as the major product (Figure 2b),
3
4
5
6
whereas only trace amounts of the adduct 11b (at m/z 258) resulting from addition of Oxo+O
• to
7
8
9
one of the π systems in CHD were observed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the Oxo+O
• reactions with both CH and CHD formation of small amounts of byꢀproducts at
m/z 148 and m/z 105 also occurred, which result from the competing unimolecular decay of
Oxo+O
•
(see Figure 1b).
The minor byꢀproducts at m/z 155 and m/z 138 in the reaction involving CH and their lighter
analogues in the reaction with CHD (m/z 153 and m/z 136) are likely formed through
fragmentation of the radical adducts 11a and 11b, respectively, which confirms that Oxo+O
•
addition to the π system in CHD indeed occurs, but only as a very minor pathway. Scheme 6
proposes a possible mechanism initiated by 1,6ꢀ or 1,5ꢀHAT, which transfers the radical site onto
the carboxonium moiety to give 12 and 17, respectively. Homolytic fragmentation leads to the
respective isomeric radical cations 13’ and 18’. 13’ can undergo a homolytic dissociation
(pathway (a)) to release the vinyloxy cation 15 at m/z 153/155 and acyl radical 14. Heterolytic
cleavage (pathway (b)) leads to radical 16 and PhC≡O+. Radical cation 18’ could rearrange
through 1,5ꢀHAT from the cyclohex(en)yl ring to give 19, followed by heterolytic fragmentation
into benzoic acid and the cationic radical fragment 20 at m/z 138/136. Thus, although formation
of the Cꢀcentred radical adducts 11a/b are only minor pathways in the reaction of Oxo+O
• with
CH and CHD, our findings clearly show that such intermediates rapidly propagate radical
damage through rearrangement and fragmentation processes.
ACS Paragon Plus Environment
18

Page 19 of 41
The Journal of Physical Chemistry
1
2
3
4
Scheme 6. Proposed mechanism for formation of the byꢀproducts at m/z 155/153 and m/z
5
6
138/136 in the reactions of Oxo+O
• with CH and CHD.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The reaction of Oxo+O
seconds timescale for the Oxo+O
•
with EA was considerably slower and required monitoring on the
signal to decrease to ~20%. Formation of Oxo+OH was the
•
major pathway (Figure 3a), but with increasing reaction time the competing unimolecular decay
Oxo+O → Oxo+C
•
• became significant, as shown in the concentration/time profiles in Figure 3b.
As will be outlined below, the product ion at m/z 177 results from the secondary reaction of
Oxo+C
•
with EA.
ACS Paragon Plus Environment
19

The Journal of Physical Chemistry
Page 20 of 41
1
2
3
4
(a)
Oxo+OH
5
6
7
8
9
Oxo+O
Oxo+C
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
*
Ph
O
(b)
Figure 3. Reaction of Oxo+O with EA ([EA] ca. 7.6 × 109 molecules cmꢀ3). (a) Mass spectrum
•
after 960 ms; the massꢀselected precursor ion is indicated by an asterisk (*). (b) Concentrationꢀ
time profile for Oxo+O , m/z = 178), Oxo+OH ( , m/z = 179), Oxo+C
PhC≡O+ (
•
(ꢀ

• (ꢁ, m/z = 148) and
ꢀ
, m/z = 105).
In contrast to this, the reaction of Oxo+O
450 ms for the Oxo+O signal to drop to ~20%. Formation of Oxo+OH was identified as the
major pathway (see Figure S6). The reaction of Oxo+O
Oxo+O
• with 1,1ꢀDME was very fast and required only about
•
•
with 1,2ꢀDME was even faster, and the
•
signal dropped to < 20% in less than 150 ms at the lowest experimentally possible
excess [1,2ꢀDME]. Oxo+OH was the only product (see Figure S7). These findings clearly
demonstrate the high susceptibility of aliphatic ethers to damage by alkoxyl radicals. The
ACS Paragon Plus Environment
20

Page 21 of 41
The Journal of Physical Chemistry
1
2
3
4
5
6
regioselectivity in the reactions with EA, 1,1ꢀDME and 1,2ꢀDME was identified using
computational studies, which will be discussed in section 3.
7
8
9
We also determined the absolute rate coefficients for various the ionꢀmolecules reactions of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Oxo+O
•, which are compiled in Table 1 (see Experimental Methods for details).
Table 1. Absolute secondꢀorder rate coefficients k for the reaction of Oxo+O
• with the neutral
model systems at 298 K.a,b
k (cm3 molecule-1 s-1)
Neutral
Oxo+O
•
consumption
Oxo+OH formation
(1.7 ± 0.4)× 10ꢀ10
(1.2 ± 0.3) × 10ꢀ9
(7.6 ± 1.9) × 10ꢀ11
(3.4 ± 0.9) × 10ꢀ10
(1.5 ± 0.4) × 10ꢀ9
Oxo+O
• addition
(3.0 ± 0.8) × 10ꢀ10
(1.5 ± 0.4) × 10ꢀ9
(9.2 ± 2.3) × 10ꢀ11
(4.6 ± 1.2) × 10ꢀ10
(1.6 ± 0.4) × 10ꢀ9
(1.1 ± 0.3) ×10ꢀ11 c
CH
CHD d
(7.0 ± 2.1) ×10ꢀ12 e
EA
ꢀꢀ
ꢀꢀ
ꢀꢀ
1,1ꢀDME
1,2ꢀDME d
aRate data for Oxo+O
•
consumption from pseudoꢀfirst order measurements; rate data for
Oxo+OH formation and Oxo+O
• addition were fitted using the Dynafit 4 program (see
c
Experimental Methods). bExperimental error (≥25%) included. Formation of product 11a at m/z
d
260. Determined by averaging k obtained from individual concentration–time profiles (see
Experimental Methods). eFormation of product 11b at m/z 258.
The rate data determined for the consumption of Oxo+O
• confirm its high reactivity, in
particular with CHD and 1,2ꢀDME, which are about 30 times higher than for EA, the least
reactive substrate explored in this study. The rate coefficient for the reaction with CH is one
order of magnitude lower than with CHD. This difference in reactivity can be rationalised by the
presence of four bisꢀallylic hydrogen atoms in CHD, which are highly susceptible for HAT.
Comparison of the kinetic data for product formation in the reaction of CHD revealed similar
rate coefficients for Oxo+OH formation and Oxo+O
• consumption, clearly confirming that HAT
ACS Paragon Plus Environment
21

The Journal of Physical Chemistry
Page 22 of 41
1
2
3
4
is the dominant pathway (kHAT/kadd ≈ 170). In contrast to this, radical addition is considerably
5
6
7
8
more favourable in the reaction of Oxo+O
The rate coefficients for Oxo+O consumption and Oxo+OH formation in the reactions involving
•
with CH, as revealed by the ratio of kHAT/kadd ≈ 15.
•
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1,1ꢀDME and 1,2ꢀDME are, within error, the same. This shows that HAT is the exclusive
reaction, in agreement with the findings from the product studies. The slightly lower rate
coefficient for Oxo+OH formation compared to Oxo+O
with EA can be rationalised by the fact that unimolecular fragmentation in Oxo+O
competitive pathway. Overall, the excellent agreement between the rate data for Oxo+O
• decay in the comparably slow reaction
•
became a
•
consumption and product formation in all reactions provide evidence that there are no
unidentified major sinks for Oxo+O
•, for example through fragmentation with loss of the
permanent charge tag, which would render these products ‘invisible’ by mass spectrometry.
1.2 Ion-molecule reactions with Oxo+C
Oxo+C was considerably less reactive in its bimolecular reactions than Oxo+O
monitoring on the time scale of several seconds was required for the Oxo+C
•
•
•, and
•
signal to drop to
~20%. Under these conditions the fragmentation to PhC≡O+ (see Scheme 4) became an
important side reaction.
In the reaction of Oxo+C with CH the HAT product Oxo+CH (at m/z 149) and the radical
•
adduct 21a (at m/z 230) were both obtained as major products (Figure 4a). However, the most
abundant species at m/z 187 could be assigned to the adduct 22a, which was formed through
electrophilic addition of the fragmentation product PhC≡O+ to CH. This was confirmed through
independent experiments where PhC≡O+ was isolated in the ion trap and reacted with CH.
ACS Paragon Plus Environment
22

Page 23 of 41
The Journal of Physical Chemistry
1
2
3
4
5
*
(a)
Oxo+C
O
6
Ph
22a
7
8
9
O
Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ph
O
Oxo+CH
O
21a
*
(b)
Ph
Oxo+C
O
Ph
Oxo+CH
O
Ph
O
22b
O
21b
185
Figure 4. Mass spectrum of the reaction of Oxo+C
with (a) CH after 5000 ms ([CH] ca. 8.7 ×
•
109 molecules cmꢀ3) and with (b) CHD after 5000 ms ([CHD] ca. 3.3 × 109 molecules cmꢀ3). The
massꢀselected precursor ion is indicated by an asterisk (*).
Oxo+CH was also obtained as major product in the reaction of CHD with Oxo+C
•, due to the
higher susceptibility of the bisallylic methylene groups for HAT, whereas radical addition to give
the adduct 21b at m/z 228 was less favourable (Figure 4b). Formation of small amounts of the
CHD – PhC≡O+ adduct 22b at m/z 185 was also observed. No attempts were made to identify the
ACS Paragon Plus Environment
23

The Journal of Physical Chemistry
Page 24 of 41
1
2
product at m/z 200, since it did not seem to have a ‘counterpart’ in the reaction of Oxo+C
• with
3
4
5
6
CH.
7
8
9
Interestingly, Oxo+C
timescale of our experiment. The only significant products were PhC≡O+, Oxo+COO
• did not react with EA through HAT to a noticeable extent on the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
•
(m/z 180;
resulting from reaction of Oxo+C
(Figure 5a).
• with residual O2 in the ion trap) and a product at m/z 177
ACS Paragon Plus Environment
24

Page 25 of 41
The Journal of Physical Chemistry
1
2
3
4
(a)
O
Oxo+C
Ph
5
*
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
Ph
Oxo+COO
23
*
(b)
Oxo+C
Oxo+CH
Oxo+COO
O
Oxo+C
(c)
Ph
*
Oxo+CH
Oxo+COOH
Oxo+COO
Figure 5. Mass spectrum of the reaction of Oxo+C
with (a) EA after 5000 ms ([EA] ca. 7.6 ×
•
109 molecules cmꢀ3), (b) 1,1ꢀDME after 5000 ms ([1,1ꢀDME] ca. 7.1 × 109 molecules cmꢀ3) and
with (c) 1,2ꢀDME after 5000 ms ([1,2ꢀDME] ca. 5.6 × 109 molecules cmꢀ3). The massꢀselected
precursor ion is indicated by an asterisk (*).
ACS Paragon Plus Environment
25

The Journal of Physical Chemistry
Page 26 of 41
1
2
3
4
5
6
7
Trapping experiments of both PhC≡O+ and Oxo+COO
•
revealed that the m/z 177 product was
8
9
not formed through reaction of EA with either of these species. It is therefore suggested that
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Oxo+C
• reacts with EA through a homolytic substitution, as shown in Scheme 7, leading to the
closedꢀshell cation 23 and carboxyl radical 24 (which could subsequently decarboxylate). Future
studies with isotopically labelled EA will reveal further insight into this unusual radical alkyl
transfer reaction.
Scheme 7. Proposed mechanism for formation of the product at m/z 177 in the reaction of
Oxo+C
• with EA.
In contrast to this, Oxo+C
Oxo+CH (Figure 5b). In addition, small amounts of Oxo+COO
• reacted readily with 1,1ꢀDME through HAT and formation of
•
were also produced. The notable
absence of PhC≡O+ could indicate a fast reaction with 1,2ꢀDME that might lead to formation of
the products at m/z 85 and m/z 117, which were not further identified.
The reaction of Oxo+C
DME and led to Oxo+CH as major product (Figure 5c). Additional products resulted from the
competing reactions of Oxo+C , e.g. the decay to PhC≡O+ and formation of Oxo+COO
• with 1,2ꢀDME was considerably faster than with the isomeric 1,1ꢀ
•
•. The
latter likely underwent a secondary HAT reaction with 1,2ꢀDME to produce the hydroperoxide
Oxo+COOH at m/z 181, which clearly highlights the high susceptibility of ethylene glycol
moieties to radicalꢀinduced damage.
ACS Paragon Plus Environment
26

Page 27 of 41
The Journal of Physical Chemistry
1
2
The decay rate data determined for the Oxo+C
•
reactions with the various neutral model
3
4
5
6
systems are about 1 – 2 orders of magnitude lower than for Oxo+O
• (Table 2). Similar to
7
8
9
Oxo+O
•, the fastest rate coefficients were found for CHD and 1,2ꢀDME, whereas the other
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
neutrals are about ten times less reactive.
Table 2. Absolute secondꢀorder rate coefficients for the reaction of Oxo+C
• with the neutral
model systems at 298 K.a,b
k (cm3 molecule-1 s-1)
Oxo+C
•
Neutral
Oxo+CH formation
Oxo+C
• addition
consumption
(7.6 ± 1.9) ×10ꢀ12
(3.1 ± 0.8) × 10ꢀ11
(5.7 ± 1.4) × 10ꢀ12
(3.8 ± 1.0) × 10ꢀ12
(1.6 ± 0.4) × 10ꢀ11
(3.3 ± 0.8) × 10ꢀ12
(1.8 ± 0.4) × 10ꢀ11
< 9 × 10ꢀ14
(3.3 ± 0.8) × 10ꢀ12 c
CH
CHD
(1.9±1.9) × 10ꢀ12 d,e
(2.1 ± 0.5) × 10ꢀ12 f
EA
(3.5 ± 0.9) × 10ꢀ12
(1.4 ± 0.4) × 10ꢀ11
1,1ꢀDME
1,2ꢀDME
aRate data for Oxo+C
ꢀꢀ
ꢀꢀ
•
consumption from pseudoꢀfirst order measurements; rate data for
Oxo+CH formation and Oxo+C
• addition were fitted using the Dynafit 4 program (see
Experimental Methods). bExperimental error (≥25%) included. Formation of product 21a at m/z
c
d
230. Determined by averaging k obtained from individual concentration–time profiles (see
Experimental Methods). Formation of product 21b at m/z 228. Formation of product 23 at m/z
e
f
177 (see text).
The data show a ratio of kHAT/kadd ≈ 1 for the reaction of Oxo+C
• with CH, whereas for the
reaction with CHD the HAT pathway is considerable more favourable (kHAT/kadd ≈ 10). Similar
to Oxo+O , the rate coefficients for Oxo+C consumption and Oxo+CH formation in the reactions
•
•
involving 1,1ꢀDME and 1,2ꢀDME, respectively, are the same, confirming that HAT is the
exclusive bimolecular reaction pathway for Oxo+C
•.
ACS Paragon Plus Environment
27

The Journal of Physical Chemistry
Page 28 of 41
1
2
While the rate coefficient determined from the consumption of Oxo+C
• in the reaction with EA
3
4
5
6
7
8
suggests a fast reaction, this is not reflected by the rate coefficient for Oxo+CH formation, which
is lower by more than two orders of magnitude (it should be noted that the rate of Oxo+CH
formation in this reaction, kHAT, is at the low end of our experimental setꢀup and only an upper
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
limit can be given). As outlined in the product studies, the dominant loss channels for Oxo+C
•
are unimolecular decay to PhC≡O+ and formation of the closedꢀshell compound 23 through an
unusual homolytic ethyl transfer from EA, which is much faster than HAT (kHAT/ktransfer < 0.04;
see Table 2).
2. Computational studies
DFT calculations were carried out to determine the most reactive site for the Oxo+O
•/C•
induced HAT in EA, 1,1ꢀDME and 1,2ꢀDME. All reactions proceed through formation of an
initial ionꢀmolecule association complex (reactant complex), which is at least 60 kJ molꢀ1 lower
in energy than the free reactants. The energy of the free reactants, or ‘entrance channel’, will be
defined as reference point, which is set to 0 kJ molꢀ1. Due to 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 energy that is released upon complex formation.
Any subsequent 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.40,41 Reaction pathways with energies above the entrance channel
cannot be accessed.
ACS Paragon Plus Environment
28

Page 29 of 41
The Journal of Physical Chemistry
1
2
Generally, the HAT reactions involving Oxo+O
•
are more exothermic than with Oxo+C
•.
3
4
5
6
Figure 6 shows the calculated potential energy diagram for the reaction of Oxo+O and Oxo+C
• •
7
8
9
with EA. HAT from the αꢀ, βꢀ, and γꢀposition and formation of the isomeric EA radicals (as
product complex with Oxo+OH or Oxo+CH, respectively) is exothermic for both Oxo+O
Oxo+C
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
•
and
•.
Figure 6. Potential energy diagram for the reaction of Oxo+O
•
and Oxo+C
(normal) and Oxo+C
• (in
• with EA. M062X/6ꢀ
31+G* enthalpies are in kJ molꢀ1 for the reaction involving Oxo+O
•
italics).
The αꢀ and βꢀradicals are the thermodynamically preferred products, whereas formation of the
γꢀradical is least favourable. HAT from the βꢀposition is the kinetically preferred pathway for
•
Oxo+O , which is associated with an activation barrier of 31 kJ molꢀ1. In contrast to this, any of
the possible TSHAT for the reaction with Oxo+C
•
are located either close to or above the entrance
channel, indicating a slow reaction through HAT, in agreement with the experimental data.
ACS Paragon Plus Environment
29