Ultrafast Observation of a Photoredox Reaction Mechanism: Photoinitiation in Organocatalyzed Atom-Transfer Radical Polymerization

Article abstract of DOI:10.1021/jacs.7b07829

Photoredox catalysis has driven a revolution in the field of organic chemistry, but direct mechanistic insights into reactions of genuine synthetic utility remain relatively scarce. Herein we report ultrafast time-resolved spectroscopic observation of a bimolecular organocatalyzed photoredox reaction, from catalyst photoexcitation through to photoinduced electron transfer (PET) and intermediate formation, using transient vibrational and electronic absorption spectroscopy with sub-picosecond time resolution. Specifically, the photochemical dynamics of initiation in organocatalyzed atom-transfer radical polymerization (O-ATRP) are elucidated for two complementary photoredox organocatalysts (N,N-diaryl-5,10-dihydrophenazines). Following photoexcitation, a dissociative bimolecular electron transfer is observed from the first excited singlet state of both photocatalysts to methyl 2-bromopropionate in dichloromethane, toluene, and dimethylformamide. The photocatalyst excited donor state, ground state, and radical cation are tracked in real time alongside the debrominated radical fragment. Our work challenges previously proposed mechanisms of initiation in O-ATRP and indicates that PET from short-lived excited singlet states can exert control of polymer molecular weight and dispersity by suppressing the steady-state concentration of the reactive debrominated radical. More broadly, we aim to demonstrate the potential of ultrafast absorption spectroscopy to observe directly transient, open-shell intermediates in mechanistic studies of photoredox catalysis.

Full text of DOI:10.1021/jacs.7b07829

Subscriber access provided by READING UNIV
Article
Ultrafast Observation of a Photoredox Reaction Mechanism: Photo-
initiation in Organocatalyzed Atom-Transfer Radical Polymerization
Daisuke Koyama, Harvey J.A. Dale, and Andrew J. Orr-Ewing
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b07829 • Publication Date (Web): 04 Jan 2018
Downloaded from http://pubs.acs.org on January 4, 2018
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.
Journal of the American Chemical Society 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 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Ultrafast Observation of a Photoredox Reaction Mechanism: Photoꢀinitiation
in Organocatalyzed AtomꢀTransfer Radical Polymerization
1
1
Daisuke Koyama , Harvey J. A. Dale and Andrew J. OrrꢀEwing*
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
School of Chemistry, University of Bristol, Cantock’s Close, Bristol, BS8 1TS, UK
*
Author for correspondence: a.orrꢀewing@bristol.ac.uk
Abstract
Photoredox catalysis has driven a revolution in the field of organic chemistry, but direct
mechanistic insights into reactions of genuine synthetic utility remain relatively scarce. Herein
we report ultrafast timeꢀresolved spectroscopic observation of a bimolecular organocatalysed
photoredox reaction, from catalyst photoexcitation through to photoinduced electron transfer
(PET) and intermediate formation, using transient vibrational and electronic absorption
spectroscopy with subꢀpicosecond time resolution. Specifically, the photochemical dynamics of
initiation in organocatalysed atomꢀtransfer radical polymerization (OꢀATRP) are elucidated for
two complementary photoredox organocatalysts (N,Nꢀdiarylꢀ5,10ꢀdihydrophenazines).
Following photoexcitation, a dissociative bimolecular electron transfer is observed from the first
excited singlet state of both photocatalysts to methyl 2ꢀbromopropionate in dichloromethane,
toluene and dimethylformamide. The photocatalyst excited donor state, ground state and radical
cation are tracked in real time alongside the debrominated radical fragment. Our work challenges
previously proposed mechanisms of initiation in OꢀATRP, and indicates that PET from shortꢀ
lived excited singlet states can exert control of polymer molecular weight and dispersity by
supressing the steadyꢀstate concentration of the reactive debrominated radical. More broadly, we
aim to demonstrate the potential of ultrafast absorption spectroscopy to observe directly
transient, openꢀshell intermediates in mechanistic studies of photoredox catalysis.
1
DK and HJAD contributed equally to this work.
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Introduction
Photoredox catalysis exploits the notion that closedꢀshell molecules become more potent electron
donors and electron acceptors when promoted to an excited electronic state. The photoexcitation
of such species can be used to promote electron transfers that would otherwise not be
thermodynamically viable in the ground electronic state. Exceptional advances in photoredox
catalysis have been made in the last decade under the umbrella of synthetic organic chemistry,
with the recent resurgence in this area often ascribed to pioneering research by the groups of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1ꢀ3
MacMillan, Yoon, and Stephenson. This work, and much that followed, has turned to the
2
+
3+
tris(bipyridine) complexes of [Ru(bpy)3] and [Ir(bpy)3] and their analogues as the
photocatalysts (PCs) of choice. The photophysics of these transition metal complexes is well
4ꢀ5
characterized, and they have since been applied as PCs to a range of problems across synthetic
6
ꢀ8
organic chemistry, including the total synthesis of natural products, heterocycle synthesis and
9
10
11ꢀ13
functionalization, asymmetric protocols and the preparation of carbonꢀcentered radicals.
1
4
They have also been used in the photocatalytic reduction of carbon dioxide to methane and in
15ꢀ22
the development of cooperative catalysis.
To circumvent the problems of natural scarcity and
2
3ꢀ24
toxicity associated with the transition metals in these photoredox systems,
researchers are
2
5ꢀ27
replacing metalꢀbased catalysts with highly conjugated organocatalysts;
applications, and
2
8ꢀ30
31
associated mechanistic studies, include oxidation using molecular oxygen,
crossꢀcoupling,
32
33
34
35
36
cyanation, addition to alkenes, trifluoromethylation, hydroxylation, Nꢀformylation, and
3
7
sulfinylation.
In addition to these synthetic uses, selected organic PCs such as perylene, phenothiazine,
dihydrophenazine and phenoxazine derivatives have recently been applied in the development of
38ꢀ41
organocatalysed atomꢀtransfer radical polymerization (OꢀATRP).
Initiation in OꢀATRP
depends upon the reductive dehalogenation of an alkyl halide initiator via an outerꢀsphere
electron transfer from a photoexcited organocatalyst; this process forms a reactive radical
species, which instigates the propagation stage of polymerization. Low dispersity and control
over the polymer molecular weight are thought to be promoted by keeping the concentration of
42
this radical low. Miyake and coꢀworkers recently proposed an oxidative quenching pathway for
4
0ꢀ41
OꢀATRP with N,N’ꢀdiarylꢀ5,10ꢀdihydrophenazines,
postulating that the first excited triplet
3
state PC*(T ) is responsible for the PET to the initiator (e.g. methyl 2ꢀbromopropionate, MBP).
1
2
ACS Paragon Plus Environment

Page 3 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
2
.
The electron transfer is dissociative, forming a dehalogenated radical fragment ( MP ) and an
2
.+
ꢀ
ionic complex ( PC Br ), and termination of the polymerization supposedly occurs via a
4
3
bimolecular, thermally mediated back electron transfer (Scheme 1). They subsequently posited
that the PCs most capable of efficiently mediating OꢀATRP, and of affording the greatest control
over polymer molecular weight and dispersity, are those whose excited states exhibit substantial
intramolecular charge transfer (CT) character. The source of such control, they suggested, is that
CT character in the excited donor state minimizes fluorescence and enhances electron transfer
rates. They further investigated effects of solvents and irradiation energy on resulting polymer
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4ꢀ45
properties.
However, there is scope for more detailed study of these OꢀATRP mechanisms to
establish a firmer basis for interpretation of the existing observations.
In photoredox reactions using metal complexes, it is widelyꢀaccepted that longꢀlived T states
1
with metalꢀtoꢀligand CT character, populated with high quantum yield by efficient intersystem
2
4
crossing (ISC), are responsible for PET.
The photochemical properties of the organic PCs
(
e.g., their redox potential, ISC quantum yields, and excited state lifetimes), as well as the
polymerization conditions used in OꢀATRP (e.g., the amount of photocatalyst required)
significantly differ from those for the metalꢀcentered photocatalysts, and their reaction
mechanism may therefore also differ. For example, synthetic studies with phenyl phenothiazines
1
have led researchers to contend that the first excited singlet state PC*(S ), as opposed to
1
3
46
PC*(T ), may be responsible for the photoinduced electron transfer step. In contrast, a recent
1
47
spectroscopic study concluded that the majority of the PET (~89%) took place from T state,
1
but was conducted at initiator concentrations 5 – 10 times lower than used for OꢀATRP
40
conditions. Exhaustive mechanistic work by Matyjaszewski and coꢀworkers with phenyl
phenothiazine photocatalysts further emphasized the importance of the radical cation lifetime
and the termination stage. These authors argued that stable radical cations with long lifetimes are
48
required for controlled polymerization. We show here that unravelling the mechanisms of
organic photoredox catalysis can benefit from scrutiny by ultrafast transient absorption
spectroscopy, which tracks the abundances of both the excited electronic states of the PC and
reactive openꢀshell intermediates directly on timescales from subꢀpicosecond upwards, using the
temporal evolution of their absorption signatures.
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Several previous transient spectroscopy studies have examined photoinduced electron transfer
49
(PET) reactions in solution. However, this prior work has largely focussed either on
intramolecular PETs, in which the electron donor and acceptor are covalently tethered, or on
intermolecular PETs that result in charge separation and recombination without the prospect of
subsequent atomꢀtransfer or onward reactivity. Ultrafast spectroscopy has previously been used
to investigate the nature of the nascent ionꢀpair, quenching pathways, and the Marcus inversion
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0ꢀ51
region in bimolecular PETs.
Herein, we demonstrate an alternative application as a broader
mechanistic tool to address the growing prominence of organic photoredox catalysis in synthesis.
Vauthey and coꢀworkers recently reported the use of ultrafast transient vibrational absorption
52
spectroscopy (TVAS) to observe the photophysical relaxation of Ru(II) complexes, but the
current study explores both the intramolecular relaxation of an organic photocatalyst and its
subsequent bimolecular reactions.
The photochemical dynamics of OꢀATRP initiation by N,N’ꢀdiarylꢀ5,10ꢀdihydrophenazine PCs
were investigated using both TVAS and transient electronic absorption spectroscopy (TEAS)
with subꢀpicosecond time resolution. Interpretation of the experimental data was supported by
KohnꢀSham density functional theory (KSꢀDFT), timeꢀdependent density functional theory (TDꢀ
DFT) and highly correlated ab initio SAꢀCASSCF/CASPT2 calculations. We consider two
complementary photocatalysts based on the N,N’ꢀdiarylꢀ5,10ꢀdihydrophenazine framework:
N,N’ꢀdiphenylꢀ5,10ꢀdihydrophenazine (PCH) and 5,10ꢀdi(4ꢀtrifluoromethylphenyl)ꢀ5,10ꢀ
dihydrophenazine (PCF). These PCs were selected to test the mechanistic postulates of Miyake
and coꢀworkers, because the emissive state in PCH has local excitation (LE, πꢀπ*) character,
whereas in PCF it has substantial intramolecular CT character, as deduced from solvatochromic
4
3
effects in emission.
To balance spectroscopic convenience with synthetic applicability, we
pursued experiments with three solvents: dichloromethane (DCM), toluene and
dimethylformamide (DMF). The outcomes offer a complete picture of the early time dynamics of
all the key intermediates in the PET process, and because some of our findings challenge
4
0ꢀ41
previous mechanistic interpretations,
organic photocatalysts for OꢀATRP.
we accordingly propose some desirable properties of
4
ACS Paragon Plus Environment

Page 5 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
PC* (Sn)
IC / ISC
hv
PC: Photocatalyst
RX: initiator
M: monomer
1
PC* (S ) /
1PC (S0)
1
3
PC* (T1)
RX = MBP
DPET
TET
2
PC Xꢀ
・+
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RM X
(n ≥ 0)
O
n
RMn+1X
H3C
OCH3
M
RMn・
RMn+1
・
PCH
R・ = MP・
PCF
propagation
Scheme 1: Organic photoredox catalysis of atomꢀtransfer radical polymerization. The photocatalytic cycle for Oꢀ
ATRP with N,N’ꢀdiarylꢀdihydrophenazines (PC) and alkyl bromide initiators, with electron transfer either from
the S or T state of the photocatalyst. IC = Internal conversion. ISC = Intersystem crossing. DPET = Dissociative
1
1
photoinduced electron transfer. TET = Thermal electron transfer. The chemical structures of the PCs used in this
study are shown on the right.
Methods
All the TVAS and TEAS experiments reported here were conducted with an ultrafast laser
system configured for pumpꢀprobe measurements. A Coherent VitaraꢀS Ti:Sapphire oscillator
was used in conjunction with a Coherent Legend Elite HE+ regenerative amplifier to generate a
single train of parent pulses with a repetition rate of 1 kHz. Each pulse was characterized by an
average duration of τ = 40 fs, a carrier wavelength of ꢀ = 800 nm and an energy of ~5 mJ. The
0
parent beam was split along three separate pathways; two beams of 2.45 mJ pulses were directed
into Coherent OPerA Solo optical parametric amplifiers (OPAs) to generate UV pump and IR
probe pulses (for TVAS), while a single train of 100 ꢁJ pulses was diverted into a CaF window
2
to generate a broadband white light supercontinuum probe (for TEAS). In TEAS and TVAS, the
UV pump wavelength was set to 370 nm and the pulses were characterized by an average energy
of 600 nJ at the sample. Solutions were prepared to the desired concentration in a 10 mL amber
glass volumetric flask and connected to a stainless steel Harrick flow cell via PTFE tubing. Two
CaF windows were sealed to the Harrick cell body by Kalrez Oꢀrings and were separated by a
2
pair of PTFE spacers of 380 ꢀm (TVAS) or 500 ꢀm (TEAS) thickness. The solution was flowed
continuously through the cell using a peristaltic pump, to ensure that consecutive pump and
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
probe pulses sampled fresh regions of the solution. Further details are presented in the ESI,
alongside computational and synthetic methodologies. Photocatalyst concentrations of 2.1 mmol
ꢀ3
ꢀ3
dm and initiator concentrations of 0.4 – 2.6 mol dm were employed throughout the study.
ꢀ3
These compare with concentrations of 5 – 10 mmol dm for the photocatalysts and 0.05 – 0.1
ꢀ3
40
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
mol dm for the initiator in OꢀATRP studies by Miyake and coꢀworkers.
Results and Discussion
SteadyꢀState Characterization
The
organocatalysts
N,N’ꢀdiphenylꢀ5,10ꢀdihydrophenazine
and
5,10ꢀdi(4ꢀ
trifluoromethylphenyl)ꢀ5,10ꢀdihydrophenazine exhibit intriguing photophysical behavior. In both
molecules, the lowestꢀenergy optically bright transition occurs with a maximum absorbance at
3
70 nm. TDꢀDFT calculations indicate that at this wavelength both PCH (D ) [S (A ) →
2
h
0
g
S (B )] and PCF (C ) [S (A ) → S (B )] undergo photoexcitation to a state with primarily LE
2
2u
2h
0
g
4
u
character (πꢀπ*) in the FranckꢀCondon region (Figure S7 and S9 of Supporting Information
(SI)). The maximum in the emission band of PCH is recorded at a wavelength of 480 nm,
41
independent of solvent, whereas emission from PCF exhibits substantial solvatochromic shifts,
with wavelengths of maximum emission varying from 480 – 650 nm depending upon solvent
4
3
polarity (Figure S4), in agreement with a prior report. The steadyꢀstate emission intensity from
PCF is essentially unchanged upon the addition of triplet quenchers 2,5ꢀdimethylhexaꢀ2,4ꢀdiene
ꢀ1 53
ꢀ1 54
(ET1 = 176 kJ mol
), cyclohexaꢀ1ꢀ3ꢀdiene (E = 219 kJ mol ) and styrene (E = 254 kJ
T1 T1
ꢀ1
mol 55) (Figure S22), or the triplet sensitizer benzophenone. The positions of the emission
bands also remain independent of both the photocatalyst concentration and excitation
wavelength. The emissive state of PCF is thus argued to be the firstꢀexcited singlet state
1
PCF*(S ), in accordance with Kasha’s rule. This assignment is supported by TEAS and TVAS
1
1
measurements discussed below. The solvatochromism is consistent with a PCF*(S ) state of
1
intramolecular CT character, whereas the solventꢀindependent emission behaviour of PCH points
4
3
to an emissive state of LE character. The TEAS measurements for PCH reported later show
evidence of slow evolution of the S to the T state, which suggests contributions to emission
1
1
1
3
from both PCH*(S ) and PCH*(T ), but prior reports indicate that this emission is not affected
1
1
6
ACS Paragon Plus Environment

Page 7 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
56ꢀ57
by the presence of oxygen.
Because our transient absorption spectroscopy measurements
establish the timescale for intersystem crossing, no further quenching studies of the PCH
emission were carried out.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Locating reactive intermediates
The transient vibrational absorption (TVA) spectra presented in Figure 1(a), and obtained over
ꢀ1
1
the 1530 – 1680 cm region, highlight a band corresponding to a ring motion of PCF*(S )
1
ꢀ1
1
(
1548 cm ) and a bleach from the photoinduced depletion of the ground state PCF(S ) (1600
0
ꢀ1
cm ) (Figure S1). The absence of recovery of the bleach feature is ascribed to an overlapped
1
transient absorption from a second ring breathing mode of PCF*(S ). Combined TVAS and
1
1
TEAS experiments (Figures S1 and S2) showed that the PCF*(S ) state: (i) is fully populated
1
1
within 2 ps via ultrafast internal conversion from the nascent excited state PCF*(S ); (ii) has an
4
1
estimated lifetime in DCM of τ[ PCF*(S )] = 3 ± 1 ns; and (iii) decays primarily via internal
1
1
conversion or fluorescence to PCF(S ), with no discernible intersystem crossing into the triplet
0
3
manifold. No transient absorptions from other species – namely PCF*(T ) – were detected in
1
any region. Upon the addition of methyl 2ꢀbromopropionate (MBP), two new transient
ꢀ1
ꢀ1
absorptions arise at 1553 cm and 1660 cm . These features are consistent with computed
2
.+
ꢀ1
vibrational frequencies for the photocatalyst radical cation PCF (D ) (1551 cm ) and the
1
2
.
ꢀ1
ꢀ1
debrominated radical fragment MP (D ) (1659 cm ). Any loss of the bleach at 1600 cm
1
1
expected for partial PCF(S ) recovery is again masked by overlapping gain features from
0
1
2
.+
PCF*(S ). Assignment of the PCF (D ) feature was confirmed by steadyꢀstate FTIR
1
1
characterization, with thermally mediated oxidation of PCF in the presence of FeCl (Figure S5).
3
None of the transient features were observed in control experiments performed on solutions of
MBP without the photocatalyst.
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
a) PCF
(b) PCF
1
Time / ps
PCF*(S1)
0
0
.4
.2
1
.0
1
1
2
0
5
0.8
0.6
100
200
1PCF*(S1)
500
1540 1550 1560
0
0
.4
.2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0.0
0.2
ꢀ
0.0
1.0
(c) PCF + MBP
(d) PCF + MBP
1
0
0
0
.4
.2
.0
PCF*(S1)
2PCF・+(D1)
2
・
MP (D )
1
0
0
0
0
0
.8
.6
.4
.2
.0
2
・
MP (D )
1
2
・+
PCF (D )
1
1540 1550 1560
1PCF*(S1)
ꢀ
0.2
1
525
1550
1575
1600
1625
1650
1675
0
200
400
600
800
1000
ꢀ1
Time / ps
Wavenumber / cm
Figure 1: TVAS measurement of electron transfer rates for PCF, with TVA spectra obtained over the wavenumber
ꢀ
1
ꢀ3
range 1525 – 1680 cm following pulsed photoexcitation at 370 nm. (a) TVA spectra for PCF (2.1 mmol dm ) in
1
DCM; (b) the corresponding kinetic trace for PCF*(S ) in the absence of MBP (○), obtained from integration of a
1
1
Gaussian function fitted to the PCF*(S ) absorption band (with floating center) at each time delay. (c) TVA
1
ꢀ
3
ꢀ3
spectra for PCF (2.1 mmol dm ) and MBP (0.9 mol dm ) in DCM; and (d) the corresponding timeꢀdependent band
1
2
.+
2
.
intensities for PCF*(S ) (○), PCF (D ) (○) and MP (D ) (○), obtained from decomposition of the spectra in (c).
1
1
1
Solid lines are fits to a concerted kinetic model for PET discussed in the main text. The absence of bleach recovery
1
in (a) is attributed to the presence of an overlapped transient absorption by PCF*(S )
1
Global fitting of all three integrated band intensity traces in Figure 1(d) to a single time constant
ꢀ
3
1
is satisfactory (e.g. 324 ±ꢂ10 ps, [MBP] = 0.9 mol dm ), showing that: (i) PCF*(S ), rather than
1
3
PCF*(T ), is indeed responsible for electron donation, because there is no kinetic evidence for
1
an S → T step; (ii) the electron transfer outꢀcompetes concomitant photophysical relaxation to
1
1
1
PCF(S ); (iii) dissociation of bromide is sufficiently rapid that it cannot be experimentally
0
distinguished from the electron transfer with our methods; and (iv) there is no significant thermal
2
ꢀ
2
+
backꢀelectron transfer from MP⋅ or Br to PCF⋅ on the timescale of our measurements (a
1
deduction confirmed by the persistance of PCF(S ) bleach features in the TVA spectra). The
0
8
ACS Paragon Plus Environment

Page 9 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
third point is consistent with prior demonstration of concerted carbonꢀhalogen bond dissociation
5
8
by electron attachment, and is further supported by the absence of an absorption signature for
2
.ꢀ
the geminate radical anion MBP (D ). It seems likely such a species would have a vibrational
1
2
.
signature distinct from MP (D ), especially in the carbonyl stretching mode, given the
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
fundamental differences in the charges of the species and the hybridization of the αꢀcarbon. We
therefore represent the electron transfer process kinetically by a single step:
1
2
.+
ꢀ
2
.
PCF*(S ) + MBP → PCF (D ) + Br + MP (D )
(1)
1
1
1
Tracking the PET: PCH vs PCF
2
.
Having located a spectroscopic signature of the MP (D ) radical, its formation was tracked for a
1
series of solutions with different concentrations of MBP (Figure 2) and the kinetics fitted to a
dissociative PET model with a pseudo firstꢀorder rate coefficient ꢃꢇ = ꢃꢄꢅꢆ[ꢈꢉꢊ]. No decay
ꢄꢅꢆ
2
.
pathways for MP (D ) were included in our kinetic model because the radical is not consumed
1
on our timescale. The resulting pseudo firstꢀorder kinetic plot (Figure 2a inset) is linear, and
9
3
ꢀ1 ꢀ1
affords a bimolecular rate coefficient of kPET(PCF) = (3.9 ± 0.2) × 10 dm mol s . Although
fast, the PET appears to remain under activation control. This result was reꢀexamined with TEAS
1
by plotting the pseudo firstꢀorder rate coefficient for PCF*(S ) decay as a function of [MBP]
1
(Figure S19); in agreement with the TVAS measurements, a bimolecular rate coefficient of
9
3
ꢀ1 ꢀ1
kPET(PCF) = (4.4 ± 0.3) × 10 dm mol s was obtained. A more sophisticated kinetic model,
accounting for static quenching at early times, is considered below. The linearity of the inset
plots in Figure 2 (and Figure S19 of SI) indicates that the electronicꢀspin character of the
*
electronꢀdonating excited state of PCF is not affected by intermolecular spinꢀorbit interactions
with the Br atom in MBP.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
a) PCF
(b) PCH
1
0
0
0
.0
.8
.6
.4
1.0
0.8
0.6
0.4
0.010
0
.04
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0
0
.005
.000
0.02
0.00
0.2
0.0
0.2
1.0 1.5 2.0 2.5
0.5 1.0 1.5 2.0
ꢀ3
ꢀ3
[MBP] / mol dm
600 800
Time / ps
0.0
[MBP] / mol dm
150
0
200
400
1000
0
50
100
200
Time / ps
2
.
Figure 2: Rates of production of MP (D ) radicals by electron transfer from the photocatalyst. TVAS
1
ꢀ3
ꢀ3
concentration dependence study for: (a) PCF (2.1 mmol dm ) and MBP in DCM; and (b) PCH (2.1 mmol dm )
and MBP in DCM. Pseudo firstꢀorder kinetic plots are inset. The solid lines represent single exponential fits for
PCF and biexponential fits for PCH (see main text for details).
Miyake and coꢀworkers argued that PET from a photocatalyst such as PCF with excited state CT
character should be faster than from structural analogues with an excited state of LE character,
40ꢀ41
with consequent greater control over polymerization.
To test this suggestion, we also
examined the photochemical dynamics of PCH and MBP in DCM. In the absence of MBP, the
1
lifetime of the PCH*(S ) state in DCM is of the order of a few nanoseconds, and appears similar
1
1
1
to PCF*(S ) (Figure S20). However, there is one notable difference: whereas the PCF*(S )
1
1
state decays to the groundꢀstate without perceptible intersystem crossing into the triplet
1
manifold, the decay of the PCH*(S ) state coincides with the concomitant growth of a new
1
transient absorption centered on 450 nm (Figure S20). The position of this feature is inconsistent
2
.+
with the absorption signature of the radical cation PCH (D ), indicating that there is no PET
1
from photoexcited PCH to DCM, and we instead assign the new feature to a triplet state
3
PCH*(T ). The quantum yield for PCH ISC was previously measured to be 0.26 in 3ꢀ
1
5
9
methylpentane solution at 77 K, supporting our assignment. In the presence of MBP a clear
2
.
transient absorption from MP (D ) was again observed (Figure S18) on timescales faster than T
1
1
growth, with a concentration dependence study affording an estimated bimolecular rate
1
0
3
ꢀ1 ꢀ1
coefficient of kPET(PCH) = (3.6 ± 0.2) × 10 dm mol s . As with PCF, electron transfer
1
appears to occur from the first excited singlet state, PCH*(S ) under our experimental
1
conditions.
10
ACS Paragon Plus Environment

Page 11 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
ꢀ3
At the relatively high initiator concentrations (0.4 – 2.2 mol dm ) considered herein, our
assumption of pseudo firstꢀorder kinetics is an approximation. The rate coefficient for the PET
2
4
will in truth be timeꢀdependent, because of a contribution from static quenching at early times.
The kinetics in the case of PCF are well fitted by monoꢀexponential functions, indicating that a
simple pseudo firstꢀorder approach is appropriate. In the case of PCH, however, curvature in
semiꢀlogarithmic plots (Figure S16) suggests that timeꢀdependent kinetics ought to be
considered. Vauthey and coꢀworkers used a formal diffusion model to deduce intrinsic rate
coefficients for bimolecular electron transfers with both static and diffusive quenching
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
51
contributions, but here we adopt a simpler procedure governed by the available data, in which
the kinetics are fitted with biꢀexponential functions. The larger rate coefficients k′PET(Stat) are
ascribed to the static PET (7 – 17 ps), which might include ET within weakly associated PCꢀ
MBP complexes, whereas the smaller rate coefficients k′PET(Dif) describe the diffusive PET. The
bimolecular rate coefficient kPET may then be obtained by plotting k′PET(Dif) as a function of
10
[MBP]; using this method, a marginally lower rate coefficient of kPET(PCH) = (2.0 ± 0.2) × 10
3
ꢀ1 ꢀ1
dm mol s was obtained. By focusing hereafter on the concentrationꢀdependent, diffusional
component of the bimolecular reaction kinetics, our discussion excludes possible contributions
from association complexes of the photocatalyst and MBP.
Irrespective of the model invoked to analyse the kinetics, the primary conclusion remains
unchanged: the rate of PET is 5 – 10 times greater for PCH than PCF, indicating that the
1
1
PCF*(S ) state, of CT character, undergoes slower PET. Electron transfer from PCH*(S )
1
1
1
0
appears to be limited by the diffusionꢀcontrolled rate coefficient, estimated to be ~ 1.6 × 10
3
ꢀ1 ꢀ1
54
dm mol s in DCM.
MarcusꢀSavéant theory states that the free energy barrier for a concerted dissociative electron
‡
transfer reaction ꢁPETG is related to the free energy change for the transfer ꢁPETG, the bond
60
dissociation energy EBD and the solvent reorganization energy ꢀ according to:
0
ꢜ
ꢜ
ꢏꢐꢑꢌ
ꢒꢓ
ꢖꢗꢘꢙ
ꢚ
ꢌ
ꢒꢓ
ꢖꢗꢘꢙ
ꢚ
‡
Δꢋꢌꢍꢎ =
ꢕ1 +
ꢛ ꢂ≈ꢂ ꢔ ꢕ1 + ꢛ
(2)
ꢔ
ꢏꢐꢑꢌꢒꢓ
ꢌꢒꢓ
To a good approximation the bond dissociation energy dominates the reorganization energy,
leading to a simpler expression that depends exclusively upon EBD and ꢁPETG. The steadyꢀstate
1
max
e
emission spectra of the two photocatalysts in DCM show that the PCF*(S ) state [ꢀ
(DCM)
1
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ꢀ1
1
max
ꢀ
=
1
562 nm, 213 kJ mol ] is lower in energy than PCH*(S ) [ꢀ
(DCM) = 480 nm, 249 kJ mol
1
e
], relative to the respective groundꢀstate photocatalysts, and KSꢀDFT calculations show that the
2
.+
ꢀ1
2
.+
ꢀ1
PCF (D ) radical cation (436 kJ mol ) is higher in free energy than PCH (D ) (421 kJ mol ).
1
1
That is to say, PCF has a more stable excited donor state and less stable radical cation compared
to PCH, leading to an estimated difference in the PET driving force for the two photocatalysts of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ1
ꢁꢁPETG = ꢁPETG(PCF) - ꢁPETG(PCH) = 51 kJ mol . The Coulombic correction term present in
the Weller equation, used to compute ꢁPETG, is unlikely to differ to such a significant degree for
the same solvent and electron acceptor, particularly given the structural similarity of the two
6
1
photocatalysts. Thus, the faster PET of PCH is consistent with MarcusꢀSavéant theory outside
the inversion region, with approximate computations (outlined in the SI) affording ꢃꢋꢌꢍ(PCH) =
1
1
3
ꢀ1 ꢀ1
9
3
ꢀ1 ꢀ1
1
.3 × 10 dm mol s and ꢃ (PCF) = 1.6 × 10 dm mol s . These values are in good
ꢋꢌꢍ
agreement with the observed diffusionꢀcontrol in PCH and activationꢀcontrolled rate in PCF, but
ꢀ
may be underestimates if the MP⋅ radical and Br products form association complexes in
62
solution.
To assess the role of solvent and consider catalytically relevant conditions, the photophysical
relaxation pathways of PCF and PCH were also investigated in DMF, which represents a typical
4
0
solvent for OꢀATRP. In the case of PCF in DMF, in the absence of the initiator, both TEAS
1
1
and TVAS show quenching of PCF*(S ), and in particular, TVAS indicates PCF(S ) groundꢀ
1
0
state recovery with a time constant of 677 ± 35 ps (Figure 3). No electronic or vibrational
2
.+
absorption signature of the radical cation PCF (D ) was observed, confirming no electron
1
transfer to DMF, nor was there any evident branching into the triplet manifold. We estimate a T1
quantum yield of less than 10% from analysis of the decays in Fig. 3(b). Accordingly, we assign
1
this substantially reduced lifetime of PCF*(S ) in DMF to a rapid internal conversion to
1
1
PCF(S ) groundꢀstate; comparable examples are known for quenching of an excited state with
0
6
3
intramolecular CT character in polar solvents. Even with added benzophenone (1.0 M) as a
triplet sensitizer, TEA spectra of solutions of PCF in DMF showed no evidence for formation of
3
1
PCF*(T ) (see Figures S23 and S24). In contrast, the PCH*(S ) state has a lifetime in DMF
1
1
that is beyond the range of our experimental setup, and the growth of a new transient absorption
3
on the nanosecond timescale suggests competitive intersystem crossing to PCH*(T ), as in
1
DCM (Figure S20).
12
ACS Paragon Plus Environment

Page 13 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
2
2
1
1
5
0
5
0
5
0
1.0
0.5
0.0
1.0
0.5
(
a) TEAS
0.10 (b) TVAS
1PCF*(S1)
1PCF(S0)
1PCF*(S1)
1PCF*(S1)
1PCF*(S1)
0.0
0.05
ꢀ0.5
ꢀ1.0
0
400
800 1200
0
400
800 1200
Time / ps
Time / ps
0.00
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Time / ps
10
100
ꢀ0.05
4
7
1
00
00
200
1
PCF(S0)
ꢀ
0.10
4
00
450
500
550
600
1540
1560
1580
1600
ꢀ1
Wavelength / nm
Wavenumber / cm
Figure 3: Photochemical dynamics of PCF in DMF. Transient electronic (a) and vibrational (b) absorption
ꢀ3
spectra of PCF (2.1 mmol dm ) in DMF following pulsed photoexcitation at 370 nm are shown, alongside the
1
1
corresponding kinetic traces of PCF*(S ) and the ground state PCF(S ), obtained from integrating the fitted
1
0
1
1
band intensities at each time. Global fitting affords a single time constant for PCF*(S ) → PCF(S ). The decay
1
0
1
of PCF*(S ) is considerably faster than in DCM, indicating more efficient singlet quenching by DMF.
1
PCF is reported to be superior to PCH as a photocatalyst for controlling polymer molecular
4
0ꢀ41
weight and dispersity in OꢀATRP.
A robust mechanistic understanding of this selective
control of OꢀATRP must consider both the dissociative PET step which is the focus of this
study, and the loss of radicals by the TET process illustrated in Scheme 1, because a
combination of the two ET steps controls the radical concentration and hence the degree of
polymer dispersity. Under otherwise identical conditions, the combination of slower PET from
1
1
2
+
PC*(S ), and more energetically favourable TET (see Scheme 1) recovering PC(S ) from PC⋅
1
0
4
0
2
+
for PCF than for PCH, will result in lower PC⋅ and organic radical concentrations when PCF
is the chosen photocatalyst. The concentration of photoactive PC molecules is expected to
2
.+
ꢀ
decrease during polymerization because of the formation of ionic complexes ( PC Br ), with
further consequences for the steady state radical concentration. The equilibrium constant for the
4
3, 45
formation of these complexes will depend on the solvent and PC used.
Nevertheless, our
comparison of the photochemical dynamics of PCH and PCF in DCM and DMF, and the known
4
0
performances of the two PCs in OꢀATRP, lead us to suggest that the superior polymer
dispersity control resulting from PCF use is a consequence of lower steadyꢀstate radical
1
concentrations. We hypothesize that short-lived PC*(S ) states, low ISC quantum yields and
1
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
slower PETs can be desirable properties for photoredox organocatalysts in OꢀATRP if low
polymer dispersity is desired. In the case of PCF, these three photochemical characteristics work
in concert to supress the production of the reactive dehalogenated radical under constant
irradiation conditions, and low radical concentration is critical to controlled polymerization, as
4
8
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
discussed by Matyjaszewski and coworkers.
3
The proposed desirability of low ISC yields rests on the assumption that any populated PC*(T )
1
1
state will be considerably longerꢀlived than PC*(S ), and of sufficiently high energy that
1
1
electron transfer outꢀcompetes further ISC or phosphorescence back to the PC(S ) groundꢀstate.
0
3
Efficient, high quantumꢀyield ISC to a longꢀlived PC*(T ) will favor radical production at lower
1
initiator concentrations, leading to poorer control during polymerization. In accordance with
MarcusꢀSavéant theory, the PET rate can be moderated by seeking catalysts with comparatively
stable excited donor states and unstable radical cations. Moreover, if the excited state of a given
photocatalyst has CT character, the rate of PET can be slowed by employing solvents that
preferentially stabilise the excited emissive state.
Our analysis of PC properties applies specifically to OꢀATRP PC design, and is based solely on
evidence from study of the electron transfer rates for two of the N,Nꢀdiarylꢀ5,10ꢀ
dihydrophenazine PCs employed by Miyake and coꢀworkers. Quantitative understanding of the
optimum conditions for use of these and other PCs in OꢀATRP requires knowledge of the rate
coefficients for the numerous competing pathways and sequential steps in the polymerization
process. Empirical investigations are currently more tractable; for example, Ryan et al. recently
4
4
examined the optimization of irradiation intensity.
For ATRP using the metal complex
photocatalyst facꢀ[Ir(ppy) ], which is known to have an ISC quantum yield close to unity, Fors
3
and Hawker found that decreasing the PC load from 0.2 mol% to 0.005 mol% significantly
64
reduced the product polymer dispersity. The low PC loads in this study, compared to the 0.5
mol% values for organic PCs typical in the OꢀATRP system investigated by Miyake and coꢀ
workers, can be qualitatively explained by the different ET and ISC efficiencies of the two PCs:
the high yield of longꢀlived T states of facꢀ[Ir(ppy) ] maintains radical concentrations
1
3
appropriate for ATRP with low polymer dispersity at much lower PC concentrations than the
organic PCs which have smaller ISC quantum yields, and therefore favor ET from shorter lived
S1 states. This interpretation is supported by the reported ISC quantum yield of only 2% for
14
ACS Paragon Plus Environment

Page 15 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
65
N,Nꢀ5,10ꢀdi(2ꢀnaphthalene)ꢀ5,10ꢀdihydrophenazine,
as well as the ultrafast transient
absorption spectroscopy measurements we report here.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
PET and Solvent Effects
To further test the conclusions from our study, the effect of solvent on the rate of PET from the
PCF and PCH photocatalysts was investigated with transient electronic absorption spectroscopy.
ꢀ
The TEA spectra in Figure 4 were obtained for PCH and PCF solutions of [MBP] = 1.8 mol dm
3
in DCM, toluene and DMF. There is no evidence from these spectra that addition of the
1
3
bromineꢀcontaining MBP promotes ISC from the PC(S ) to the PC(T ) state. The time
1
1
constants τ′PET reported in the figure correspond to the pseudo firstꢀorder time constants for
electron transfer, obtained using monoꢀexponential fits for PCF and biꢀexponential fits for PCH,
1
as discussed earlier. In each case, the initial spectrum at 2 ps is assigned to the PC*(S ) state,
1
whereas the 1200ꢀps final spectrum corresponds to the absorption signature of the radical cation
2
.+
PC (D ), as confirmed by steadyꢀstate UVꢀVis characterization following thermally mediated
1
oxidation of PCF by FeCl (Figure S5 and S6).
3
1
The absorption signatures of the PC*(S ) states for the two photocatalysts are strikingly
1
different, with the broad absorption extending beyond 625 nm in the case of PCH tentatively
1
1
assigned as an excitation from PCH*(S ) to the CT A state ( PCH*(S ) in the FranckꢀCondon
1
u
3
region). For such a concentration of MBP, τ′PET is wellꢀapproximated by the observed time
1
constant for the decay of the PC*(S ) state, τ , with the exception of PCF in DMF. In this case,
1
obs
1
the decay of the PCF*(S ) state in the absence of MBP – quantified by the time constant τ =
1
1
6
77 ps (Figure 3) – is competitive with the PET, such that:
ꢞ
ꢟꢠꢡ
ꢤ
ꢝ′ꢄꢅꢆ
=
(3)
ꢟꢠꢡ
ꢢꢣ
ꢤ
ꢥ
For PCH, τ′PET is essentially invariant over the three solvents, which is consistent with electron
transfer being diffusionꢀlimited. Subtle variations are likely to be a consequence of small
modulations in the groundꢀstate redox potentials. In the case of PCF, τ′PET is inversely correlated
1
with the energy of the PCF*(S ) state [ꢀmax(Emission) = 585 nm (DMF), 562 nm (DCM), 508
1
2
.+
nm (toluene)]. The reduction potentials of PCF (D ) and MBP will also depend upon the
1
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
polarity of the solvent, but it is noteworthy that the observed trend in the PET rate is consistent
1
with the trend in the PCF*(S ) energy. Although solvent effects modulate the rate of PET in the
1
case of PCF, as is expected for an excited donor state of CT character, it is apparent that
substituent effects (ꢀH vs ꢀCF ) dominate the photochemical dynamics.
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
20
PCF / DMF
Time / ps
PCF / DCM
PCF / Toluene
20
τʹ
= 198 ± 7 ps
2
10
50
τʹ
= 170 ± 7 ps
τʹ
= 135 ± 7 ps
1
0
5
0
PET
max
PET
max
PET
max
λe
= 585 nm
λe
= 562 nm
λe
= 508 nm
3
1000
00
10
10
0
0
PCH / DMF
τʹPET = 28 ± 3 ps
PCH / DCM
τʹPET = 28 ± 4 ps
PCH / Toluene
τʹPET = 23 ± 3 ps
20
20
10
0
1
5
0
5
0
1
10
0
400
450
500
550
600
400
450
500
550
600
400
450
500
550
600
Wavelength / nm
Wavelength / nm
Wavelength / nm
Figure 4: Solvent effects on the photochemical dynamics of PCF (top) and PCH (bottom) in the presence of MBP
initiator after pulsed photoexcitation at 370 nm. DMF (left), DCM (middle) and toluene (right) solutions were
ꢀ3
ꢀ3
examined, with PCF or PCH (2.1 mmol dm ) and MBP (1.8 mol dm ). The wavelength of maximum emission for
e
1
PCH
λ
= 480 nm remains largely independent of solvent, due to the LE character of the PCH*(S ) state,
max
1
e
1
whereas
λ
in the case of PCF is sensitive to solvent polarity due to the CT character of PCF*(S ). The time
max
1
constants shown for PCH correspond to the slower k′PET(Dif) components obtained by biexponential fits, as
discussed in the main text.
Conclusions
The photochemical dynamics of initiation in OꢀATRP have been probed with transient electronic
and vibrational absorption spectroscopy, using two complementary N,Nꢀdiarylꢀ5,10ꢀ
dihydrophenazine photoredox organocatalysts (PCF, PCH) and methyl 2ꢀbromopropionate
(MBP) as a representative radical initiator. For PCF, we report direct spectroscopic observation
1
of a bimolecular dissociative photoinduced electron transfer from PCF*(S ) to MBP in DCM,
1
1
1
2
.+
2
.
with PCF(S ), PCF*(S ), PCF (D ) and MP (D ) tracked simultaneously in realꢀtime from the
0
1
1
1
temporal evolution of their absorption signatures. Global fitting of the kinetics, and the absence
2
.ꢀ
of a distinct transient absorption for the geminate radical anion MBP (D ), indicate that electron
1
16
ACS Paragon Plus Environment

Page 17 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
transfer and bromide dissociation steps cannot be distinguished on our timescale. Studies of
MBP concentration dependence with TVAS and TEAS in DCM afford a mean bimolecular rate
9
3
ꢀ1 ꢀ1
3
coefficient of kPET(PCF) = (4.2 ± 0.4) × 10 dm mol s . No spectroscopic trace of PCF*(T )
1
was detected within the temporal range of the experiment (1.3 ns) in either DCM or DMF; in the
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
absence of MBP, the recovery of PCF(S ) accounts completely for the decay of PCF*(S ). The
0
1
1
lifetime of the PCF*(S ) state is particularly short in DMF (677 ± 35 ps), which is significant as
1
this represents a typical solvent for OꢀATRP. In accordance with MarcusꢀSavéant theory, the rate
1
of PET to MBP is activationꢀcontrolled and varies with the energy of the PCF*(S ) state, which
1
is sensitive to solvent polarity on account of its intramolecular CT character.
10
3
Under the same conditions, a bimolecular rate coefficient of kPET(PCH) = (2.0ꢂ± 0.2) × 10 dm
ꢀ1 ꢀ1
1
mol s was determined for PET from the PCH*(S ) state to MBP. This is ~5 times faster than
1
the PCF and comparable to the diffusionꢀlimited rate coefficient in DCM under ambient
conditions. Contrary to the postulate of Miyake and coꢀworkers, we thus find that PCH
1
undergoes faster PET than PCF, despite the local excitation character of the PCH*(S ) state.
1
The disparity in the rates of PET is more reliably explained with MarcusꢀSavéant theory, given
the substantial difference in the estimated PET driving force of ꢁꢁPETG = ꢁPETG(PCF) -
ꢀ1
3
ꢁPETG(PCH)
≈ 51 kJ mol . In PCH, ISC to the first excited triplet state PCH*(T ) is
1
competitive with fluorescence in both DCM and DMF; at the lower concentrations of MBP
ꢀ3
employed under synthetic conditions (0.05 – 0.10 mol dm ), therefore, it seems probable that
3
electron transfer from PCH*(T ) is significant. Unlike PCF, the rate of PET remains
1
independent of the solvent polarity, consistent with a diffusionꢀlimited PET from an excited state
of PCH of local excitation character.
Comparison of the photochemical dynamics of PCF and PCH suggests that a combination of
1
shorter PC*(S ) lifetimes, low ISC quantum yields and slower PETs favors control of polymer
1
molecular weight and dispersity in OꢀATRP using these N,Nꢀdiarylꢀ5,10ꢀdihydrophenazines as
photocatalysts, by supressing the steadyꢀstate concentration of the dehalogenated radical. With
this in mind, it would be prudent for synthetic chemists to consider the design of organocatalysts
for OꢀATRP that are characterized by short fluorescence lifetimes, lowꢀlying excited donor states
and high energy radical cations in a given solvent if control of polymer dispersity is a primary
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
objective. Our results further suggest PET rates for different photocatalysts are better interpreted
in terms of the free energies of excited states than their CT/LE electronic characters.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Acknowledgements
Financial support for this research from ERC Advanced Grant CAPRI 290966 is gratefully
acknowledged. We thank Philip Coulter (University of Bristol) for help with the acquisition of
some of the transient spectroscopy data.
Associated Content
Additional analysis, steadyꢀstate absorption and emission spectra, computational methodology,
photocatalyst syntheses and TEAS/TVAS spectra and kinetics and are located in the Electronic
Supporting Information.
Experimental Data
All experimental data are archived in the University of Bristol’s Research Data Storage Facility
(DOI 10.5523/bris.1wp4y75121tre2728hptf1hk55).
18
ACS Paragon Plus Environment

Page 19 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
References
1. Nicewicz, D. A.; MacMillan, D. W. C., Science 2008, 322, 77ꢀ80.
2
3
. Ischay, M. A.; Anzovino, M. E.; Du, J.; Yoon, T. P., J. Am. Chem. Soc. 2008, 130, 12886.
. Narayanam, J. M. R.; Tucker, J. W.; Stephenson, C. R. J., J. Am. Chem. Soc. 2009, 131,
8
756.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
5
. Prier, C. K.; Rankic, D. A.; MacMillan, D. W. C., Chem. Rev. 2013, 113, 5322ꢀ5363.
. Shaw, M. H.; Twilton, J.; MacMillan, D. W. C., J. Org. Chem. 2016, 81, 6898ꢀ6926.
6. Karkas, M. D.; Porco, J. A.; Stephenson, C. R. J., Chem. Rev. 2016, 116, 9683ꢀ9747.
7. Nicholls, T. P.; Leonori, D.; Bissember, A. C., Nat. Prod. Rep. 2016, 33, 1248ꢀ1254.
8
9
1
1
1
. Staveness, D.; Bosque, I.; Stephenson, C. R. J., Acc. Chem. Res. 2016, 49, 2295ꢀ2306.
. Chen, J. R.; Hu, X. Q.; Lu, L. Q.; Xiao, W. J., Acc. Chem. Res. 2016, 49, 1911ꢀ1923.
0. Yoon, T. P., Acc. Chem. Res. 2016, 49, 2307ꢀ2315.
1. Ghosh, I.; Marzo, L.; Das, A.; Shaikh, R.; Konig, B., Acc. Chem. Res. 2016, 49, 1566ꢀ1577.
2. Goddard, J. P.; Ollivier, C.; Fensterbank, L., Acc. Chem. Res. 2016, 49, 1924ꢀ1936.
13. Ravelli, D.; Protti, S.; Fagnoni, M., Chem. Rev. 2016, 116, 9850ꢀ9913.
14. Takeda, H.; Ishitani, O., Coord. Chem. Rev. 2010, 254, 346ꢀ354.
15. Lang, X. J.; Zhao, J. C.; Chen, X. D., Chem Soc Rev 2016, 45, 3026ꢀ3038.
16. Skubi, K. L.; Blum, T. R.; Yoon, T. P., Chem. Rev. 2016, 116, 10035ꢀ10074.
17. Hopkinson, M. N.; Sahoo, B.; Li, J. L.; Glorius, F., Chem. Eur. J. 2014, 20, 3874ꢀ3886.
18. Hopkinson, M. N.; TlahuextꢀAca, A.; Glorius, F., Acc. Chem. Res. 2016, 49, 2261ꢀ2272.
19. Levin, M. D.; Kim, S.; Toste, F. D., ACS Cent. Sci. 2016, 2, 293ꢀ301.
20. Luo, J.; Zhang, J., ACS Catal. 2016, 6, 873ꢀ877.
2
2
2
1. Terrett, J. A.; Cuthbertson, J. D.; Shurtleff, V. W.; MacMillan, D. W. C., Nature 2015, 524,
30ꢀ334.
2. Zoller, J.; Fabry, D. C.; Ronge, M. A.; Rueping, M., Angew. Chem. Int. Edit. 2014, 53,
3264ꢀ13268.
3. Tsarevsky, N. V.; Matyjaszewski, K., Chem. Rev. 2007, 107, 2270ꢀ2299.
3
1
24. AriasꢀRotondo, D. M.; McCusker, J. K., Chem. Soc. Rev. 2016, 45, 5803ꢀ5820.
25. Romero, N. A.; Nicewicz, D. A., Chem. Rev. 2016, 116, 10075ꢀ10166.
2
2
2
2
3
6. Margrey, K. A.; Nicewicz, D. A., Acc. Chem. Res. 2016, 49 (9), 1997ꢀ2006.
7. Hari, D. P.; Konig, B., Chem. Comm. 2014, 50, 6688ꢀ6699.
8. Fukuzumi, S.; Ohkubo, K., Chem. Sci. 2013, 4, 561ꢀ574.
9. Ohkubo, K.; Fujimoto, A.; Fukuzumi, S., J. Phys. Chem. A 2013, 117, 10719ꢀ10725.
0. Ohkubo, K.; Mizushima, K.; Iwata, R.; Fukuzumi, S., Chem. Sci. 2011, 2, 715ꢀ722.
31. Romero, N. A.; Margrey, K. A.; Tay, N. E.; Nicewicz, D. A., Science 2015, 349, 1326ꢀ1330.
32. McManus, J. B.; Nicewicz, D. A., J. Am. Chem. Soc. 2017, 139, 2880ꢀ2883.
3
3
3
3. Romero, N. A.; Nicewicz, D. A., J. Am. Chem. Soc. 2014, 136, 17024ꢀ17035.
4. Pitre, S. P.; McTiernan, C. D.; Ismaili, H.; Scaiano, J. C., ACS Catal. 2014, 4, 2530ꢀ2535.
5. Pitre, S. P.; McTiernan, C. D.; Ismaili, H.; Scaiano, J. C., J. Am. Chem. Soc. 2013, 135,
1
3286ꢀ13289.
3
6. Ghosh, T.; Das, A.; Konig, B., Org. Biomol. Chem. 2017, 15, 2536ꢀ2540.
37. Meyer, A. U.; Wimmer, A.; Konig, B., Angew. Chem. Int. Edit. 2017, 56, 409ꢀ412.
3
3
8. Treat, N. J.; Sprafke, H.; Kramer, J. W.; Clark, P. G.; Barton, B. E.; de Alaniz, J. R.; Fors, B.
P.; Hawker, C. J., J. Am. Chem. Soc. 2014, 136, 16096ꢀ16101.
9. Miyake, G. M.; Theriot, J. C., Macromolecules 2014, 47, 8255ꢀ8261.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 21
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
0. Theriot, J. C.; Lim, C. H.; Yang, H.; Ryan, M. D.; Musgrave, C. B.; Miyake, G. M., Science
2
016, 352, 1082ꢀ1086.
4
1. Pearson, R. M.; Lim, C. H.; McCarthy, B. G.; Musgrave, C. B.; Miyake, G. M., J. Am.
Chem. Soc. 2016, 138, 11399ꢀ11407.
42. Wang, J. S.; Matyjaszewski, K., J. Am. Chem. Soc. 1995, 117, 5614ꢀ5615.
43. Lim, C. H.; Ryan, M. D.; McCarthy, B. G.; Theriot, J. C.; Sartor, S. M.; Damrauer, N. H.;
Musgrave, C. B.; Miyake, G. M., J. Am. Chem. Soc. 2017, 139, 348ꢀ355.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
4. Ryan, M. D.; Pearson, R. M.; French, T. A.; Miyake, G. M., Macromolecules 2017, 50,
616ꢀ4622.
4
4
5. Ryan, M. D.; Theriot, J. C.; Lim, C. H.; Yang, H. S.; Lockwood, A. G.; Garrison, N. G.;
Lincoln, S. R.; Musgrave, C. B.; Miyake, G. M., J. Polym. Sci. Pol. Chem. 2017, 55, 3017ꢀ
3027.
6. Discekici, E. H.; Treat, N. J.; Poelma, S. O.; Mattson, K. M.; Hudson, Z. M.; Luo, Y. D.;
Hawker, C. J.; de Alaniz, J. R., Chem. Commun. 2015, 51, 11705ꢀ11708.
7. Jockusch, S.; Yagci, Y., Polym. Chem. 2016, 7, 6039ꢀ6043.
4
4
4
8. Pan, X. C.; Fang, C.; Fantin, M.; Malhotra, N.; So, W. Y.; Peteanu, L. A.; Isse, A. A.;
Gennaro, A.; Liu, P.; Matyjaszewskit, K., J. Am. Chem. Soc. 2016, 138, 2411ꢀ2425.
49. Kumpulainen, T.; Lang, B.; Rosspeintner, A.; Vauthey, E., Chem. Rev. 2017, 117, 10826ꢀ
10939.
5
5
5
0. Dereka, B.; Koch, M.; Vauthey, E., Acc. Chem. Res. 2017, 50, 426ꢀ434.
1. Rosspeintner, A.; Angulo, G.; Vauthey, E., J. Am. Chem. Soc. 2014, 136, 2026ꢀ2032.
2. Sun, Q. C.; Dereka, B.; Vauthey, E.; Daku, L. M. L.; Hauser, A., Chem. Sci. 2017, 8, 223ꢀ
2
30.
5
3. Kellogg, R. E.; Simpson, W. T., J. Am. Chem. Soc. 1965, 87, 4230ꢀ&.
54. Montali, M.; Credi, A.; Pradi, L.; Gandolfi, M. T., Handbook of Photochemistry. 3rd ed.;
CRC Press: Boca Raton, 2006.
5. Gorman, A. A.; Gould, I. R.; Hamblett, I., J. Am. Chem. Soc. 1981, 103, 4553ꢀ4558.
6. Evans, D. F., J. Chem. Soc. 1957, 1351ꢀ1357.
7. Ni, T.; Caldwell, R. A.; Melton, L. A., J. Am. Chem. Soc. 1989, 111, 457ꢀ464.
8. Isse, A. A.; Gennaro, A.; Lin, C. Y.; Hodgson, J. L.; Coote, M. L.; Guliashvili, T., J. Am.
Chem. Soc. 2011, 133, 6254ꢀ6264.
5
5
5
5
59. Morris, J. V.; Bruhlman, U.; Serafimov, O.; Huber, J. R., Ber Bunsen Phys Chem 1974, 78,
348ꢀ1353.
6
6
6
1
0. Saveant, J. M., J. Am. Chem. Soc. 1987, 109, 6788ꢀ6795.
1. Rehm, D.; Weller, A., Berich. Bunsen. Gesell. 1969, 73, 834.
2. Cardinale, A.; Isse, A. A.; Gennaro, A.; Robert, M.; Saveant, J. M., J. Am. Chem. Soc. 2002,
1
24, 13533ꢀ13539.
63. Dereka, B.; Rosspeintner, A.; Li, Z. Q.; Liska, R.; Vauthey, E., J. Am. Chem. Soc. 2016,
138, 4643ꢀ4649.
6
4. Fors, B. P.; Hawker, C. J., Angew. Chem. Int. Edit. 2012, 51, 8850ꢀ8853.
5. Du, Y.; Pearson, R. M.; Lim, C. H.; Sartor, S. M.; Ryan, M. D.; Yang, H. S.; Damrauer, N.
H.; Miyake, G. M., Chem. Eur. J. 2017, 23, 10962ꢀ10968.
6
20
ACS Paragon Plus Environment

Page 21 of 21
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Table of Contents Graphic
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
ACS Paragon Plus Environment