Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene

Article abstract of DOI:10.1016/j.chempr.2018.10.006

Divergent synthesis that enables a catalytic reaction to selectively produce different products from common substrates will allow the charting of wider chemical space and the unveiling of distinct mechanistic paradigms. A common strategy for it employs different ligands to modulate organometallic catalysts. Dramatic developments in photocatalysis have enabled previously inaccessible transformations. In particular, photoredox catalysis modulates the oxidation state of transition-metal complexes, offering enormous opportunities for methodology development. Herein, we developed a photo-mediated divergent ethylene difunctionalization via modulating oxidation states of the nickel catalyst by using different photoredox catalysts. This work will inspire new perspectives for value-added chemical synthesis using ethylene as a feedstock and shed light on photoredox-catalyst-based divergent synthesis, which fundamentally differs from ligand-controlled transition-metal catalysis.Divergent synthesis represents a powerful strategy for directly accessing different molecular scaffolds originating from the same starting materials. Access to different end products via transition-metal catalysis is conventionally achieved by ligand control. We herein demonstrate the use of ethylene feedstock and commercially available aryl halides to accomplish the divergent synthesis of 1,2-diarylethanes, 1,4-diarylbutanes, or 2,3-diarylbutanes in a highly selective fashion through the synergistic combination of nickel and photoredox catalysis. Mechanistic studies suggest that the observed selectivity was due to different active states of Ni(I) and Ni(0) modulated by Ru- and Ir-based photoredox catalysts, respectively. The ability to access different organometallic oxidation states via photoredox catalysis promises to inspire new perspectives for synergistic transition-metal-catalyzed divergent synthesis.Functionalization of ethylene without polymerization is challenging under photo-irradiation conditions. We have demonstrated that the photo-transformation of ethylene can be controllable by merging photoredox and transition-metal catalysis. In our study, the use of different photoredox catalysts was able to modulate the oxidation state of the nickel catalyst. Through different oxidation states, the nickel-catalyzed couplings proceeded via distinct pathways to generate divergent ethylene difunctionalization products selectively from the same feedstock.

Full text of DOI:10.1016/j.chempr.2018.10.006

Article
Photoredox-Catalysis-Modulated, Nickel-
Catalyzed Divergent Difunctionalization of
Ethylene
Jiesheng Li, Yixin Luo, Han Wen
Cheo, Yu Lan, Jie Wu
lanyu@cqu.edu.cn (Y.L.)
chmjie@nus.edu.sg (J.W.)
HIGHLIGHTS
Photo-mediated selective
divergent difunctionalization of
ethylene with aryl halide
Modulation of Ni catalyst
oxidation states via photoredox
catalysts
Mechanistic studies supported by
DFT calculations
Inexpensive and readily available
feedstocks
Functionalization of ethylene without polymerization is challenging under photo-
irradiation conditions. We have demonstrated that the photo-transformation of
ethylene can be controllable by merging photoredox and transition-metal
catalysis. In our study, the use of different photoredox catalysts was able to
modulate the oxidation state of the nickel catalyst. Through different oxidation
states, the nickel-catalyzed couplings proceeded via distinct pathways to generate
divergent ethylene difunctionalization products selectively from the same
feedstock.
Li et al., Chem 5, 1–12
January 10, 2019 ª 2018 Elsevier Inc.
https://doi.org/10.1016/j.chempr.2018.10.006

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
Article
Photoredox-Catalysis-Modulated,
Nickel-Catalyzed Divergent
Difunctionalization of Ethylene
1,3,
Jiesheng Li,1 Yixin Luo,2 Han Wen Cheo,1 Yu Lan,2, and Jie Wu
*
*
SUMMARY
The Bigger Picture
Divergent synthesis that enables a
catalytic reaction to selectively
produce different products from
common substrates will allow the
charting of wider chemical space
and the unveiling of distinct
mechanistic paradigms. A
Divergent synthesis represents a powerful strategy for directly accessing
different molecular scaffolds originating from the same starting materials.
Access to different end products via transition-metal catalysis is conventionally
achieved by ligand control. We herein demonstrate the use of ethylene
feedstock and commercially available aryl halides to accomplish the divergent
synthesis of 1,2-diarylethanes, 1,4-diarylbutanes, or 2,3-diarylbutanes in a high-
ly selective fashion through the synergistic combination of nickel and photore-
dox catalysis. Mechanistic studies suggest that the observed selectivity was due
to different active states of Ni(I) and Ni(0) modulated by Ru- and Ir-based
photoredox catalysts, respectively. The ability to access different organome-
tallic oxidation states via photoredox catalysis promises to inspire new perspec-
tives for synergistic transition-metal-catalyzed divergent synthesis.
common strategy for it employs
different ligands to modulate
organometallic catalysts.
Dramatic developments in
photocatalysis have enabled
previously inaccessible
transformations. In particular,
photoredox catalysis modulates
the oxidation state of transition-
metal complexes, offering
INTRODUCTION
Development of catalytic methodology to selectively produce targeted products is
of essential focus in organic synthesis. In this regard, divergent synthesis serves as an
appealing yet challenging strategy to selectively access multiple scaffolds starting
from the same materials.1,2 Subtle variations on catalysts, additives, solvents, or
temperature can account for distinct reaction pathways. In particular, transition-
metal catalysts have conventionally relied on the proper choice of ligands bearing
different electronic or steric properties to achieve divergent reactivity from the
same starting materials (Figure 1A).3,4
enormous opportunities for
methodology development.
Herein, we developed a photo-
mediated divergent ethylene
difunctionalization via modulating
oxidation states of the nickel
catalyst by using different
As the simplest alkene, ethylene is widely utilized in the chemical industry, with an
estimated annual global production over 150 million tons, far exceeding that of
any other organic compound.5 The high-volume usage of ethylene in industry in-
cludes the production of polyethylene materials and other commodities and feed-
stocks, such as ethylene oxide, propionic acid, acetaldehyde, and ethylbenzene.6
However, relatively few methods used ethylene as a reactant for the synthesis of
fine chemicals, probably due to its inherent simplicity resulting in molecules of
modest complexity and the apprehensions of handling such flammable gas. Existing
catalytic transformations of ethylene into fine chemicals are typically limited to
monofunctionalization, including Heck or reductive-Heck-type reactions, hydroacy-
lations, hydrovinylations, Wacker oxidations, enyne metathesis, and ethenolysis
(Figure 1B).7 To achieve effective ethylene difunctionalization, there is a need to
avoid potential competing side reactions, such as polymerization and oligomeriza-
tion, in the presence of excess ethylene8 or b-hydride elimination of in-situ-formed
alkyl metallic intermediates to give unwanted Heck-type products. In this regard,
photoredox catalysts. This work
will inspire new perspectives for
value-added chemical synthesis
using ethylene as a feedstock and
shed light on photoredox-
catalyst-based divergent
synthesis, which fundamentally
differs from ligand-controlled
transition-metal catalysis.
Chem 5, 1–12, January 10, 2019 ª 2018 Elsevier Inc.
1

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
Sigman and coworkers have pioneered the palladium-catalyzed ethylene difunction-
alization providing 1,1-arylvinylation9 and 1,1-diarylation products.10,11 In 2015, the
Ogoshi group disclosed a 1,2-carbodifunctionalization of ethylene via a selective
cross-trimerization of tetrafluoroethylene, ethylene, and aldehydes.12 Recently,
Bower and Russell reported an oxidative ethylene 1,2-difunctionalization via
gold-catalyzed oxyarylation.13 Despite these elegant developments, effective
catalytic strategies for difunctionalization of ethylene remain scarce and are highly
desirable.
Visible-light photocatalysis has emerged as a powerful technique for organic
synthesis over the past decade.14,15 In particular, the synergistic combination of
photoredox catalysis and transition-metal catalysis provides distinct modes for
activation of the organometallic complexes through single-electron transfer (SET)
or energy transfer.16 However, to the best of our knowledge, modulating oxidation
state of organometallic catalysts by photoredox catalysts to access different catalytic
pathways leading to divergent products has not been previously reported. This
strategy fundamentally differs from ligand-controlled transition-metal catalysis.17
Additionally, literature reports on photo-mediated transformations involving
gaseous reagents are limited.18–22 Our group has recently developed a ‘‘stop-
flow’’ micro-tubing (SFMT) reactor for efficient screening of gas-involved photo-
mediated transformations in a convenient and safe manner.23 As part of our ongoing
interests in developing visible-light-promoted transformations using inexpensive
gaseous feedstocks,23–25 we herein report the light-mediated ethylene difunctional-
ization through the synergistic combination of photoredox and Ni catalysis. Assisted
by photocatalysts with different redox potentials, Ni-catalyzed reductive coupling
between aryl halides and ethylene produced 1,2-diarylethanes, 1,4-diarylbutanes,
and 2,3-diarylbutanes in a highly selective manner (Figure 1C).
RESULTS AND DISCUSSION
Condition Exploration for Divergent Difunctionalization of Ethylene
Earth-abundant Ni catalysts were selected for the investigation because of their
intriguing properties: (1) readily accessible multiple oxidation states, (2) favorable
binding with olefins, (3) difficult-to-trigger b-hydride elimination,26 and (4) privilege
in merging with photoredox catalysis.16,27,28 We commenced our study by evalu-
ating visible-light-promoted difunctionalization of ethylene with 4-iodotoluene 1a
as the model substrate. Through the synergistic merger of a photocatalyst and a
Ni catalyst, three different types of products, including one ethylene insertion prod-
uct 2a, two ethylene insertion linear product 3a, and two ethylene insertion branched
product 4a, were generated. After extensive evaluation of reaction parameters
(Figure 2 and Tables S1–S16), all three types of products could be achieved in
high selectivity from identical starting materials. As shown in Figure 2, 1,2-diaryl-
ethane 2a was obtained in 90% yield with high selectivity (2a:3a = 13:1) in the
presence of Ru(bpy)3Cl2 (0.5 mol %), NiCl2$glyme (10 mol %), N,N,N0,N0-tetramethy-
lethylenediamine (TMEDA) (2 equiv), and Hunig’s base (2 equiv) in DMSO (0.2 M)
¨
1Department of Chemistry, National University of
Singapore, 3 Science Drive 3, Singapore 117543,
Republic of Singapore
under blue light-emitting diode (LED) irradiation at 50ꢀC for 24 hr (entry 1, condition
A). On the other hand, 3a could be prepared in excellent selectivity and yield
(90%, 2a:3a = 1:30) using [Ir(ppy)2(dtbbpy)]PF6 (1 mol%) as the photocatalyst in
MeCN (0.05 M) at 25ꢀC under blue LED irradiation (entry 6, condition B). Increasing
the reaction temperature to 100ꢀC and replacing NiCl2$glyme with NiI2 in DMSO
(0.1 M) delivered 4a exclusively in high yield (entry 8, condition C). Notably, no diaryl
coupling of 1a or ethylene oligomerization was observed under these visible-light-
mediated conditions. The holistic choice of photocatalysts, Ni catalysts, reaction
2School of Chemistry and Chemical Engineering,
Chongqing University, Chongqing 400044, P.R.
China
3Lead Contact
*Correspondence: lanyu@cqu.edu.cn (Y.L.),
chmjie@nus.edu.sg (J.W.)
https://doi.org/10.1016/j.chempr.2018.10.006
2
Chem 5, 1–12, January 10, 2019

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
Figure 1. Divergent Functionalization of Ethylene
(A) Strategies for transition-metal-catalyzed divergent synthesis.
(B) Reported approaches and challenges for ethylene functionalization.
(C) Divergent and selective difunctionalization of ethylene by regulating Ni catalytic behavior with
photoredox catalysts.
temperature, concentration, and solvents was critical and exhibited a profound
effect on the product selectivity (Tables S1–S16).29
Effect of Photoredox Catalysts
As shown in Figure 2, both Ru(bpy)3Cl2 and Ru(phen)3Cl2 with a similar oxidative po-
tential (E1/2Ru(I)/Ru(II) = ꢁ1.26 V versus SCE in DMSO) could selectively produce 2a in
good selectivity and yield under condition A (entries 1 and 3). However, when the
photocatalyst was changed to Ir(dFCF3ppy)2(dtbbpy)PF6 or [Ir(ppy)2(dtbbpy)]PF6,
the selectivity decreased with more 1,4-diarylbutane 3a generated (entries 4
and 5). In addition, increasing the Ru(bpy)3Cl2 catalyst loading from 0.5 mol% to
1 mol% dramatically eroded the selectivity of 2a (entry 2). Conversely, the use of
[Ir(ppy)2(dtbbpy)]PF6 under condition B delivered product 3a in excellent selectivity,
whereas Ru(bpy)3Cl2 gave low selectivity between 2a and 3a (entries 6 and 7).
Chem 5, 1–12, January 10, 2019
3

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
Figure 2. Condition Evaluation for Visible-Light-Driven Difunctionalization of Ethylene
Ratios of 2a:3a:4a were determined by gas chromatography analysis of crude reaction mixtures. The yields listed are the combined isolated yields, and
the diastereomeric ratio (d.r.) of 4a = 1:1. The reaction for entry 22 was conducted in the stop-flow microtubing reactor at 250 psi pressure. TMEDA,
N,N,N0,N0-tetramethylethylenediamine; DIPEA, N,N-diisopropylethylamine.
Notably, both Ir and Ru photocatalysts could explicitly deliver 4a at 100ꢀC under
condition C, albeit in lower yield using the Ru system (entries 8 and 9). Even though
the correlation of the photocatalyst and the product selectivity between 2a and 3a is
still not fully understood at the current stage, and requires further investigation, we
did observe that the Ru catalyst favored the formation of product 2a while the Ir cata-
lyst had a strong tendency toward 3a.
Effect of Ni Catalysts
A Ni concentration dependence was observed in both selective formations of 2a and
3a. A higher catalyst loading of NiCl2$glyme resulted in an improved selectivity of 2a
under condition A (entries 10–12) and 3a under condition B (entries 14 and 15), which
suggested that both catalytic processes might involve a bi-Ni-species interaction.
Reaction using Ni(COD)2 in place of NiCl2$glyme gave no desired product 2a under
condition A (entry 13), indicating that Ni(0) might not be operative in the active
catalytic cycle. In contrast, product 3a was generated in high yield using Ni(COD)2
under condition B (entry 16). Using NiCl2$glyme at 100ꢀC under condition C led to
4
Chem 5, 1–12, January 10, 2019

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
a very high selectivity of 4a, with 4-chlorotoluene as the sole side product (entry 17).
Replacing NiCl2$glyme with NiI2 avoided halogen exchange of 1a (entry 18).
Ni(COD)2 was also applicable to selectively prepare 4a (entry 19).
Effect of Solvent
The formation of 2a could only be achieved in DMSO or MeCN under condition A
(Table S14), and MeCN resulted in much lower selectivity than DMSO (Figure 2, entry
20). Product 3a could be generated under condition B with other polar solvents such
as DMSO, dimethylformamide (DMF), and N-methyl-2-pyrrolidone (NMP) (Table
S15). However, these solvents were incomparable with MeCN for the selective
preparation of 3a, resulting in lower yield or selectivity. The different selectivity
may be attributed to the change of redox potentials of photocatalysts when in
different solvents. The formation of 4a under condition C was more tolerant toward
different solvents, as polar solvents such as DMSO, MeCN, DMF, dimethylaceta-
mide, and NMP could all produce 4a exclusively (Table S16).
Effect of Temperature and Concentration
Temperature of 50ꢀC and 0.2 M concentration were the optimum criteria for the
selective generation of 2a under condition A. Decreasing the reaction tempera-
ture to 25ꢀC or lowering concentration to 0.05 M resulted in lower product
selectivity (Figure 2, entries 23 and 26), whereas increasing reaction temperature
to 80ꢀC afforded 2a and a significant amount of 2,3-diarylbutane 4a (entry 24).
The formation of 3a under condition B proceeded selectively at 25ꢀC and
0.05
M concentration. Increasing reaction temperature and concentration
resulted in higher yield of 2a and 4a (entries 25 and 27). Further increasing the
reaction temperature to 100ꢀC delivered 4a exclusively in high yield. The selec-
tivity between 2a and 3a may be attributed to the varying amount of ethylene
in the reaction solution at different temperature and concentration. 4a was selec-
tively generated above 100ꢀC.
Effect of Base and Others
Hunig’s base could be replaced by Et N, whereas other bases investigated were not
¨
3
effective (Tables S8–S10). Omission or decreased loading of Hunig’s base led to
¨
lower productivity for compounds 2a, 3a, and 4a (Tables S11–S13). Finally, no prod-
uct was detected in the absence of either photocatalyst, Ni catalyst, TMEDA, or light,
demonstrating the need for all these components (entry 28).29 Excess TMEDA was
required, which served not only as a critical ligand for Ni but also as a ‘‘sink’’ for
the halide leaving groups. The formation of TMEDAH-I salts between protonated
TMEDA and the halide were detected and characterized by X-ray crystallographic
analysis (see Data and Software Availability).
Reaction Scope Investigation
As outlined in Scheme 1A, para- and meta-substituted aryl iodides and bromides
with either electron-rich or electron-poor functionality could effectively react with
ethylene in the presence of Ru/Ni or Ir/Ni dual catalysts under blue LED irradiation
to furnish either one ethylene or two ethylene linear insertion products (2 or 3) in
good yield and selectivity. However, both transformations were drastically affected
by steric effects, as ortho-substituted aryl halides reacted sluggishly.
The initial substrate expansion of 2,3-diarylbutane product 4 using condition C was
only applicable toward para-functionalized bromo- and iodoarenes, and other arene
and heteroarene variants afforded either no reactivity or dehalogenation byprod-
ucts. Our speculation was that most ethylene escaped out from DMSO solution at
Chem 5, 1–12, January 10, 2019
5

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
Scheme 1. Scope of Visible-Light-Driven Divergent Ethylene Difunctionalization
All reactions were conducted under optimized conditions; see the Supplemental Information for details of conditions A, B, D, and E. Reported yields
were isolated yields. Ratios of 2:3 were determined by 1H NMR of isolated product mixtures, except where otherwise noted. Ratios of 2a:3a were
determined by gas chromatography analysis of crude reaction mixtures. The d.r. for branched product 4 = 1:1 (meso:dl) as determined by 1H NMR
spectrum analysis.
100ꢀC in batch reactors. The SFMT reactor was thus applied to keep ethylene in the
reaction solution at high temperature and high pressure for the substrate scope eval-
uation of 4 (condition D).23 The SFMT reactor proved to be efficient for the
generation of 2,3-diarylbutanes (4a–4m) from a wide range of para- or meta-
substituted aryl iodides and bromides (Scheme 1B). Further substrate expansion
to arene-fused heterocycles was investigated, and 4n–4t were generated effectively.
The structure of 4t was confirmed by single-crystal structural analysis, and other
products 4 were assigned by analogy. Finally, reductive-Heck products 5a–5f were
generated when arenes bearing electron-deficient substituents were subjected
6
Chem 5, 1–12, January 10, 2019

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
to the Ir/Ni dual catalytic system under blue LED irradiation at 80ꢀC in the presence
of water (condition E).
Mechanistic Elucidation with Supporting Evidence
Reaction discovery studies suggested that the utilization of [Ir(ppy)2(dtbbpy)]PF6, a
dilute reaction mixture, and reaction at 25ꢀC would favor the generation of two
ethylene insertion product 3, whereas the use of Ru(bpy)3Cl2, a concentrated reac-
tion mixture, and slight heating at 50ꢀC would prefer the formation of one ethylene
insertion product 2. In addition, the use of [Ir(ppy)2(dtbbpy)]PF6 at high temperature
(100ꢀC), would selectively generate two ethylene insertion branched product 4. A
critical question raised by these studies relates to tuning reaction conditions result-
ing in divergent products initiated by the same aryl halide and ethylene.
Selective Generation of 2 (Catalytic Cycle A)
Control experiments and cyclic voltammetry (CV) analysis suggested Ni(I) as the
active catalyst for the generation of 1,2-diarylethane 2 in the Ru/Ni dual catalyst
system.30,31 CV analysis of NiCl2-TMEDA complex (EP
Ni(I)/Ni(0)
= ꢁ0.68 V and
Ni(II)/Ni(I)
Ru(I)/Ru(II)
EP
= ꢁ1.66 V versus SCE in DMSO) and Ru(bpy)3Cl2 (E1/2
=
ꢁ1.26 V versus SCE in DMSO), where Ru(I) can partially reduce Ni(II) to Ni(I) with
roughly 0.58 V of thermodynamic driving force, while full reduction of Ni(II) to
Ni(0) is unfavorable with roughly ꢁ0.40 V (Scheme 2B and Figures S22–S25). These
data are in accordance with the reducing ability of Ru(bpy)3Cl2, which is inefficient in
conjunction with an amine sacrificial reductant for Ni-catalyzed reductive coupling
reactions.32,33 The reaction was totally suppressed in the presence of stoichiometric
TEMPO (Scheme 2D [A]). Furthermore, no ethylene insertion product was observed
when equal molarities of 1a and 2-iodoethylbenzene were added (Scheme 2D [B]).
This could indicate a better reactivity of alkyl iodide than of aryl iodide toward the
in situ generated Ni(I) species and that ethylene insertion into alkyl Ni intermediate
may be difficult. Intriguingly, reaction in MeCN resulted in a significant amount of 3a
(2a:3a = 2:1; Figure 2, entry 20). This was probably due to the similar reduction
Ru(I)/Ru(II)
potential of Ru(bpy)3Cl2 in MeCN (E1/2
= ꢁ1.33 V versus SCE) versus
Ni(II) to Ni(0) (EP = ꢁ1.27 V versus SCE in MeCN) as illustrated in Scheme 2C, where
Ru(bpy)3Cl2 was capable of fully reducing Ni(II) to Ni(0), thereby altering the
selectivity.34 Lastly, light on/off experiments indicated that light was essential for
this transformation (Figure S9).
Selective Generation of 3 (Catalytic Cycle B)
CV analysis indicated that [Ir(ppy)2(dtbbpy)]PF6 (E1/2Ir(II)/Ir(III) = ꢁ1.51 V versus SCE in
MeCN) gave roughly 0.24 V of thermodynamic driving force to fully reduce Ni(II)
complex to Ni(0) (EP = ꢁ1.27 V versus SCE in MeCN; Scheme 2C), which indicated
that Ni(0) was involved in the formation of 3a. The reaction proceeded to afford
3a in the presence of 1.1 equiv TEMPO (Scheme 2D [C]), suggesting a non-radical
pathway. Starting with 2-iodoethylbenzene instead of 4-iodotoluene 1a exclusively
generated homo-dimerization product 3b (Scheme 2D [D]). The absence of ethylene
insertion again suggested that the transient alkyl Ni intermediate might not undergo
migratory insertion with ethylene, thus preventing ethylene oligomerization and
polymerization.
Selective Generation of 4 (Catalytic Cycle C)
Prevailing mechanistic studies indicates that Ni(0) will be generated with Ir-photoca-
talyst. Radical may not be involved in the catalytic cycle as desired product 4a was
generated in the presence of 1.1 equiv TEMPO (Scheme 2D [E]). 4b could be
obtained when 2-iodoethylbenzene instead of 4-iodotoluene 1a was used in the
Chem 5, 1–12, January 10, 2019
7

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
Scheme 2. Proposed Mechanisms with Supporting Evidence
(A) Proposed plausible mechanisms. Photo-reductive quenching cycle was omitted for clarity; see Figures S18–S20 for detail mechanisms of each cycle.
(B) Cyclic voltammetry analysis of Ni catalyst and Ru photocatalyst in DMSO.
(C) Cyclic voltammetry analysis of Ni catalyst, Ru and Ir photocatalysts in MeCN.
(D) Control experiments to elucidate the mechanisms.
presence of Ni(COD)2 catalyst at 30ꢀC (Scheme 2D [F] and Figure S11), which
supported that the isomerization of alkyl Ni intermediate through b-hydride elimina-
tion to generate nickel hydride was possible even at room temperature to deliver
branched product 4. Surprisingly, increasing temperature to 100ꢀC exclusively deliv-
ered tetramer 6, which indicated that the in situ generated nickel hydride would
preferentially undergo reductive elimination in the presence of amine base instead
of re-inserting back onto styrene (see Figure S13 for proposed mechanism). Further
investigations are underway to explore this unprecedented transformation. In addi-
tion, the nickel hydride species was capable of intercepting another olefin present in
the reaction mixture (Scheme 2D [G] and Figure S12).35 Finally, organo-photocata-
lyst 1,2,3,5-tetrakis(carbazol-9-yl)-4,6-dicyanobenzene (4CzIPN)36 could be
applied to replace the Ir catalyst to enable a more economical transformation
8
Chem 5, 1–12, January 10, 2019

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
Figure 3. Free-Energy Profiles for the Ni(I)-Ni(III) Catalytic Cycle of the Diarylethane Product 2a Generation Process
Values for bond lengths are given in angstroms.
(Scheme 2D [H]). Deuteration experiments suggested that the mechanism for the
generation of product 5 was built on catalytic cycle C, where alkyl Ni was quenched
by D2O instead of DMSO after isomerization (Figure S14).
To further reveal the Ni(I)-based mechanistic pathway for the selective generation of
diarylethane product 2, density functional theory (DFT) method M11-L was applied.
As shown in Figure 3, the active catalyst Ni(I) species CP1 was chosen as the relative
zero point for the free energy profiles, which could be generated via SET with
+
reduced photocatalyst Ru(bpy)3 (a detailed mechanism is shown in Figure S18).
The generation of an aryl radical via Ni(I)-induced SET oxidative addition was ruled
out as the relative free energy was too high (50.2 kcal/mol; Figure S16). 4-Iodoto-
luene 1a was thus activated by coordinating with CP1 to give the intermediate
CP2 with a free energy increase of 2.7 kcal/mol. The subsequent oxidative addition
occurred via a three-membered ring transition state TS1 with a 7.4 kcal/mol overall
barrier, where an aryl-Ni(III) complex CP3 was generated. The geometry of TS1 is
˚
shown in Figure 3, in which the Ni-I distance (3.11 A) was shorter than the sum of their
van der Waals radii. The calculated spin density of this transition state illustrated that
the unpaired electrons were mainly localized on the Ni atom. Further intrinsic
reaction coordinate calculations demonstrated that iodide coordinated with Ni(III)
to form intermediate CP3 following TS1. Then the ligand exchange with ethylene
would form intermediate CP5 with 15.1 kcal/mol endergonic with the release of
iodide. The irreversible migratory insertion of coordinated ethylene into aryl-Ni
bond occurred via transition state TS2 with an overall barrier of 20.8 kcal/mol to
form an alkyl Ni(III) intermediate CP6. The theoretical calculation revealed that the
ethylene insertion step was considered to be the rate-determining step in catalytic
cycle A. Further coordination of CP6 followed by insertion of another ethylene was
unfavorable with a 26.6 kcal/mol endergonic barrier (Figure S17). CP6 could react
with intermediate CP3 to form a dichloro-bridged binuclear Ni(III) intermediate
Chem 5, 1–12, January 10, 2019
9

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
CP7. Then rapid intramolecular transmetallation of the aryl group could take place
via TS3. The calculated barrier of this step was lower than zero, which could be
attributed to the computational error of the solvation effect and the usage of entropy
correction in the gas phase. The following reductive elimination could take place
from a mononuclear Ni(III) intermediate CP8 with a free energy barrier of only
10.3 kcal/mol through TS4. After releasing product 2a, the active catalyst CP1 was
regenerated. Moreover, the possibility for the formation of 1,4-diarylbutane 3a
through the Ni(I)–Ni(III) catalytic cycle was considered theoretically (Figure S15).
However, the overall activation free energy for the formation of 3a was 34.8 kcal/mol
via transition state TS6, and the relative free energy was 19.8 kcal/mol higher than
via TS4.
Tentative Proposed Mechanistic Cycle
In light of all the experimental data and DFT calculation, plausible mechanistic path-
ways for the divergent ethylene functionalization are proposed in Scheme 2A.
Different from previously reported Ni(0)-catalyzed reductive coupling reactions,31,37
the synthesis of 1,2-diarylethane 2 (catalytic cycle A) is initiated by Ru(bpy)3Cl2-pro-
moted Ni(I) formation. The generated Ni(I) intermediate I subsequently undergoes
oxidative addition with aryl halides to form aryl-Ni(III) species II. Migratory insertion
of ethylene into aryl-Ni(III) with the dissociation of a halide ligand (trapped by
TMEDAH+) generates a cationic alkyl Ni(III) species III. The aryl transfer by a transme-
tallation between III and aryl-Ni(III) species II delivers an aryl alkyl Ni(III) species IV,
which can undergo a rapid reductive elimination to yield diarylethane product 2
and regenerate the active Ni(I) catalyst I. The transmetallation process would also
generate a cationic Ni(III) species V, which could be reduced by a Ru-based
photoredox cycle to regenerate Ni(I) I. As for Ir-catalyzed generation of 1,4-diarylbu-
tanes 3 (catalytic cycle B), Ni(II) precursor is fully reduced to Ni(0) through the
Ir-based photoredox cycle. Oxidative insertion of aryl halides to Ni(0) followed by
ethylene migratory insertion furnishes alkyl Ni(II) complex VII. As suggested by
control experiments, additional ethylene insertion would not occur with alkyl Ni spe-
cies VII, thereby preventing unwanted oligomerization and polymerization.
Transmetallation between two molecules of complex VII delivers the dialkyl Ni(II) in-
termediate VIII and a Ni(II) halide, the latter will be reduced by photocatalyst and
amine reductant to regenerate Ni(0). Reductive elimination of VIII achieves 3 and
Ni(0). In the proposed catalytic cycle C for the selective generation of 2,3-diarylbu-
tanes 4, Ni(II) complex VII is generated via Ir-photocatalyst through a similar process
as catalytic cycle B. b-Hydride elimination is triggered at high temperature to give a
Ni hydride species IX, which then undergoes a reinsertion to form the more thermo-
dynamically favored benzyl Ni intermediate X.38 Transmetallation between two
molecules of X followed by reductive elimination accomplishes the synthesis of 4
and regenerates the Ni(0) catalyst. Alternatively, in the presence of water, interme-
diate X could be quenched to deliver reductive Heck-type product 5.
Conclusion
In summary, we have discovered a divergent synthesis of 1,2-diarylethanes, 1,4-di-
arylbutanes, 2,3-diarylbutanes, and ethylarenes from a common aryl halide and
ethylene through the synergistic combination of photoredox and Ni catalysis. All
products can be generated in a highly selective fashion by proper selection of photo-
catalysts and tuning the reaction parameters. Various control experiments and DFT
calculations were conducted to explore the mechanisms for the selective product
generation. Our study showcases the difunctionalization of ethylene under photore-
dox conditions, which will inspire new perspectives for value-added chemical
synthesis using ethylene as a feedstock. Furthermore, distinct catalytic pathways
10
Chem 5, 1–12, January 10, 2019

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
can be obtained via photoredox modulation of organometallic intermediates to ac-
cess desired oxidation states, which will shed new light on transition-metal-catalyzed
divergent synthesis.
EXPERIMENTAL PROCEDURES
Full experimental procedures are provided in the Supplemental Information.
DATA AND SOFTWARE AVAILABILITY
The structures of TMEDAH-I, meso-4t, and 6 reported in this article have been
deposited in the Cambridge Crystallographic Data Centre under accession numbers
CCDC: 1825640, 1825641, and 1825642, respectively.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures, 123
figures, 38 tables, and 3 data files and can be found with this article online at
https://doi.org/10.1016/j.chempr.2018.10.006.
ACKNOWLEDGMENTS
We are grateful for the financial support provided by the National University of
Singapore and the Ministry of Education of Singapore (R-143-000-665-114 and
R-143-000-696-114), GlaxoSmithKline and the Singapore Economic Development
Board (R-143-000-687-592), the A*STAR RIE2020 Plan for Advanced Manufacturing
and Engineering (A1783c0013), and the National Natural Science Foundation of
China (21822303 and 21772020). We thank Prof. Zhu Shaolin (Nanjing University)
and Dr. Yang Yang (UC-Berkeley) for their helpful discussions.
AUTHOR CONTRIBUTIONS
J.L. and H.W.C. conducted the experiments and analyzed the data. Y. Luo and Y. Lan
conducted the DFT calculation and mechanism study. J.W. and J. Li designed the
project, analyzed the data, and wrote the manuscript.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: May 30, 2018
Revised: September 3, 2018
Accepted: October 11, 2018
Published: November 8, 2018
REFERENCES AND NOTES
1. Ping, L., Chung, D.S., Bouffard, J., and Lee,
S.G. (2017). Transition metal-catalyzed site-
and regio-divergent CꢁH bond
functionalization. Chem. Soc. Rev. 46, 4299–
4328.
4. Liu, J., Liu, Q., Franke, R., Jackstell, R., and
Beller, M. (2015). Ligand-controlled palladium-
catalyzed alkoxycarbonylation of allenes:
regioselective synthesis of a,b- and b,g-
unsaturated esters. J. Am. Chem. Soc. 137,
8556–8563.
7. Saini, V., Stokes, B.J., and Sigman, M.S. (2013).
Transition-metal-catalyzed laboratory-scale
carbon–carbon bond-forming reactions of
ethylene. Angew. Chem. Int. Ed. 52, 11206–
11220.
2. Lee, Y.C., Kumar, K., and Waldmann, H. (2018).
Ligand-directed divergent synthesis of carbo-
and heterocyclic ring systems. Angew. Chem.
Int. Ed. 57, 5212–5226.
8. For UV-light-promoted cycloaddition and free-
radical polymerization using ethylene, see
Mirbach, M.F., Mirbach, M.J., and Saus, A.
(1982). High-pressure photochemistry and
Ultraviolet spectroscopy in gas-liquid systems.
Chem. Rev. 82, 59–76.
5. McCoy, M., Reisch, M., Tullo, A.H., Short, P.L.,
Tremblay, J.F., and Storck, W.J. (2006).
Production: growth is the norm. Chem. Eng.
News 84, 59–236.
3. Meng, Q.Y., Wang, S., Huff, G.S., and Konig, B.
¨
(2018). Ligand-controlled regioselective
hydrocarboxylation of styrenes with CO2 by
combining visible light and nickel catalysis.
J. Am. Chem. Soc. 140, 3198–3201.
6. Independent Chemical Information Service
(2007). Ethylene uses and market data. https://
www.icis.com/resources/news/2007/11/05/
9075777/ethylene-uses-and-market-data.
9. Saini, V., and Sigman, M.S. (2012). Palladium-
catalyzed 1,1-difunctionalization of ethylene.
J. Am. Chem. Soc. 134, 11372–11375.
Chem 5, 1–12, January 10, 2019
11

Please cite this article in press as: Li et al., Photoredox-Catalysis-Modulated, Nickel-Catalyzed Divergent Difunctionalization of Ethylene, Chem
(2018), https://doi.org/10.1016/j.chempr.2018.10.006
10. Saini, V., Liao, L., Wang, Q., Jana, R., and
Sigman, M.S. (2013). Pd(0)-catalyzed 1,1-
diarylation of ethylene and allylic carbonates.
Org. Lett. 15, 5008–5011.
21. Shimomaki, K., Murata, K., Martin, R., and
Iwasawa, N. (2017). Visible-light-driven
carboxylation of aryl halides by the combined
use of palladium and photoredox catalysts.
J. Am. Chem. Soc. 139, 9467–9470.
thiols with aryl and heteroaryl iodides via thiyl
radicals. J. Am. Chem. Soc. 138, 1760–1763.
31. Oderinde, M.S., Jones, N.H., Juneau, A.,
Frenette, M., Aquila, B., Tentarelli, S., Robbins,
D.W., and Johannes, J.W. (2016). Highly
chemoselective iridium photoredox and nickel
catalysis for the cross-coupling of primary aryl
amines with aryl halides. Angew. Chem. Int. Ed.
55, 13219–13223.
11. For the first nickel-catalyzed 1,1-diarylation of
ethylene, see Rollin, Y., Meyer, G., and Troupel,
M. (1982). Catalyse par des complexes du
nickel l’electrosynthese de 1,1-diarylethane a
partir d’halogenures aromatiques et
22. Meng, Q.-Y., Wang, S., and Konig, B. (2017).
¨
Carboxylation of aromatic and aliphatic
bromides and triflates with CO2 by dual visible-
light–nickel catalysis. Angew. Chem. Int. Ed. 56,
13426–13430.
d’ethylene. Tetrahedron Lett. 23, 3573–3576.
32. Paul, A., Smith, M.D., and Vannucci, A.K. (2017).
Photoredox-assisted reductive cross-coupling:
mechanistic insight into catalytic aryl-alkyl
12. Ohashi, M., Shirataki, H., Kikushima, K., and
Ogoshi, S. (2015). Nickel-catalyzed formation
of fluorine-containing ketones via the selective
cross-trimerization reaction of
23. Xue, F., Deng, H.P., Xue, C.W., Mohamed,
D.M.B., Tang, K.Y., and Wu, J. (2017). Reaction
discovery using acetylene gas as the chemical
feedstock accelerated by the ‘‘stop-flow’’
micro-tubing reactor system. Chem. Sci. 8,
3623–3627.
cross-couplings. J. Org. Chem. 82, 1996–2003.
tetrafluoroethylene, ethylene, and aldehydes.
J. Am. Chem. Soc. 137, 6496–6499.
33. Cyclic voltammetry (CV) studies have been
widely used in photoredox catalysis to
elucidate the oxidation state of transition
metals. However, we cannot exclude
13. Harper, M.J., Emmett, E.J., Bower, J.F., and
Russell, C.A. (2017). Oxidative 1,2-
difunctionalization of ethylene via gold-
catalyzed oxyarylation. J. Am. Chem. Soc. 139,
12386–12389.
24. Hou, J., Ee, A., Feng, W., Xu, J.H., Zhao, Y., and
Wu, J. (2018). Visible-light-driven alkyne
hydro-/carbocarboxylation using CO2 via
iridium/cobalt dual catalysis for divergent
heterocycle synthesis. J. Am. Chem. Soc. 140,
5257–5263.
endothermic steps in catalytic cycles Studer,
A., and Curran, D.P. (2016). Catalysis of radical
reactions: a radical chemistry perspective.
Angew. Chem. Int. Ed. 55, 58–102.
14. Prier, C.K., Rankic, D.A., and MacMillan, D.W.C.
(2013). Visible light photoredox catalysis with
transition metal complexes: applications in
organic synthesis. Chem. Rev. 113, 5322–5363.
34. For an example of a solvent effect study in dual
Ir-photoredox/Ni dual catalysis, see Oderinde,
M.S., Varela-Alvarez, A., Auila, B., Robbins,
D.W., and Johannes, J.W. (2015). Effects of
molecular oxygen, solvent, and light on
iridium-photoredox/nickel dual-catalyzed
cross-coupling reactions. J. Org. Chem. 80,
7642–7651.
25. Deng, H.P., Zhou, Q., and Wu, J. (2018). Micro-
tubing reactor-assisted aliphatic CꢁH
functionalization utilizing HCl as the hydrogen
atom transfer catalyst precursor in conjunction
with an organo photoredox catalyst. Angew.
Chem. Int. Ed. 57, 12661–12665.
15. Narayanam, J.M., and Stephenson, C.R. (2011).
Visible light photoredox catalysis: applications
in organic synthesis. Chem. Soc. Rev. 40,
102–113.
26. Tasker, S.Z., Standley, E.A., and Jamison, T.F.
(2014). Recent advances in homogeneous
nickel catalysis. Nature 509, 299–309.
16. Twilton, J., Le, C., Zhang, P., Shaw, M.H., Evans,
R.W., and MacMillan, D.W.C. (2017). The
merger of transition metal and photocatalysis.
Nat. Rev. Chem. 1, 0052.
35. Chen, F., Chen, K., Zhang, Y., He, Y., Wang,
Y.M., and Zhu, S. (2017). Remote migratory
cross-electrophile coupling and olefin
hydroarylation reactions enabled by in situ
generation of NiH. J. Am. Chem. Soc. 139,
13929–13935.
27. Tellis, J.C., Kelly, C.B., Primer, D.N., Jouffroy,
M., Patel, N.R., and Molander, G.A. (2016).
Single-electron transmetalation via
17. Corcoran, E.B., Pirnot, M.T., Lin, S., Dreher,
S.D., DiRocco, D.A., Davies, I.W., Buchwald,
S.L., and MacMillan, D.W.C. (2016). Aryl
amination using ligand-free Ni(II) salts and
photoredox catalysis. Science 353, 279–283.
photoredox/nickel dual catalysis: unlocking a
new paradigm for sp3-sp2 cross-coupling. Acc.
Chem. Res. 49, 1429–1439.
36. Luo, J., and Zhang, J. (2016). Donor-acceptor
fluorophores for visible-light-promoted
organic synthesis: photoredox/Ni dual catalytic
C(sp3)-C(sp2) cross-coupling. ACS Catal. 6,
873–877.
28. Deng, H.P., Fan, X.Z., Chen, Z.H., Xu, Q.H., and
Wu, J. (2017). Photo-induced nickel-catalyzed
chemo- and regioselective hydroalkylation of
internal alkynes with ether and amide a-hetero
C(Sp3)-H bonds. J. Am. Chem. Soc. 139, 13579–
13584.
18. Seo, H., Katcher, M.H., and Jamison, T.F.
(2017). Photoredox activation of carbon
dioxide for amino acid synthesis in continuous
flow. Nat. Chem. 9, 453–456.
37. Tsou, T.T., and Kochi, J.K. (1979). Mechanism of
oxidative addition. Reaction of nickel(0)
complexes with aromatic halides. J. Am. Chem.
Soc. 101, 6319–6332.
19. Majek, M., and Wangelin, A.J. (2015). Metal-
free carbonylations by photoredox catalysis.
Angew. Chem. Int. Ed. 54, 2270–2274.
29. A detailed study can be found in the
Supplemental Information.
20. Yatham, V.R., Shen, Y., and Martin, R. (2017).
Catalytic intermolecular
30. For Ni(I)-based catalytic cycle via Ir-
photoredox/Ni dual catalysis, see Oderinde,
M.S., Frenette, M., Robbins, D.W., Aquila, B.,
and Johannes, J.W. (2016). Photoredox
mediated nickel catalyzed cross-coupling of
38. He, Y., Cai, Y., and Zhu, S. (2017). Mild and
regioselective benzylic C–H functionalization:
Ni-catalyzed reductive arylation of remote and
proximal olefins. J. Am. Chem. Soc. 139, 1061–
1064.
dicarbofunctionalization of styrenes with CO2
and radical precursors. Angew. Chem. Int. Ed.
56, 10915–10919.
12
Chem 5, 1–12, January 10, 2019