Mechanistic Investigations of an α-Aminoarylation Photoredox Reaction

Bernard G. Stevenson, Ethan H. Spielvogel, Emily A. Loiaconi, Victor Mulwa Wambua,

Article

Mechanistic Investigations of an α‑Aminoarylation Photoredox

Reaction

Roman V. Nakhamiyayev, and John R. Swierk*

Cite This: J. Am. Chem. Soc. 2021, 143, 8878 −8885

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: While photoredox catalysis continues to transform

modern synthetic chemistry, detailed mechanistic studies involving

direct observation of reaction intermediates and rate constants are

rare. By use of a combination of steady state photochemical

measurements, transient laser spectroscopy, and electrochemical

methods, an α-aminoarylation mechanism that is the inspiration for

a large number of photoredox reactions was rigorously charac-

terized. Despite high product yields, the external quantum yield

(

QY) of the reaction remained low (15−30%). By use of transient

absorption spectroscopy, productive and unproductive reaction

pathways were identiﬁed and rate constants assigned to develop a

comprehensive mechanistic picture of the reaction. The role of the

cyanoarene, 1,4-dicyanobenzne, was found to be unexpectedly complex, functioning both as initial proton acceptor in the reaction

and as a neutral stabilizer for the 1,4-dicyanobenzene radical anion. Finally, kinetic modeling was utilized to analyze the reaction at

an unprecedented level of understanding. This modeling demonstrated that the reaction is limited not by the kinetics of the

individual steps but instead by scattering losses and parasitic absorption by a photochemically inactive donor−acceptor complex.

INTRODUCTION

diﬃcult to achieve thermally. Typically, photoredox reactions

proceed via single electron transfer (SET) to a target substrate,

which then generates one or more radical species involved in

the bond-forming step. Examples of photoredox reactions

involving energy transfer instead of electron transfer are also

■

Photoredox catalysis allows for the activation of highly stable

bonds by utilizing visible light. Since the ﬁrst simultaneous

reports of photoredox from MacMillan et al. and Yoon et al.,

−7

the scope of photoredox reactions has expanded greatly. Yet

while the development of photoredox methods has accelerated

at a remarkable pace, a comprehensive mechanistic under-

standing has lagged behind. Most reports oﬀer varying degrees

of mechanistic insights through a combination of Stern−

Volmer measurements, redox potential measurements, bond

dissociation energies, and variation in reagent concentra-

known. Importantly, visible light allows reactions to occur

without the use of harsh reaction conditions or reagents that

are classically used in synthetic transformations.

While mechanistic information about the initial steps of

photoredox reactions is common, complete characterizations

of reaction mechanisms and kinetics are rare. Nocera and co-

workers used a combination of spectroscopic, electrochemical,

and computational methods to fully characterize the catalytic

cycle and rate constants of a hydroamidation reaction and, with

that information, were able to improve the low quantum yield

QY) of the reaction. Orr-Ewing and co-workers were able to

determine the rate constants for electron transfer (ET) and

radical propagation steps for an atom transfer radical

−11

tion.

While this can give information about the initial steps

in the reaction, subsequent steps are less well characterized. A

complete mechanistic understanding involves direct knowledge

of intermediates as well as the kinetics of the productive and

unproductive steps. This information can then be used to

(

2,13

improve the reaction yields and reaction completion times.

Despite a wide variation in the reaction substrates and

ancillary reagents, all photoredox methods rely on a photo-

catalyst to absorb light and initiate the reaction. This

photocatalyst is often an organometallic complex with long-

lived excited states, though there are many examples using

18,19

polymerization using transient IR techniques.

Romero

Received: April 7, 2021

Published: June 2, 2021

,14,15

purely organic photocatalysts.

Once a photon is

absorbed, the excited state of the photocatalyst can function

as either a potent oxidant or reductant, which allows the

photocatalyst to generate high energy intermediates that are

2021 American Chemical Society

J. Am. Chem. Soc. 2021, 143, 8878−8885

878

Journal of the American Chemical Society

Article

and Nicewicz characterized an alkene hydrofunctionalization

reaction, observing transient radical intermediates and kinetic

information related to a hydrogen atom transfer cycle.

METHODS

■

Fac-tris(2-phenylpyridine C2,N)iridium(III) (Ir(ppy) ), N,N-dime-

thylactemide (DMA), (4,4′-di-tert-butyl-2,2′-bipyridine)bis(2-

Martinez-Haya and co-workers evaluated both the thermody-

namics and kinetics for the reductive dehalogenation of several

brominated substrates using riboﬂavin, ﬁnding good agreement

between thermodynamic predictions and measured rate

phenylpyridine)iridium(III) hexaﬂuorophosphate (Ir(dtbbpy)ppy ),

sodium acetate (NaOAc), and 1,4-dicyanobenzene (DCB) were

purchased from Sigma-Aldrich. N-Phenylpyrrolidine (NPP) was

purchased from Alfa Aesar and used as received. DMA was dried

over molecular sieves, and sodium acetate was dried at 100 °C before

use. DCB and sodium acetate were also crushed with a mortar and

pestle. All other reagents were used as received.

Quantum Yield (QY) Measurements. A stirring ﬂea was placed

into a screwtop 1 cm path length cuvette along with Ir(ppy) (2.5

μmol, 0.005 equiv), DCB (0.5 mmol, 1 equiv), and sodium acetate

1.0 mmol, 2 equiv). The solvent, DMA (2 mL), was purged for 30

min with argon before being added to the cuvette via syringe along

with NPP (1.5 mmol, 3 equiv). The complete solution was then

bubbled with argon in the dark for an additional 45 min. For reactions

that varied in light intensity, the cuvette was placed on a stirring plate

with a 3D-printed cuvette holder in front of a collimated 415 nm LED

Thor Labs M15LP1) for a speciﬁed length of time (1−30 h). The

LED light intensity was measured using a calibrated photodiode

Thor Labs S120C). For wavelength-dependent reactions, the cuvette

was placed on a stir plate in front of a 950 W Xe arclamp (Oriel

6921) equipped with a monochromator (Spectral Products CM110)

constants. Several other groups have utilized transient

absorption spectroscopy to observe reaction intermediates in

22−30

photoredox reactions.

Substituted nitrogen heterocycles and α-arylamines are

important structural motifs in medicinal and pharmaceutical

1,35

chemistry.

Generation of these motifs using latent sp C−

(

H bonds via cross-coupling of amines and aryl building blocks

2,33

has attracted signiﬁcant attention,

reactions representing a particularly attractive approach.

The pioneering work on the photoredox generation of α-

with photoredox

arylamines was ﬁrst reported by MacMillan and co-workers,

(

though the use of electron-deﬁcient arenes and α-amino

radicals was subsequently generalized for a host of other

(

6−44

photoredox transformations

and, more recently, to

electrosynthetic chemistry. All of these reactions build on

the mechanism ﬁrst proposed by MacMillan et al., using a

prototypical coupling of 1,4-dicyanobenzene (DCB) and N-

phenylpyrrolidine (NPP) to generate 4-(1-phenyl-2-

pyrrolidinyl)benzonitrile (Scheme 1). On the basis of Stern−

for 2 h. After illumination, 0.25 mmol of triphenylmethane was added

as an internal standard and the reaction was stirred for 30 min in the

dark. 100−200 μL of reaction mixture was then dissolved in d-

acetonitrile and the reaction yield calculated with quantitative H

nuclear magnetic resonance with a Bruker 400 MHz NMR. The

external QY was then calculated according to the following equation:

Scheme 1. α-Aminoarylation Reaction Scheme

moles of product

quantum yield =

moles of incident photons

Transient Absorption Spectroscopy (TAS) Experimentation.

Transient absorption experiments were carried out using a Spectra-

Physics Quanta-Ray Pro-290 pulsed Nd:YAG laser (10 Hz) ﬁtted

with a PrimoScan OPO. An excitation wavelength of 415 nm (900 μJ/

Volmer emission quenching studies, redox potentials, and

bond dissociation energies, they proposed that, upon excitation

of an iridium photocatalyst, an electron is transferred to DCB

cm ) was used for all experiments. Laser pulses were chopped at every

other pulse to improve the signal-to-noise ratio per the method of

•

−

Rimshaw et al. The sample was illuminated with a broadband white

light source (Energetiq EQ-99X), with a shutter before the sample to

minimize light exposure. After the sample, probe light was collected

by a monochromator (Spectral Products DK240) and passed onto a

silicon photodiode (ThorLabs DET10A). Data were collected with a

Pico Technology 6404C oscilloscope and analyzed using software

written in LabView.

to generate a radical anion (DCB ) and Ir(IV) species. The

Ir(IV) species subsequently oxidizes NPP to generate the NPP

•

radical cation (NPP ) and regenerate the ground state

photocatalyst. In their mechanistic proposal, this NPP radical

cation is then deprotonated by sodium acetate to give the

•

•−

neutral NPP radical (NPP ), which couples with DCB to

form the target product.

Single wavelength traces were collected at 6.4 ns intervals up to 12

μs and at 1 μs intervals up to 10 ms. The short-time and long-time

data traces were stitched together before ﬁtting. For short-time traces,

data were collected with the probe on and oﬀ to remove any residual

laser scattering. For TAS experiments, solution concentrations of 37

Despite the importance of α-arylamines and related

photoredox reactions, the reaction mechanism and kinetics

are not well characterized. Recently, Walker and co-workers

examined the coupling of DCB and various substituted

μM Ir(ppy) , 50 mM DCB, and 150 mM NPP in DMA were used.

piperidines. They determined the rate constant for back

Solutions were prepared under an argon atmosphere for 90 min in a

four-sided screw top cuvette with cap and septum before TAS

experiments. Samples were changed every 2 h with stability conﬁrmed

by comparing single wavelength traces at the same wavelength

collected at diﬀerent times throughout the experiment. The TAS

traces were ﬁt to a kinetic model, described in the Supporting

Information.

Spectrochemical Studies. All spectroelectrochemical experi-

ments were performed using a BioLogic SP-50 potentiostat, a

platinum honeycomb spectroelectrochemical cell (Pine) with a path

•

−

electron transfer between DCB and [Ir(ppy) ] , as well as

the rate constant for the oxidation of the piperidine by

[

Ir(ppy) ] , but the kinetics of subsequent steps were left

unexplored. In this work, a combination of reaction QY

measurements, transient absorption spectroscopy (TAS), and

electrochemistry was utilized to characterize productive and

unproductive pathways in the catalytic coupling of DCB and

NPP and assign rate constants to all steps. These measure-

ments reveal a signiﬁcantly more complex reaction mechanism

than previously suspected, speciﬁcally in regard to DCB.

Kinetic modeling of the reaction indicates that the QY of the

reaction is not limited by the kinetics of the reaction but

instead by both scattering and parasitic absorption by a

photochemically inactive donor−acceptor complex.

^l^eⁿ^g^t^h^o^f¹^.⁷^m^m^,^aⁿ^d^a^S^hⁱ^m^a^d^z^u^U^V^-²⁶⁰⁰^U^V⁻^vⁱ^s^s^p^e^c^t^r^o^p^h^o+^-

tometer. Electrochemical potentials were applied relative to a Ag/Ag

reference electrode. For all spectroelectrochemical studies, 0.1 M

tetrabutylammonium hexaﬂuorophosphate (TBAPF ) in DMA was

used as the electrolyte. Concentrations of 83 and 590 μM were used

for Ir(ppy) and DCB, respectively. Spectroelectrochemical studies of

879

J. Am. Chem. Soc. 2021, 143, 8878−8885

Journal of the American Chemical Society

Article

NPP failed to produce a stable spectrum for the radical cation and

were not pursued.

RESULTS AND DISCUSSION

QY Measurements. At an excitation wavelength of 415

nm, a QY was observed in the range of 0.15 −0.3 (Figure 1),

■

Figure 2. Transient absorption diﬀerence spectra measured at

indicated delay times after 415 nm excitation pulse with 37 μM

Figure 1. Percent yield (black circles) and external QY (orange

squares) of reaction illuminated by a 415 nm LED (10.4 mW cm ).

Ir(ppy) and 50 mM DCB. At 100 ns, the bleach at 380 nm and

−

changes in absorption at wavelengths of >460 nm were assigned to the

generation of [Ir(ppy) ] , while the 430 nm peak was assigned to

•‑

generation of DCB radical, DCB .

depending on the length of the reaction, which is consistent

with a reaction that does not proceed through a radical chain

pathway. After 24 h, the reaction reached a maximum percent

yield of 88% when measured by quantitative NMR (Figures S1

and S2 ) and was independent of wavelength (Figure S3). It

was observed that the reaction rate was the same at high (10.4

mW) and low (5.06 mW, Figure S4) light intensities when

normalized for photon ﬂux.

in the TAS spectrum, there were positive absorption features at

wavelengths longer than 460 nm and a bleach centered at 380

nm. Both features are consistent with the formation of Ir(IV)

(Figure S7). There was also a new absorption at ∼430 nm that

was assigned to the DCB radical anion (Figure S8). At 100 ns,

the transient spectrum could be reproduced by combining the

diﬀerence spectra obtained from the spectroelectrochemical

A reaction QY of 0.15−0.3 is in good agreement with other

studies on nonradical chain propagation reactions, where

typical QYs of photoredox reactions can range from 0.19 to

•−

spectrum for [Ir(ppy) ] and DCB . From this, an initial

concentration of 3.55 μM was determined for both species

(Figure S9).

8−50

.43.

Ellman and co-workers partly explored the

mechanism of a closely related α-aminoarylation reaction

Unexpectedly, at longer times (>1 μs) changes were

observed in the transient spectra that are consistent with the

formation of a new species. Most notably, a new absorption at

∼860 nm formed on a tens of microseconds time scale that

involving substituted piperidine analogues and observed a QY

of 0.4−0.6 at early time scales, which is in relatively good

agreement with the QY in this study. These results were

collected without the use of insoluble sodium acetate, which

actinometry experiments suggest scatter 10−20% of incoming

photons.

•−

cannot be assigned to either [Ir(ppy) ] or DCB . In

addition, the bleach at 380 nm was replaced by a new

absorption, and the absorption at 600 nm blueshifted and

changed shape. These new absorbances were assigned to the

formation of a new species, which was formulated as

(DCB)₂^•. Increasing the concentration of DCB, while

maintaining a constant pump ﬂuence, led to a larger change

in absorbance (Figure S10). Fits of single wavelength traces at

diﬀerent concentrations demonstrated a second order reaction

dependence on the concentration of DCB^•and neutral DCB,

and each trace could be ﬁt using the same set of kinetic

parameters. In addition, spectroelectrochemistry at high

concentrations of DCB revealed an absorption feature that

peaked at 860 nm and was not present at lower concentrations

of DCB (Figure S11). While further studies are planned to

elucidate the nature of this intermediate, previous work

suggests that the solution chemistry of cyanoarene radical

Over the course of the reaction, the QY showed signiﬁcant

variation. The gradual decrease in QY from 3+ hours was

assigned to the formation of acetic acid. NMR data (Figure S5)

showed that acetic acid is formed from sodium acetate over the

course of the reaction (vide infra), which eventually competes

with DCB as an electron acceptor for the excited photocatalyst.

A quenching rate constant was determined for acetic acid of

−

−1 −1

.8 × 10 M s , which is comparable to the quenching rate

−1 −1

constant for DCB (2.2 × 10 M s ). Furthermore, TAS data

demonstrated that, in the presence of excess acetic acid, charge

transfer to DCB does did not occur (Figure S6).

It is possible that the increase in QY over the ﬁrst 3 h was

caused by a decrease in scattering by sodium acetate. As

sodium acetate was converted to acetic acid, the particle size

may have decreased, which would lead to a decrease in

scattering and an increase in QY. By comparison, when the

soluble base tetrabutylammonium acetate was used, the QY

was ∼52% at 30 min and then decreased to 42% at 1 h.

anions can involve multiple species, including dimers.

In order to understand the kinetics of electron transfer

•−

between Ir(ppy) , DCB, and (DCB) , a kinetic model was

developed that incorporates back electron transfer between

Ir(IV) and DCB^•(k_r_e_c_o_m_b), pairing of DCB and DCB (k_p_a_i_r),

−

•−

Electron Transfer between DCB and Ir(ppy) . TAS data

•

−

were collected from 100 ns to 10 ms and from 340 to 1100 nm

and electron transfer between Ir(IV) and (DCB)₂(k_r_e_c_o_m_b₂)

(

Figure 2) to examine the rate at which the DCB radical anion

and applied to the single wavelength absorption traces (Figure

3, Table S1 ). An average rate constant for recombination,

can undergo back electron transfer (BET) with Ir(IV). Initially,

880

Journal of the American Chemical Society

J. Am. Chem. Soc. 2021, 143, 8878−8885

Article

Figure 3. Single wavelength traces of 3.7 μM Ir(ppy) and 50 mM DCB at 360 nm, 620 nm, and 860 nm at 415 nm excitation (0.9 mJ/cm per

pulse, gray). Orange line is the ﬁt to a kinetic model.

k_r_e_c_o_m_b, of 5.4 (±0.7) × 10 M s⁻¹was calculated. Though

−1

kinetic modeling can be used to estimate a minimum value of

−1 −1

this recombination is nearly diﬀusion controlled, the low

concentration of Ir(IV) and DCB means it is not the

dominant pathway. Instead, the model suggests that DCB

forms (DCB) with a rate constant of 1.1 (±0.1) × 10 M

s . Back electron transfer then occurs between (DCB ) and

Ir(IV) (k_r_e_c_o_m_b₂= 6.0 (±0.6) × 10 M s ). The absorption

^s ^p ^e ^c ^t ^r ^u ^m^f^o^r⁽^D^C^B⁾2 was determined from these ﬁts (Figure

k as 4 × 10 M s . From ﬁtting the transient spectrum at

•

−

•−

100 ns, an initial concentration of 3.1 μM of DCB and

NPP (Figure S14 ) was estimated. Though the pump ﬂuence

•

−

•+

•

−

−1

is constant, this decrease in the concentration of the charge-

separated state is likely due to formation of a photochemically

inactive donor−acceptor complex between DCB and NPP that

parasitically absorbs at 415 nm (Figure S15) but does not

exhibit any photochemical response in the TAS. Also,

−1

•−

−1 −1

•

−

S12).

^E^l^e^c^t^r^oⁿ^T^r^aⁿ^s^f^e^r^b^e^t^w^e^eⁿ^D^C^B^,^N^P^P^,^aⁿ^d^I^r⁽^p^p^y⁾3^.

Following TAS studies with only DCB and Ir(ppy) , NPP was

then included in the reaction mixture. It was immediately

obvious that the transient spectra were remarkably diﬀerent

with NPP added (Figure 4 ). On short time scales (<10 μs)

illumination of the reaction without Ir(ppy) does not result

in product formation.

At 10 μs and longer, the transient spectra change

dramatically. Most notably, a large transient absorption

develops in the near IR from 700 to 1100 nm, as well as an

increase in absorption between 400 and 600 nm. While, to the

•

best of our knowledge, the absorption spectrum for NPP is

not available, these absorption features are generally consistent

2,53

with alkane radicals.

Direct oxidation of NPP with

Ir(dtbbpy)(ppy) generates the same species (Figure S16).

Single wavelengths traces (Figure 5) conﬁrmed the

formation of a new species on a 10−100 μs time scale,

followed by decay over milliseconds. These traces could not be

ﬁt using a simple exponential model, so the kinetic model used

in the Ir(ppy) /DCB experiments was expanded to include

•+

•

deprotonation of NPP , k

DCB) , k_c_o_u_p_l_e. Though coupling between NPP and

DCB was initially included, the curve ﬁtting consistently

set the rate constant for this coupling to zero, suggesting that

DCB^•does not undergo coupling. Also, while not included in

, and coupling of NPP and

deprot

•

−

•+

(

•

−

the original mechanistic proposal, it was found that the data

could not be ﬁt unless terms were included that described

•

•−

unproductive electron transfer between NPP and DCB ,

k_r_a_d_r_e_c_o_m_b, and (DCB)₂^•, k_r_a_d_r_e_c_c_o_m_b₂, which resulted in the

regeneration of DCB and NPP. By use of the expanded model,

the TAS traces were ﬁt and rate constants extracted (Figure 6,

Table S2).

−

Figure 4. Transient absorption diﬀerence spectra measured at

indicated delay times after 415 nm excitation pulse with 35 μM

Ir(ppy) , 50 mM DCB, and 150 mM NPP. At 100 ns, the broad

absorption between 400 and 600 nm was assigned to the formation of

The kinetic scheme in Figure 6 reveals several key features

about this reaction. First, the slowest step is the deprotonation

•

•‑

NPP and the peak at 440 nm to DCB . No features were observed

that could be assigned to [Ir(ppy) ] .

•+

of NPP , which serves as a necessary precursor to coupling.

While the reaction waits for deprotonation to occur,

unproductive electron transfer between the radical ions can

occur, leading to a decrease in QY. It is particularly interesting

that this unproductive step is nearly an order of magnitude

there was a broad absorbance from 400 to 600 nm, with a

sharp peak at 440 nm. There was also a slight increase in

absorption in the near-IR. The broad 400−600 nm absorption

•

•−

was assigned to the formation of NPP (Figure S13 ) and the

slower with (DCB)

than with DCB . This suggests that

•−

peak at 440 nm to DCB . No features that could be assigned

pairing stabilizes the radical anion, allowing it to persist until

•

to [Ir(ppy) ] were observed, which suggests that oxidation of

the formation of NPP . It is noted that the rapid oxidation of

NPP is complete within 100 ns. While this precludes making a

deﬁnitive assignment for the rate constant of oxidation, k_o_x,

NPP by Ir(IV) means that back electron transfer to the

oxidized photocatalyst plays no role in the complete reaction

881

Journal of the American Chemical Society

J. Am. Chem. Soc. 2021, 143, 8878−8885

mJ/cm per pulse, gray). Orange line is the ﬁ t to a kinetic model.

Article

Figure 5. Single wavelength traces of 3.7 μM Ir(ppy) , 50 mM DCB, and 150 mM NPP at 540 nm, 800 nm, and 900 nm at 415 nm excitation (0.9

Figure 6. Overall kinetic scheme with rate constants for the coupling of N-phenylpyrrolidine (NPP) and 1,4-dicyanobenzene (DCB). Red arrows

indicate steps that are catalytically unproductive.

and demonstrates that electron transfer between radical ions is

the primary unproductive step that can limit the reaction.

Role of the Base in the Reaction. Deprotonation of

NPP^•is the key mechanistic step in the reaction. In the

Instead, DCB may function as the primary proton acceptor

in the reaction and then transfer the proton to sodium acetate

on a slower time scale. As noted above, without added base,

the reaction yield did not exceed 37%. If NPP was the primary

proton acceptor, this would correspond to roughly 14% of the

remaining NPP being protonated, whereas if DCB was the

primary proton acceptor, then 60% of the remaining DCB

would be protonated. This is consistent with the NMR data for

the reaction without sodium acetate. While the percent yield

was 37%, the conversion of DCB was 99% (Figures S19 and

S20), indicating that DCB was consumed via some other

pathway. This pathway may function as a base to deprotonate

NPP . Electrochemical experiments also provide further

indication of DCB acting as a base. When DCB was added

to a solution of NPP, the anodic peak potential for NPP

oxidation shifts to more positive potentials. This suggests that

the kinetic control of deprotonation occurs at higher scan

rates, outside of our window of measurement, and is consistent

with a faster rate of deprotonation.

original report by MacMillan and co-workers, it was

proposed that sodium acetate was the primary base; however,

sodium acetate may not be the initial proton acceptor from

•

NPP . For one, sodium acetate is insoluble in DMA, which

•

would make rapid deprotonation of NPP unlikely. Also, the

•

apparent deprotonation rate of NPP was measured electro-

chemically and no diﬀerence in the rate with or without

sodium acetate was observed (Figure S17). Lastly, Walker et al.

observed that removal of sodium acetate from the coupling of

•

DCB and piperidines had no eﬀect on the reaction. Inclusion

of sodium acetate is apparently necessary for the reaction

between NPP and DCB, as the maximum percent yield

without added base was 37 ± 1% after 48 h, compared to 88%

with sodium acetate (Figure S18). Also, NMR of the ﬁnal

reaction mixture shows that acetic acid is generated during the

reaction (Figure S5), indicating that sodium acetate is the

eventual proton acceptor.

Kinetic Modeling. By use of the values in Figure 6, a

kinetic model was developed for the reaction. Under reaction

882

J. Am. Chem. Soc. 2021, 143, 8878−8885

Journal of the American Chemical Society

was determined (Figure 7 ) before decaying over the course of

Article

conditions, an internal QY of nearly 1 at short times (<1 h)

to impact the QY of the reaction. These observations may

explain the QY results reported by Walker et al. for the

coupling of 2-methyl-1-phenylpiperidine with DCB (0.4−0.6).

These authors report that oxidation of the piperidine, k , is at

−1 1

least one magnitude slower (2.4 × 10 M s ) than what was

−1 −1

observed (>4 × 10 M s ) in the present study. This

diﬀerence in rate may be related to the presence of a methyl

group adjacent to the amine nitrogen. Kinetic modeling

−1 −1

suggests that k can be as slow as 10 M

without having a

signiﬁcant impact on the overall QY. Assuming the formation

of a similarly absorbing donor−acceptor complex between the

DCB and 2-methyl-1-phenylpiperidine, a QY of 0.4−0.6 would

correspond to an internal QY near 1.

CONCLUSION

■

Through a combination of steady state photochemical

measurements, laser spectroscopy, and electrochemical meth-

ods, a photoredox reaction that is broadly relevant to a wide

range of transformations was rigorously characterized. While

the overall product yield of the reaction is high, the external

QY of the reaction is more modest, only 0.15−0.3, depending

on the length of the reaction. Characterization of the rate

constants for productive and unproductive pathways permitted

the development of a kinetic model for this reaction. Unlike

previous mechanistic studies of photoredox reactions, this

model was used to predict reaction behavior and gave

predicted QY in good agreement with experimental results.

The kinetic modeling demonstrates that nearly all photons

Figure 7. Simulated external quantum and percent yields using the

values in Figure 6 and accounting for scattering losses and parasitic

absorption by NPP/DCB donor−acceptor complex.

4 h due to the generation of acetic acid and depletion of

DCB. This means that, in the absence of acetic acid, nearly all

absorbed photons result in product formation. Next, the QY

was calculated based on the number of incident photons (i.e.,

external QY). The parasitic absorbance was ﬁrst accounted for

by the DCB/NPP donor−acceptor complex. On the basis of

the concentrations of Ir(ppy) and DCB/NPP, it is estimated

that 44.2% of incoming photons were absorbed by the DCB/

NPP donor−acceptor complex and resulted in no productive

photochemistry. This resulted in a maximum external QY of

that are absorbed by Ir(ppy) result in product formation. An

.54, which is in excellent agreement with our observations

experimental QY of much less than 1 was observed because

using TBA acetate, which gave a QY of 0.52 at 30 min. Finally,

scattering by sodium acetate was accounted for, which

actinometry experiments estimated at ∼20% of incoming

photons. This results in a maximum external QY of 0.43, which

decays to 0.15 at 24 h. In general, this is in good agreement

with the experimental QY, except at times less than 4 h where

the model overestimates the QY (Figure 7).

scattering and parasitic photon absorption prevent ∼57% of

the incoming photons from reaching Ir(ppy) . While scattering

can be overcome by using a soluble base, the formation of the

NPP/DCB donor−acceptor complex provides a unique

challenge to the reaction. Decreasing the concentration of

NPP and/or DCB will reduce the absorbance of the donor−

acceptor complex and increase the fraction of photons

Taken together, the modeling demonstrates that the

limitations on this reaction are not related to the coupling

chemistry but instead to scattering losses and parasitic

absorption by the NPP/DCB donor−acceptor complex. The

latter is particularly problematic, as increasing the concen-

tration of photocatalyst will have little impact on the partition

of photons between the photocatalyst and donor−acceptor

complex. Likewise, decreasing the concentration of NPP and

DCB will lead to a decrease in the overall QY as well.

Modeling the concentration of NPP and DCB as 150 and 50

mM, respectively, led to a maximum internal QY of 89%. At

lower concentrations, unproductive pathways can better

compete with productive pathways and lead to a decrease in

QY.

absorbed by Ir(ppy) . At the same time, key pathways in the

reaction, such as reduction of [Ir(ppy) ] and deprotonation of

•

NPP , rely on high concentrations of NPP and DCB to

outcompete unproductive pathways. Instead, eﬀorts to

improve the QY of this reaction should focus on identifying

conditions that may disrupt the formation of this donor−

acceptor species (e.g., diﬀerent solvent, inert electrolyte).

Finally, these results also introduce questions about the role

of the cyanoarene in this coupling reaction. Not only was it

demonstrated that DCB^•pairs with a neutral DCB, but our

evidence also suggests that DCB is in fact the primary proton

−

•+

acceptor from NPP . However, our model fails to explain why

other cyanoarenes do not achieve high product yields. While

Interestingly, variation of the diﬀerent rate constants

revealed that there was a signiﬁcant amount of kinetic

redundancy in this reaction. For example, it is predicted that

the oxidation of NPP by Ir(IV) can be slowed by several orders

poor quenching kinetics may result in a low rate of product

formation, eventually the product yield of the reaction should

approach 100%. Experiments demonstrate that this is not the

case outside of DCB, suggesting a yet unknown pathway that

of magnitude without having an impact on the reaction (Figure

_S ₂ ₁ ₎_._L_i_k_e_w_i_s_e_,_t_h_e_r_a_t_e_o_f_d_e_p_r_o_t_o_n_a_t_i_o_n_o_f_N_P_P_•+ or

leads to an irreversible loss of the cyanoarene. Other

cyanoarenes may function as poorer bases than DCB or may

undergo decomposition pathways when protonated. Studies

are currently underway to understand how the mechanism

changes when cyanoarenes other than DCB are used.

quenching of Ir(ppy) * could both be slowed by more than

an order of magnitude without impacting the QY (Figures S22

and S23 ). Unproductive pathways (e.g., radical recombination)

would need to be signiﬁcantly faster than diﬀusion-controlled

883

J. Am. Chem. Soc. 2021, 143, 8878−8885

Journal of the American Chemical Society

Article

2) Ischay, M. A.; Anzovino, M. E.; Du, J.; Yoon, T. P. Efficient

ASSOCIATED CONTENT

sı Supporting Information

The Supporting Information is available free of charge at

https://pubs.acs.org/doi/10.1021/jacs.1c03693.

■

Visible Light Photocatalysis of [2 + 2] Enone Cycloadditions. J. Am.

Chem. Soc. 2008, 130, 12886−12887.

Photoredox Catalysis: Applications in Organic Synthesis. Chem. Soc.

Rev. 2011, 40, 102−113.

3) Narayanam, J. M. R.; Stephenson, C. R. J. Visible Light

Detailed procedures for ﬁtting TAS traces, determi-

•

4) Prier, C. K.; Rankic, D. A.; MacMillan, D. W. C. Visible Light

nation of the spectra of NPP and NPP , electro-

chemical measurement of the apparent deprotonation

rate, estimation of scattering losses, and kinetic

modeling; NMR data of reaction mixtures; QY as a

function of wavelength and light intensity; spectroelec-

trochemical diﬀerence spectra; NMR showing acetic

acid formation; TAS data showing impact of acetic on

Photoredox Catalysis with Transition Metal Complexes: Applications

in Organic Synthesis. Chem. Rev. 2013 , 113 , 5322 −5363.

(5) Romero, N. A.; Nicewicz, D. A. Organic Photoredox Catalysis.

Chem. Rev. 2016 , 116 , 10075 −10166.

6) Yoon, T. Photochemical Stereocontrol Using Tandem Photo-

redox-Chiral Lewis Acid Catalysis. Acc. Chem. Res. 2016, 49, 2307−

315.

(7) Matsui, J. K.; Lang, S. B.; Heitz, D. R.; Molander, G. A.

•

charge separation; spectra of NPP , NPP , and

Photoredox-Mediated Routes to Radicals: the Value of Catalytic

Radical Generation in Synthetic Methods Development. ACS Catal.

2017, 7, 2563−2575.

−

(

DCB)₂^•; quantum and percent yield without sodium

acetate; percent conversion of DCB without added

sodium acetae; simulated rate constant data from kinetic

modeling; electrochemical data for the rate of deproto-

nation (PDF )

(8) Arias-Rotondo, D. M.; McCusker, J. K. The Photophysics of

Photoredox Catalysis: A Roadmap for Catalyst Design. Chem. Soc.

Rev. 2016, 45, 5803−5820.

(9) Pitre, S. P.; McTiernan, C. D.; Scaiano, J. C. Understanding the

AUTHOR INFORMATION

Corresponding Author

John R. Swierk − Department of Chemistry, Binghamton

University, Binghamton, New York 13902, United States;

orcid.org/0000-0001-5811-7285; Email: jswierk@

binghamton.edu

Kinetics and Spectroscopy of Photoredox Catalysis and Transition-

■

Metal-Free Alternatives. Acc. Chem. Res. 2016 , 49 , 1320 −1330.

(10) Blanksby, S. J.; Ellison, G. B. Bond Dissociation Energies of

Organic Molecules. Acc. Chem. Res. 2003, 36, 255−263.

(11) Waidmann, C. R.; Miller, A. J. M.; Ng, C.-W. A.; Scheuermann,

M. L.; Porter, T. R.; Tronic, T. A.; Mayer, J. M. Using Combinations

of Oxidants and Bases as PCET Reactants: Thermochemical and

Practical Considerations. Energy Environ. Sci. 2012, 5, 7771−7780.

Authors

(12) Ruccolo, S.; Qin, Y.; Schnedermann, C.; Nocera, D. G. General

Bernard G. Stevenson − Department of Chemistry,

Binghamton University, Binghamton, New York 13902,

United States

Strategy for Improving the Quantum Efficiency of Photoredox

Hydroamidation Catalysis. J. Am. Chem. Soc. 2018, 140, 14926−

14937.

Ethan H. Spielvogel − Department of Chemistry, Binghamton

University, Binghamton, New York 13902, United States

Emily A. Loiaconi − Department of Chemistry, Binghamton

University, Binghamton, New York 13902, United States

Victor Mulwa Wambua − Department of Chemistry,

Binghamton University, Binghamton, New York 13902,

United States; orcid.org/0000-0003-1597-7394

Roman V. Nakhamiyayev − Department of Chemistry,

Binghamton University, Binghamton, New York 13902,

United States

13) Thompson, W. A.; Fernandez, E. S.; Maroto-Valer, M. M.

Review and Analysis of CO ₂ Photoreduction Kinetics. ACS

Sustainable Chem. Eng. 2020, 8, 4677 −4692.

(14) Ravelli, D.; Fagnoni, M. Dyes as Visible Light Photoredox

Organocatalysts. ChemCatChem 2012 , 4 , 169 −171.

15) Larsen, C. B.; Wenger, O. S. Photoredox Catalysis with Metal

Complexes Made from Earth-Abundant Elements. Chem. - Eur. J.

018, 24, 2039−2058.

16) Teegardin, K.; Day, J. I.; Chan, J.; Weaver, J. Advances in

Photocatalysis: A Microreview of Visible Light Mediated Ruthenium

and Iridium Catalyzed Organic Transformations. Org. Process Res. Dev.

2016, 20, 1156−1163.

Complete contact information is available at:

https://pubs.acs.org/10.1021/jacs.1c03693

(17) Yoon, T. P. Visible Light Photocatalysis: The Development of

Photocatalytic Radical Ion Cycloadditions. ACS Catal. 2013, 3, 895−

902.

Notes

18) Koyama, D.; Dale, D. J. A.; Orr-Ewing, A. J. Ultrafast

The authors declare no competing ﬁnancial interest.

Observation of a Photoredox Reaction Mechanism: Photoinitiation in

Organocatalyzed Atom-Transfer Radical Polymerization. J. Am. Chem.

Soc. 2018, 140, 1285−1293.

ACKNOWLEDGMENTS

This work was supported in part by an ACS Petroleum

Research Fund Doctoral New Investigator Award (Grant

■

(

19) Lewis-Borrell, L.; Sneha, M.; Bhattacherjee, A.; Clark, I. P.; Orr-

Ewing, A. J. Mapping the Multi-step Mechanism of a Photoredox

Catalyzed Atom-Transfer Radical Polymerization Reaction by Direct

Observation of the Reactive Intermediates. Chem. Sci. 2020, 11,

0013-DNI4). The authors also thank Binghamton University

for start-up funding. E.H.S. and B.G.S. thank the Department

of Chemistry for summer fellowships. We thank Profs. Jennifer

Hirschi and Matthew Vetticatt (Binghamton University) for

helpful advice on quantitative NMR measurements and

discussions on the reaction mechanism. We also thank Prof.

Malcolm Forbes (Bowling Green State University) for helpful

discussions on the manuscript.

20) Romero, N. A.; Nicewicz, D. A. Mechanistic Insight into the

475−4481.

Photoredox Catalysis of Anti-Markovnikov Alkene Hydrofunctional-

ization Reactions. J. Am. Chem. Soc. 2014, 136, 17024−17035.

Photocatalytic Reductive Dehalogenation Using Visible-Light: A

Time-Resolved Mechanistic Study. Eur. J. Org. Chem. 2017, 2017,

21) Martinez-Haya, R.; Miranda, M. A.; Marin, M. L. Metal-Free

164−2169.

REFERENCES

■

(22) Herzog, W.; Bronner, C.; Löffler, S.; He, B.; Kratzert, D.;

Stalke, D.; Hauser, A.; Wenger, O. S. Electron Transfer between

Hydrogen-Bonded Pyridylphenols and a Photoexcited Rhenium(I)

Complex. ChemPhysChem 2013, 14, 1168−1176.

1) Nicewicz, D. A.; MacMillan, D. W. C. Merging Photoredox

Catalysis with Organocatalysis: The Direct Asymmetric Alkylation of

Aldehydes. Science 2008, 322, 77−80.

884

J. Am. Chem. Soc. 2021, 143, 8878−8885

Journal of the American Chemical Society

(41) Ahneman, D. T.; Doyle, A. G. C-H Functionalization of Amines

Article

(

23) Rueda-Becerril, M.; Mahé, O.; Drouin, M.; Majewski, M. B.;

with Aryl Halides by Nickel-Photoredox Catalysis. Chem. Sci. 2016, 7,

7002−7006.

West, J. G.; Wolf, M. O.; Sammis, G. M.; Paquin, J.-F. Direct C −F

Bond Formation Using Photoredox Catalysis. J. Am. Chem. Soc. 2014,

24) Majek, M.; Jacobi von Wangelin, A. Mechanistic Perspectives

36, 2637−2641.

(42) Heitz, D. R.; Tellis, J. C.; Molander, G. A. Photochemical

Nickel-Catalyzed C −H Arylation: Synthetic Scope and Mechanistic

Investigations. J. Am. Chem. Soc. 2016, 138, 12715−12718.

on Organic Photoredox Catalysis for Aromatic Substitutions. Acc.

Chem. Res. 2016, 49, 2316−2327.

(

43) Ide, T.; Barham, J. P.; Fujita, M.; Kawato, Y.; Egami, H.;

Hamashima, Y. Regio- and Chemoselective Csp -H Arylation of

Benzylamines by Single Electron Transfer/Hydrogen Atom Transfer

Synergistic Catalysis. Chem. Sci. 2018, 9, 8453−8460.

25) Ng, Y. Y.; Tan, L. J.; Ng, S. M.; Chai, Y. T.; Ganguly, R.; Du, Y.;

Yeow, E. K. L.; Soo, H. S. Spectroscopic Characterization and

Mechanistic Studies on Visible Light Photoredox Carbon-Carbon

Bond Formation by Bis(Arylimino)Acenaphthene Copper Photo-

sensitizers. ACS Catal. 2018, 8 (12), 11277 −11286.

44) Revathi, L.; Ravindar, L.; Fang, W.-Y.; Rakesh, K. P.; Qin, H.-L.

Visible Light-Induced C-H Bond Functionalization: A Critical

Review. Adv. Synth. Catal. 2018, 360, 4652−4698.

26) Ma, J.; Zhang, X.; Phillips, D. L. Time-Resolved Spectroscopic

45) Ma, Y.; Yao, X.; Zhang, L.; Ni, P.; Cheng, R.; Ye, J. Direct

Observation and Characterization of Water-Assisted Photoredox

Reactions of Selected Aromatic Carbonyl Compounds. Acc. Chem.

Res. 2019, 52, 726−737.

Arylation of α -Amino C(sp )-H Bonds by Convergent Paired

Electrolysis. Angew. Chem., Int. Ed. 2019, 58, 16548−16552.

(46) Walker, M. M.; Koronkiewicz, B.; Chen, S.; Houk, K. N.;

(

27) Ting, S. I.; Garakyaraghi, S.; Taliaferro, C. M.; Shields, B. J.;

Mayer, J.; Ellman, J. A. Highly Diastereoselective Functionalization of

Piperidines by Photoredox-Catalyzed α -Amino C-H Arylation and

Epimerization. J. Am. Chem. Soc. 2020, 142, 8194 −8202.

Scholes, G. D.; Castellano, F. N.; Doyle, A. G. d-d Excited States of

Ni(II) Complexes Relevant to Photoredox Catalysis: Spectroscopic

Identification and Mechanistic Implications. J. Am. Chem. Soc. 2020,

(47) Rimshaw, A.; Grieco, C.; Asbury, J. B. Note: Using Fast

(

42, 5800−5810.

Digitizer Acquisition and Flexible Resolution to Enhance Noise

Cancellation for High Performance Nanosecond Transient Absorb-

ance Spectroscopy. Rev. Sci. Instrum. 2015, 86, 066107.

28) Zheng, S.; Zhang, S.-Q.; Saeednia, B.; Zhou, J.; Anna, J. M.;

Hong, X.; Molander, G. A. Diastereoselective Olefin Amidoacylation

via Photoredox PCET/Nickel-Dual Catalysis: Reaction Scope and

Mechanistic Insights. Chem. Sci. 2020, 11, 4131−4137.

(48) Dethe, D. H.; Srivastava, A.; Dherange, B. D.; Kumar, B. V.

Unsymmetrical Disulfide Synthesis through Photoredox Catalysis.

29) Coles, M. S.; Quach, G.; Beves, J. E.; Moore, E. G. A

Adv. Synth. Catal. 2018, 360, 3020−3025.

Photophysical Study of the Sensitization-Initiated Electron Transfer:

Insights into the Mechanism of Photoredox Activity. Angew. Chem.,

Int. Ed. 2020, 59 (24), 9522−9526.

(49) Xu, W.; Wang, W.; Liu, T.; Xie, J.; Zhu, C. Late-stage

trifluoromethylthiolation of benzylic C-H bonds. Nat. Commun. 2019,

0, 4867.

50) Lei, Z.; Banerjee, A.; Kusevska, E.; Rizzo, E.; Liu, P.; Ngai, M.-

30) Wang, K.; Lu, H.; Zhu, X.; Lin, Y.; Beard, M. C.; Yan, Y.; Chen,

X. Ultrafast Reaction Mechanisms in Perovskite Based Photocatalytic

C-C Coupling. ACS Energy Lett. 2020, 5, 566−571.

Y. β -Selective Aroylation of Activated Alkenes by Photoredox

Catalysis. Angew. Chem., Int. Ed. 2019, 58 (22), 7318−7323.

31) Vitaku, E.; Smith, D. T.; Njardarson, J. T. Analysis of the

(51) Yildiz, A.; Sertel, M.; Gambert, R.; Baumgärtel, H. On the

Structural Diversity, Substitution Patterns, and Frequency of Nitrogen

Heterocycles among U.S. FDA Approved Pharmaceuticals. J. Med.

Chem. 2014, 57, 10257−10274.

Stabilities of the Anion Radicals, Dianions, and Dimerdianions of

Benzenedicarbonitriles. Electrochim. Acta 1988, 33, 169 −170.

(52) van den Bosch, A.; Stienlet, D.; Ceulemans, J. Spectroscopic

Characterization of Alkane Radical Cations − I. Electron Absorption

Spectra of 3-Methylalkane Radical Cations. Radiat. Phys. Chem. 1989,

33, 371−379.

(53) Ichikawa, T.; Ota, N. ESR and Electronic Spectra of Alkane

Radical Cations Formed in γ -Irradiated 3-Methylpentane and 3-

Methylhexane Glasses Containing Alkane Solutes. J. Phys. Chem.

1987, 91, 3244−3248.

32) Chen, W.; Ma, L.; Paul, A.; Seidel, D. Direct α -C-H Bond

Functionalization of Unprotected Cyclic Amines. Nat. Chem. 2018,

0, 165−169.

33) Spangler, J. E.; Kobayashi, Y.; Verma, P.; Wang, D.-H.; Yu, J.-Q.

α -Arylation of Saturated Azacycles and N -Methylamines via

Palladium(II)-Catalyzed C(sp )-H Coupling. J. Am. Chem. Soc.

(

015, 137, 11876−11879.

34) Shaw, M. H.; Shurtleff, V. W.; Terrett, J. A.; Cuthbertson, J. D.;

MacMillan, D. W. C. Native Functionality in Triple Catalytic Cross-

Coupling: sp C-H Bonds as Latent Nucleophiles. Science 2016, 352,

35) McNally, A.; Prier, C. K.; MacMillan, D. W. C. Discovery of an

304−1308.

α -Amino C −H Arylation Reaction Using the Strategy of Accelerated

Serendipity. Science 2011, 334 (6059), 1114−1117.

(

36) Pirnot, M. T.; Rankic, D. A.; Martin, D. B. C.; MacMillan, D.

W. C. Photoredox Activation for the Direct β -Arylation of Ketones

and Aldehydes. Science 2013, 339, 1593 −1596.

37) Noble, A.; MacMillan, D. W. C. Photoredox α -Vinylation of α -

Amino Acids and N-Aryl Amines. J. Am. Chem. Soc. 2014, 136,

1602−11605.

38) Qvortrup, K.; Rankic, D. A.; MacMillan, D. W. C. A General

Strategy for Organocatalytic Activation of C −H Bonds via Photo-

redox Catalysis: Direct Arylation of Benzylic Ethers. J. Am. Chem. Soc.

(

014, 136, 626−629.

39) Zuo, Z.; Ahneman, D. T.; Chu, L.; Terrett, J. A.; Doyle, A. G.;

MacMillan, D. W. C. Merging Photoredox with Nickel Catalysis:

Coupling of α -carboxyl sp -carbons with Aryl Halides. Science 2014,

40) Jin, J.; MacMillan, D. W. C. Direct α -Arylation of Ethers

45, 437−440.

through the Combination of Photoredox-Mediated C-H Functional-

ization and Minisci Reaction. Angew. Chem., Int. Ed. 2015, 54, 1565−

569.

885