Palladium-Catalyzed Homocoupling of Highly Fluorinated Arylboronates: Studies of the Influence of Strongly vs Weakly Coordinating Solvents on the Reductive Elimination Process

Article abstract of DOI:10.1021/jacs.9b11871

C-C reductive elimination from [PdL₂(C₆F₅)₂] to form polyfluorinated biaryls has been a challenge for over 50 years. Thus, palladium-catalyzed homocoupling of arylboronates (Ar_F-Bpin) containing two ortho-fluorine substituents is very difficult, as the reaction typically stops at the [PdL₂(Ar_F)₂] stage after two transmetalation steps. The transmetalated complexes cis-[Pd(MeCN)₂(C₆F₅)₂] (3a), cis-[Pd(MeCN)₂(2,4,6-C₆F₃H₂)₂] (3b), and cis-[Pd(MeCN)₂(2,6-C₆F₂H₃)₂] (3e) have been isolated from the reaction of Ar_F-Bpin with Pd(OAc)₂ in acetonitrile solvent, with no homocoupling observed. However, catalytic homocoupling proceeds smoothly in a "weakly coordinating" arene solvent as long as no ancillary ligands or coordinating solvents are present. DFT computations reveal that the active catalyst formed by arene solvent coordination leads to an overall reduced barrier for the reductive elimination step compared to the formation of stable [PdL₂(Ar_F)₂] complexes in the presence of a donor ligand or solvent L.

Full text of DOI:10.1021/jacs.9b11871

Subscriber access provided by Murdoch University Library
Article
Palladium-Catalyzed Homocoupling of Highly Fluorinated
Arylboronates: Studies of the Influence of Strongly vs. Weakly
Coordinating Solvents on the Reductive Elimination Process
Yudha Prawira Budiman, Arumugam Jayaraman, Friedrich
Alexandra, Florian Kerner, Udo Radius, and Todd B. Marder
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b11871 • Publication Date (Web): 05 Mar 2020
Downloaded from pubs.acs.org on March 6, 2020
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 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 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.
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 40
Journal of the American Chemical Society
1
2
3
4
Palladium-Catalyzed Homocoupling of Highly Fluorinated Arylboronates:
5
6
7
8
9
Studies of the Influence of Strongly vs. Weakly Coordinating Solvents on
the Reductive Elimination Process
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Yudha P. Budiman,a,b,‡ Arumugam Jayaraman,a,‡ Alexandra Friedrich,a Florian
Kerner,a Udo Radius,a,* Todd B. Mardera,*
aInstitut für Anorganische Chemie, and Institute for Sustainable Chemistry & Catalysis with Boron,
Julius-Maximilians-Universität Würzburg, Am Hubland, 97074 Würzburg, Germany
bDepartment of Chemistry, Faculty of Mathematics and Natural Sciences,
Universitas Padjadjaran, 45363 Jatinangor, Indonesia
ABSTRACT: C–C reductive elimination from [PdL2(C6F5)2] to form polyfluorinated biaryls has been
a challenge for over 50 years. Thus, palladium-catalyzed homocoupling of arylboronates (ArF–Bpin)
containing two ortho-fluorine substituents is very difficult as the reaction typically stops at the
[PdL2(ArF)2] stage after two transmetalation steps. The transmetalated complexes cis-
[Pd(MeCN)2(C6F5)2] (3a), cis-[Pd(MeCN)2(2,4,6-C6F3H2)2] (3b), and cis-[Pd(MeCN)2(2,6-C6F2H3)2]
(3e) have been isolated from the reaction of ArF–Bpin with Pd(OAc)2 in acetonitrile solvent, with no
homocoupling observed. However, catalytic homocoupling proceeds smoothly in a “weakly-
coordinating” arene solvent as long as no ancillary ligands or coordination solvents are present. DFT
computations performed reveal that the active catalyst formed by arene solvent coordination leads to
an overall reduced barrier for the reductive elimination step compared to the formation of stable
[PdL2(ArF)2] complexes in the presence of a donor ligand or solvent L.
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 40
1
2
3
4
INTRODUCTION
5
6
7
8
9
Many transition metal-catalyzed processes consist of steps such as oxidative addition, reductive
elimination, migratory insertion, transmetalation, and β-hydride elimination.1 Reductive elimination is
often important for the final product-releasing step. Generally, two electronic effects have been
observed for the reductive elimination from the [ML2(Ar)(Ar)] complexes containing group-10
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
metals. As the metal is formally reduced in the process, reductive elimination of Ar–Ar from these
complexes is usually faster if the ancillary ligand L is only weakly electron donating or strongly π-
accepting.2 Furthermore, the rates of reductive elimination from these complexes are higher when an
electron-rich and an electron-poor aryl group are involved in the reaction compared to two electron-
rich aryl groups. The slowest reductive elimination rate steps are typically found when both aryl
groups are electron-poor.2 In particular, [PdL2(C6F5)2] complexes are well known for their high
stability and, thus, reluctantly undergo reductive elimination,3 due to their strong Pd–Ar bonds.3g,h
Consequently, palladium-catalyzed homocoupling of C6F5B(OR)2,4 and Suzuki-Miyaura cross-
coupling of C6F5B(OR)2 with 2,6-C6F2+nH3-nX are found to be difficult.5 For example, Oliveira et al.
reported optimized conditions for the Pd-catalyzed homocoupling of Ar-BF3K in an aqueous solvent
which leads to coupled products in good to very good yields.4 Notably, if the substrates have two
ortho-fluorine substituents, e.g., C6F5BF3K or 2,6-C6F2H3BF3K, the corresponding homocoupling
products were not observed.4 Regarding Suzuki-Miyaura cross-coupling, Osakada et al.5b reported the
reaction of 2,4,6-trifluorophenylboronic acid with trans-[Pd(PEt3)2(C6F5)I] in the presence of Ag2O in
toluene/H2O which led to the formation of a stable intermediate trans-[Pd(PEt3)2(C6F5)(2,4,6-C6F3H2)]
and no reductive elimination was observed.
As early as in 1970, Rausch and Tibbets reported the synthesis of [PdL2(C6F5)2] (L = PPh2Me;
L2 = bpy)3f and, in 1964 and 1965, Rosevear and Stone reported that [NiL2(C6F5)2] (L = PEt3, PPh3; L2
= bpy)6 and [PtL2(C6F5)2] (L = PEt3, PPh3; L2 = dppe)7 are stable complexes. Jones et al.8a and Perutz
et al.8b,c reported that a high M–Ar bond energy can be achieved when two ortho-fluorine substituents
are present on the aryl group. Thus, as long as the two aryl groups contain two ortho-fluorine
2
ACS Paragon Plus Environment

Page 3 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
substituents, [PdL2(ArF)2] complexes are stable, e.g., [PdL2(C6F4(4-C5H10N))2] (L2 = dppe, cod)9 and
[PdL2(2,4,6-C6F3H2)2] (L = PhCN, PEt3, AsPh3, P(OMe)3, P(OEt)3, P(OPh)3; L2 = cod, tmen)10.
7
8
9
Scheme 1. Previous efforts to enhance reductive elimination from [PdL2(C6F5)2].
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Yamamoto et al., 2008
HNO3
(quantitative)
C6F5 C6F5
[PdL2(C6F5)2]
Sol, Temp
if L2 = cod; Sol = Solvent = THF:CH3Cl; HNO3 (1:3:1.3); RT; 4 h,
or L2 = dppe; Sol = toluene; 80 oC; 4 h
Espinet et al., 2016
+ 2L
C6F5H
C6F5 C6F5
+
+
PdL2
cis-[Pd(THF)2(C6F5)2]
toluene
25 o
C
L = phosphine ligand
if L = PtBu3, 4 h, C6F5-C6F5 = 98%, C6F5H = 2%
Alcarazo et al., 2017
Me
N
Me
Me
N
Me
H
N
N
N
N
N
Me
Me
P
P
N
Me
C6F5
C6F5
C6F5
C6F5
+
Pd
Pd
Me
MeCN
70 o
N
P
P
C
Ph
Ph
Ph
Ph
Me
Unsuitable for catalysis, as oxidative addition to Pd(0) occured at one of the C-P(H2Im)
bonds formed after reductive elimination to give the Pd(II) phosphinidene complex.
Due to the high stability of [PdL2(C6F5)2] complexes, researchers were encouraged to develop
efficient methods to enhance the rate of the reductive elimination step (Scheme 1). In 2008,
Yamamoto et al. demonstrated that addition of a Brønsted acid, HNO3, promotes the reductive
elimination of C6F5–C6F5 from cis-[PdL2(C6F5)2] (L2 = cod, bpy, and dppb).3g Although this strategy is
successful under stoichiometric conditions, it is not compatible with the catalytic homocoupling of
arylboronates, as basic reagents are required to activate Ar–B(OR)2 for transmetalation to the Pd
center.11
In 2016, Espinet et al.12 reported that reductive elimination of C6F5–C6F5 from cis-
[Pd(THF)2(C6F5)2] could be promoted by introduction of phosphine ligands, (PtBu3, o-TolPEWO-F,
tBuXPhos, P(C6F5)3, PhPEWO-F, P(o-Tol)3, tBuBrettPhos, Xantphos, and PhPEWO-H), in toluene
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 40
1
2
3
4
5
6
7
8
solvent. They suggested that immediate reductive elimination to form C6F5–C6F5 occurred after rapid
ligand substitution. Among different phosphines examined for efficient stoichiometric reductive
elimination, PtBu3 was ranked as the best ligand according to initial coupling rates.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In 2017, Alcarazo et al.3h attempted to solve the problem by synthesizing [PdL2(C6F5)2] and
employing two different α-dicationic chelating ancillary ligands, both containing a PPh2 moiety and a
strong π-acceptor [P(Me2Im)2]+2 unit (Me2Im = 1,3-dimethylimidazolidine-2-ylidene) connected via
o-phenylene or 2,2´-biphenylene linkers, respectively. Both complexes were calculated to have a low
energy barrier (~23 kcal/mol) for reductive elimination due to the strong acceptor character of the
ancillary ligands which enhances electron deficiency at the Pd center. However, they found these
complexes unsuitable for catalysis as, after reductive elimination, the resulting Pd(0) complex inserted
into one of the P–C bonds of the dicationic ligands to form Pd(II) phosphinidene complexes.
We recently reported the thermal13a and photocatalytic13b defluoroborylation of fluoroarenes
using an NHC-nickel complex as catalyst (NHC = N-Heterocyclic Carbene) and B2pin2 (pin =
pinacolato) as the boron source. Subsequently, we focused on applications of the resulting ArF–Bpin
products in organic synthesis. Very recently, we reported optimized conditions for Suzuki-Miyaura
cross-coupling of ArF–Bpin with aryl–iodide / bromide as a coupling partner, using a combination of
CuI and phenanthroline as the catalyst precursor,14 and the Ni-catalyzed Suzuki-Miyaura cross-
coupling of arylboronates with polyfluorinated arenes via C-F activation.13c Herein we expand the
applications of ArF–Bpin species to the homocoupling reaction using a palladium catalyst, as the
reductive elimination from Pd(II)(2,6-C6F2+nH3-n)2 complexes to generate symmetrical fluorinated
biphenyls is still very challenging. Symmetrical polyfluorinated biphenyls have been receiving
considerable attention as they have potential applications in organic semiconductors,15 electron-
transport materials,16 crystal-engineering,17 and supramolecular chemistry.18 Fluorinated organic
compounds, as well as some metal complexes with a fluorinated aryl ligand, are prospective
antiproliferative agents against HT-29 (colon carcinoma) and MCF-7 (breast adenocarcinoma).9
Thus, we report a strategy to enhance the reductive elimination rate of Pd(II)(2,6-C6F2+nH3-n)2
complexes in the Pd-catalyzed homocoupling of ArF–Bpin compounds. Using acetonitrile as the
4
ACS Paragon Plus Environment

Page 5 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
reaction solvent, we also isolated the complexes cis-[Pd(MeCN)2(C6F5)2] (3a), cis-[Pd(MeCN)2(2,4,6-
C6F3H2)2] (3b), and cis-[Pd(MeCN)2(2,6-C6F2H3)2] (3e) after double transmetalation, as these proved
resistant to reductive elimination. However, the Pd-catalyzed homocoupling of 2,6-C6F2+nH3-nBpin
substrates proceeds smoothly if no ancillary ligand or strongly coordinating solvent is used. DFT
computations were performed to gain additional insight into the reductive elimination step and to
understand why the catalytic homocoupling reactions conducted in arene solvents is efficient, whereas
it is difficult in the presence of ancillary ligand or coordinating solvents.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS AND DISCUSSION
Experimental studies. We began by examining the homocoupling reaction of 0.6 mmol of C6F5Bpin
catalyzed by 2 mol% of Pd(OAc)2 in the presence of stoichiometric Ag2O, which not only acts as an
oxidizer for in situ formed Pd(0), but is also suggested to act as an accelerator for the transmetalation
process.5b,19,20 For example, Osakada et al. reported that trans-[Pd(PEt3)2(C6F5)I] reacts with Ag2O in
toluene/H2O to generate trans-[Pd(PEt3)2(C6F5)(OH)], which undergoes facile transmetalation with
the boronic acid 4-MeOC6H4B(OH)2.5b As it is well documented that employing strong Lewis basic
ligands, such as mono- or bidentate phosphorus or nitrogen based ligands, stops the coupling reaction
by formation of stable [PdL2(C6F5)2] complexes,3 to facilitate the reductive elimination of C6F5–C6F5
we planned to conduct the coupling reaction in weakly coordinating solvents without adding any
ancillary ligands. It was anticipated that these solvents, upon coordination to the Pd center, would
provide a less electron-rich environment at the metal center, thereby facilitating reductive elimination
from the presumably four-coordinate cis-[Pd(solvent)2(C6F5)2] complex or three-coordinate cis-
[Pd(solvent)(C6F5)2] complex. It was previously proposed that acetonitrile can act as a labile ligand in
cis-[Pd(MeCN)(PMe3)(R)2] (R = Me, Ph, vinyl).21 This lability enables facile generation of the three-
coordinate complex [Pd(PMe3)(R)2] and the ensuing reductive elimination of R2. Conducting the
reaction in acetonitrile solvent, which was neither anhydrous nor dried or distilled, however, did not
lead to homocoupling (Scheme 2). Using dry acetonitrile did not help. Likewise, no homocoupling
product was formed with the arylboronates 2,4,6-C6F3H2Bpin (1b), 2,3,4,6-C6F4HBpin (1c), 2,3,5,6-
C6F4HBpin (1d), or 2,6-C6F2H3Bpin (1e) in this solvent (Scheme 2). However, with 2,3,4-C6F3H2Bpin
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 40
1
2
3
4
5
6
7
8
9
(1f), which has only one ortho-fluorine substituent, a quantitative formation of the homocoupled
product 2f was observed. From these data, it is clear that the reductive elimination is difficult only if
the ArF-Bpin contains two ortho-fluorine substituents.4,5b,9,10 This phenomena is in accordance with
the theoretical research by Jones et al.8a and Perutz et al.8b,c who showed that the strength of the M–
aryl bond strongly depends on the number of ortho-fluorine substituents on the aryl ring. Thus, it can
be suggested that diaryl complexes containing two ortho-fluorines on each aryl ring have a high Pd-
Ar bond energy and, in turn, a higher energy barrier for the reductive elimination step than those
containing only one or no ortho-fluorine substituent.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Pd-catalyzed homocoupling of ArFn–Bpin in MeCN.a
Fn
Fn
Fn
Pd(OAc)2 (2 mol%)
Ag2O (0.6 mmol)
Bpin
MeCN
5 h, 75 oC
2
1
(0.6 mmol)
F
F
F
F
F
F
F
F
F
Bpin
F
Bpin
Bpin
F
F
1a
1b
1c
Substrate:
Product:
b,c
b,c
b,c
2a
2b
( : 0%)
(
: 0%)
2c
(
: 0%)
F
F
F
F
F
Bpin
Bpin
F
Bpin
F
F
F
1e
1f
1d
Substrate:
Product:
b,c
c
b,c
2e
(
2f
: 92%)
: 0%)
(
2d
(
: 0%)
aGeneral condition: 0.6 mmol of ArF-Bpin (1), 2 mol% of Pd(OAc)2, 0.6 mmol of Ag2O, in 1.5 ml of MeCN, 5
h, at 75 ˚C, air. bNo product was observed. cIsolated yield.
The reaction of 1.2 mmol of C6F5Bpin (1a), 2,4,6-C6F3H2Bpin (1b), or 2,6-C6F2H3Bpin (1e)
with 0.4 mmol of Pd(OAc)2 and 1.6 mmol of Ag2O in 4 mL of MeCN for 16 h at 60 C stopped after
the double transmetalation step (Scheme 3), and the resulting complexes cis-[Pd(MeCN)2(C6F5)2]
(3a), cis-[Pd(MeCN)2(2,4,6-C6F3H2)2] (3b), and cis-[Pd(MeCN)2(2,6-C6F2H3)2] (3e) were isolated in
6
ACS Paragon Plus Environment

Page 7 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
good yields. All these products were characterized by NMR and IR spectroscopy, elemental analysis
and single-crystal X-ray diffraction (Figure 1 and Supporting Information).
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Synthesis of cis-[Pd(MeCN)2(ArFn)2].a
Fn
Fn
Pd(OAc)2 (0.4 mmol)
NCMe
Ag2O (1.6 mmol)
Bpin
Pd
MeCN
NCMe
16 h, 60 o
C
Fn
1
3
(1.2 mmol)
b
3a
: 75%)
ArFn = C6F5
(
b
3b
2,4,6-C6F3H2 ( : 80%)
b
3e
: 93%)
2,6-C6F2H3
(
aGeneral conditions: 1.20 mmol of ArF-Bpin, 0.40 mmol of Pd(OAc)2, 1.60 mmol of Ag2O, in 4 mL of MeCN,
16 h, 60 ˚C, in air. bIsolated yield.
Single crystals of cis-[Pd(MeCN)2(C6F5)2] (3a), cis-[Pd(MeCN)2(2,4,6-C6F3H2)2] (3b), and
cis-[Pd(MeCN)2(2,6-C6F2H3)2] (3e) were grown by vapor diffusion of CD3CN/Et2O solutions placed
in a larger vessel containing toluene (3a, 3e) and by vapor diffusion of an Et2O solution with hexane
(3b). The crystal structures, as shown in Figure 1, reveal that the complexes have cis-conformations,
as reported for other complexes such as cis-[PdL2(C6F5)2], (L = THF,22a CO,22a NH3,22b NH=CMe2,22b
3i
NH=C(OMe)Me,22c 7-azaindolyl22d and PPh3 ).
Figure 1. Molecular structures of 3a (left), 3b (middle), and 3e (right) determined by single-crystal X-ray
diffraction at 100 or 103 K. Ellipsoids are drawn at the 50% probability level, and H atoms as well as the
disordered diethyl ether solvent molecule are omitted for clarity. Only one of two symmetry independent
molecules of 3b is shown. Colors: red (palladium), green (fluorine), blue (nitrogen), white (carbon).
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 40
1
2
3
4
5
6
7
8
9
The isolation of complexes 3a, 3b, and 3e shows that even a weakly donating ligand such as
MeCN is sufficient to stabilize the [Pd(C6F5)2], [Pd(2,4,6-C6F3H2)2], and [Pd(2,6-C6F2H3)2]
complexes, formed after double transmetalation, and thus to inhibit reductive elimination. Yamamoto
et al.3g suggested that effective reductive elimination from [PdLn(C6F5)2] depends on weak Pd–
Cipso(C6F5) bonds and shorter Cipso(C6F5)–Cipso(C6F5) distances. Jones et al.8a and Perutz et al.8b,c
reported that the maximum increase in M–aryl bond strength is achieved with two ortho-fluorine
substituents, while the total number of fluorines only has a minor effect on the M-aryl bond strength.
As shown in Table 1, the Pd–Cipso(ArF) lengths, as well as Cipso(ArF)–Cipso(ArF) distances and
Cipso(ArF)–Pd–Cipso(ArF) angles are almost the same for cis-[Pd(MeCN)2(C6F5)2] (3a), cis-
[Pd(MeCN)2(2,4,6-C6F3H2)2] (3b), and cis-[Pd(MeCN)2(2,6-C6F2H3)2] (3e).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Selected bond lengths (Å) and angles (°) of 3a, 3b, and 3e.
3a
3b
3e
Molecule 1 Molecule 2
Pd–Cipso(ArF)
1.998(3)
1.988(2)
2.002(2)
2.773(3)
88.06(9)
98.77(8)
1.997(2)
1.997(2)
2.755(3)
87.23(9)
73.38(8)
1.996 (4)
1.998 (4)
2.752 (5)
87.09 (15)
86.76(11)
Cipso(ArF)·· Cipso(ArF) 2.774(5)
Cipso–Pd–Cipso
87.94(15)
72.75(13)
 Ar–Ar
We then continued to screen conditions for the homocoupling of C6F5Bpin (1a) (Table 2).
Intriguingly, shifting from coordinating solvents to non-coordinating ones23 such as toluene, benzene,
or m-xylene generated the homocoupling product C6F5–C6F5 (2a) in nearly quantitative yield (entry
1). The reaction is found to be slow in perfluorobenzene solvent, as it required 3 days to afford good
yield (entry 2). It should be noted that the arene solvents used were neither anhydrous nor dried or
distilled. In contrast to the outcomes obtained using arene solvents, no product was observed when
coordinating solvents such as THF or NEt3 were used (entry 3); presumably, stable [PdL2(C6F5)2] (L =
THF, NEt3) complexes were formed. In fact, the stable cis-[Pd(THF)2(C6F5)2]22a and cis-
[Pd(NH3)2(C6F5)2]22b complexes are well known. Using other coordinating solvents such as DMF or
8
ACS Paragon Plus Environment

Page 9 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
DMSO also failed to afford homocoupling product. In contrast to Pd(OAc)2, heterogenous Pd(0),
namely Pd/C, did not catalyze the homocoupling reaction (entry 4), indicating that our catalytic
process is not promoted by clusters, nanoparticles, etc. No homocoupling product was formed when
using PdCl2 instead of Pd(OAc)2 (entry 5). This unsuccessful outcome is likely due to the poor
solubility of PdCl2 in toluene.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Espinet et al. reported that addition of 2 equivalents of PtBu3 to cis-[Pd(THF)2(C6F5)2] in
toluene enhances the reductive elimination of C6F5–C6F5 under much milder conditions (Scheme 1).12
The authors suggested that fast ligand substitution with the bulky, strongly -donating PtBu3 ligand
leads to a three-coordinate Pd complex,24 cis-[Pd(PtBu3)(C6F5)2], and enhances the elimination of
C6F5–C6F5. Inspired by that, we added 4 mol% of PtBu3 as a ligand to the reaction mixture to examine
for the possibility of achieving efficient catalytic reaction under ambient conditions but,
unfortunately, it failed to generate any homocoupling product, even at a high temperature (Table 2,
entry 6). We did not screen any other phosphine ligands, as many complexes of the form
[Pd(PR3)2(C6F5)2] are known and are stable with respect to reductive elimination.3,5b
Table 2. Reaction conditions screened for Pd-catalyzed homocoupling of ArF-Bpin.a
F
F
Pd(OAc)2 (2 mol%)
Ag2O (0.6 mmol)
F
Bpin
C6F5 C6F5
Solvent, 75 oC
F
F
2a
1a
(0.6 mmol)
Entry
Solvent
Time (h) Yield (%)b
1
2
Toluene, benzene, or m-xylene
5
5
> 94
36
C6F6
24
72
5
54
75
0c
trace d
trace e
0c (0)c,f
3
4
5
6
THF, NEt3, DMF, or DMSO
Toluene
24
5
Toluene
Toluene + 4 mol% of PtBu3
24
aGeneral conditions: 0.60 mmol of 1a, 2 mol% of Pd(OAc)2, 0.60 mmol of Ag2O, in 1.5 mL of solvent, 5-72 h,
75 ˚C, in air. bIsolated yield. c Product was not observed. dUsing Pd/C (0.5, 5 or 10 mol%) in place of Pd(OAc)2.
eUsing PdCl2 (2 mol%) in place of Pd(OAc)2. fUnder argon and using dried toluene.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 40
1
2
3
4
5
6
7
8
9
We attempted to employ C6F5H instead of C6F5Bpin as a substrate in toluene under the above
optimized conditions, but no homocoupling product was generated (Scheme 4A),25 consistent with a
process in which the C-Bpin moiety is important for transmetalation, and that C-H oxidative addition
did not occur in toluene solvent. Interestingly, however, conducting the same reaction in MeCN
solvent afforded a 74% isolated yield of complex 3a (Scheme 4B), indicating that MeCN could
function as a base and/or a sufficiently strong donor ligand on Pd to activate the C-H bond of
C6F5H,25d but MeCN solvent has to be avoided if the reductive elimination product 2a is desired.
Notably, no conversion was observed when employing 1,3,5-trifluorobenzene and 1,5-
difluorobenzene as substrates, even in MeCN solvent (Scheme 4B).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Using toluene as the solvent under the above optimized conditions, we expanded the scope to
other fluorinated aryl-Bpin derivatives (Scheme 5). Compounds in which the C–Bpin bond is flanked
by two C–F bonds (1a, 1b, 1c, 1d, 1e), only one C–F bond (1f, 1g) or no ortho-fluorines (1h), all
proved to be suitable substrates, and the respective homocoupled products 2a, 2b, 2c, 2d, 2e, 2f, 2g,
2h were formed in excellent yields.
Scheme 4. Attempted homocoupling reaction using C6F5H, 1,3,5-C6F3H2, and 1,5-C6F2H3 substrates.
A
)
Pd(OAc)2 (2 mol%)
Ag2O (0.6 mmol)
F
F
F
H
C6F5
C6F5
Toluene
5 h, 75 oC
a
F
F
2a
(0%)
(0.6 mmol)
Fn
B
)
F
Pd
F
F
Pd(OAc)2 (0.4 mmol)
Ag2O (1.6 mmol)
NCMe
NCMe
Fn
F
F
H
MeCN
16 h, 60 oC
F
(4.0 mmol)
Fn
3
b
3a
ArFn : C6F5 ( , 74%)
b
3b
2,4,6-C6F3H2 ( , 0%)
b
3e
2,6-C6F2H3 ( , 0%)
aCondition: C6F5H (0.6 mmol), Pd(OAc)2 (2 mol%), Ag2O (0.6 mmol), toluene (1.5 ml), 75 C, 5 h, in air.
bConditions: ArFn (4 mmol), Pd(OAc)2 (0.4 mmol), Ag2O (1.6 mmol), MeCN (4 ml), 60 C, 16 h, in air.
10
ACS Paragon Plus Environment

Page 11 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Scheme 5. Reaction scope for the homocoupling of ArFn–Bpin in toluene.a
Fn
Fn
Fn
Pd(OAc)2 (2 mol%)
Ag2O (0.6 mmol)
Bpin
Toluene
5 h, 75 o
C
n: 0, 2-5
2
1
(0.6 mmol)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
F
F
F
F
F
F
F
F
F
2c
F
F
F
F
Bpin
F
Bpin
Bpin
F
Bpin
F
F
1a
1c
1d
Substrate:
Product:
1b
b
b
b
b
2a
(
, 97%)
F
2d
, 97%)
(
, 93%)
(
2b
(
, 96%)
CF3
F
F
F
Bpin
Bpin
F
Bpin
Bpin
CF3
F
1e
1g
1h
Substrate:
Product:
1f
b
b
b
b
2e
(
, 92%)
2g
2h
, 92%)
(
, 97%)
(
2f
(
, 97%)
aGeneral conditions: ArF-Bpin (0.6 mmol), Pd(OAc)2 (2 mol%), Ag2O (0.6 mmol), toluene (1.5 mL), 75 C,
5 h, in air. bIsolated yield.
Crystal and molecular structures of the fluorinated biaryl products: Intermolecular ππ
stacking interactions. Crystal structures of the products 2,2',3,3',5,5',6,6'-octafluorobiphenyl (2d) and
2,2',3,3',4,4'-hexafluorobiphenyl (2f) were obtained via single-crystal X-ray diffraction. The
molecular geometries of these compounds in their crystal structures are shown in Figure 2. The
central C–C bonds are equal within one standard deviation and are in the range 1.485(2) – 1.487(4) Å
(Table 3) which is typical of biphenyl compounds.26 The twist between the aryl moieties of the biaryl
compounds is slightly stronger in 2d (58.06(4)°) than in 2f (47.02(6)°) (Table 3). This difference is
likely due to the larger steric demand of fluorine in close vicinity to the central C–C bond joining the
rings and, hence, to the stronger repulsion between the fluorine atoms of both aryl units in 2d. A
similar twist angle to that of 2f was also reported for 2,3,4-trifluorobiphenyl (49.76(7)°).14
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 40
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Solid-state molecular structures of 2d (left) and 2f (right) determined by single-crystal X-ray
diffraction at 100 K. Ellipsoids are drawn at the 50% probability level, and H atoms are omitted for clarity.
Compound 2f shows a two-fold rotational symmetry. Colors: white (carbon), green (fluorine).
Table 3. Selected bond lengths (Å) and angles (°) of 2d and 2f, and ππ stacking distances in 2f.
2d
2f
conjunction bond CAryl–CAryl
 Aryl-Aryl
1.4852(16)
1.487(4)
58.06(4)
47.02(6)
3.6980(1)
3.3911(16)
1.475(4)
centroid-centroid distance
interplanar separation
offset shifta
-
-
-
aThe offset shift, also called inter-centroid shift, is the distance within a plane of an aryl ring between the
centroid of the respective aryl ring and the intersection point with the normal to the plane through the centroid
of the other aryl ring.
Particularly interesting in the crystal structure analysis are the intermolecular interactions and, hence,
molecular packing in these compounds. In compound 2d there are no π-π stacking interactions
present. A weak C–HF interaction with an angle of 124.62(8)° and a few weak FF interactions
with distances of 2.7078(12)–2.9169(15) Å are observed between the molecules (Table S2 in the
Supporting Information). However, in compound 2f, the biaryls form columns of offset face-to-face
π-stacked 2,3,4-trifluoro phenyl rings along the c axis. The interplanar separation between the
trifluorophenyl rings is typical for π-π stacking interactions (3.3911(16) Å, Table 3). Due to the large
twist angle of ca. 47° of the biaryl, the columnar stacks are cross-like (Figure 3). Similar cross-like
columnar stacks are observed in 2,3,4,5,6-pentafluorobiphenyl,27a 1,2,4,5-tetrafluoro-3-phenyl-
6(trifluoromethyl)benzene,27a
4-bromo-2,3,5,6-tetrafluorobiphenyl-4-carbonitrile,27a
and
2-
(perfluorophenyl)naphthalene14 which show biphenyl twist angles of 52, 50, 41, and 51°, respectively.
12
ACS Paragon Plus Environment

Page 13 of 40
Journal of the American Chemical Society
1
2
3
4
Weak intermolecular C–HF and FF interactions are formed between the stacks (Table S2 in the
5
6
Supporting Information).
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Crystal structure of 2f projected along the c axis (top), which is the stacking direction of the
molecules, and the a axis (bottom), at 100 K. Trifluorophenyl rings are -stacked along the c axis. The
intramolecular angle between the aryl rings of the biaryl is 47° leading to the formation of cross-like stacks
(bottom). Only a section of the crystal structure, i.e., one fourth along the a direction, is plotted in the bottom
projection, while four unit cells are shown along the c axis. Ellipsoids are drawn at the 50% probability level,
and H atoms are omitted for clarity. Colors: white (carbon), green (fluorine).
Computational studies. The observation that stable highly fluorinated diaryl palladium(II) complex
ligated with two acetonitriles, cis-[Pd(MeCN)2(C6F5)2] (3a), can be isolated from our attempted
catalytic homocoupling reactions conducted in acetonitrile suggests that the transmetalation process is
facile under the reaction conditions employed. It is also apparent that the transmetalation step is facile
for the reactions performed in other solvents, including THF, benzene, toluene, m-xylene and
probably perfluorobenzene. Thus, it can be proposed that the rate determining step in these
homocoupling reactions is reductive elimination. To understand why the catalytic homocoupling
reactions conducted in acetonitrile and THF failed, while they are successful in arene solvents,
computations on the reductive elimination step were performed using DFT at the B3LYP-
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 40
1
2
3
4
5
6
D3/Def2TZVP/6-311+g(2d,p)/IEFPCM // B3LYP-D3/SDD/6-31g**/IEFPCM level (see Supporting
Information for details).
7
8
9
Brief overview of previous theoretical studies on reductive elimination from cis-[PdL2RR’].
Several theoretical studies on C-H, C-C, C-E, and C-X reductive elimination from palladium
complexes of type cis-[PdL2R2] have been reported. However, in comparison to the oxidative addition
step, reductive elimination has been less well studied theoretically.28 For a long time, the reductive
elimination step was generally assumed to be fast and irreversible. However, recent experimental
studies have shown that there are systems in which reductive elimination is a slow step for the
formation of C-C bonds.29 Based on previous theoretical studies, reductive elimination from cis-
[PdL2R2] is generally accepted to be a concerted step occurring from a cis-complex. The earliest
detailed theoretical study on reductive elimination from d8-metal complexes of group 10 metals was
the 1981 report by Tatsumi et al.30 In this study, the authors employed extended Hückel calculations
with a weighted Hij approximation and analyzed how switching the transition metal within the same
group, as well as the ligands on the square planar complexes influences the activation barrier for the
reductive elimination of two cis-positioned alkyl groups. This study disclosed that, in comparison to
Pd(II) and Pt(II) complexes, Ni(II) complexes can readily facilitate the reductive elimination of ethane
from two cis-positioned methyl ligands. Moreover, this study also suggested that, upon dissociation of
one of the neutral ligands from cis-[ML2R2] complexes, the resulting three-coordinate complexes (cis-
[MLR2]) can adopt either a T- or Y-shape geometry. Complexes with a T-shape geometry, that have
the two eliminating groups positioned cis to each other, tend to undergo reductive elimination readily,
whereas complexes with a Y-shape geometry tend to facilitate the cis/trans isomerization of the two
alkyl groups.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Goddard et al. have explored theoretically, using the Hartree-Fock and generalized valence
bond approximations, the factors which facilitate the reductive coupling of H–H, C–H and C–C bonds
to produce molecular hydrogen and alkanes from the four-coordinate d8-complexes cis-[M(PH3)2RR]
(M = Pd, Pt; R, R’ = H, Me), and reported that the reductive elimination barrier increases in the order
H-H < C–H < C–C.31 For this trend, the authors noted that the faster reductive elimination of H2
14
ACS Paragon Plus Environment

Page 15 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
results from the better orbital overlap between the two eliminating hydride ligands, whereas the extent
of such overlap is reduced as we move towards more carbon-based eliminating ligands for which the
orbitals are more directed. For the C–C reductive elimination, Espinet et al. conducted a combined
theoretical and experimental study on Csp3–Csp3 and Csp2–Csp2 reductive elimination from cis-
[Pd(PMe3)2R2], and cis-[Pd(PMe3)LR2] complexes possibly formed in situ from addition of coupling
promoters (L = acetonitrile, ethene, maleic anhydride) to cis-[Pd(PMe3)2R2].21 Their computational
study at the B3LYP/SDD/6-31g* level of theory predicted a reductive elimination barrier of ~29
kcal/mol for the Csp3–Csp3 coupling from cis-[Pd(PMe3)2(Me)2], and a barrier of ~12 kcal/mol for Csp2–
Csp2 coupling from complexes of the type cis-[Pd(PMe3)2R2] (R = vinyl and Ph). Morukuma et al.
carried out a similar theoretical study, using the ONIOM(B3LYP/SDD/6-311g*:B3LYP/LANL2DZ)
level of theory, on the reductive elimination from Pd(II) and Pt(II) complexes by employing the
simplest phosphine ligand PH3 as a model.24b Their computations predicted a lower barrier for the
model complexes of the type cis-[Pd(PH3)2R2] (R = Me, vinyl, Ph, ethynyl) compared to their PMe3
analogues. For example, the reductive elimination barrier for Csp3–Csp3 coupling was reduced from
~29 kcal/mol to ~24 kcal/mol upon switching the phosphine ligand from PMe3 to PH3. Likewise, for
the vinyl case, the barrier was reduced from 12 kcal/mol to 6 kcal/mol when PMe3 is replaced with
PH3. However, for the bis-phenyl complex, the reductive elimination barrier was predicted to be
similar in both cases. While the reductive elimination barrier for Csp–Csp coupling from the PMe3
complex was not computed, it was calculated for the PH3 complex cis-[Pd(PH3)2(CCH)2] which was
found to have a barrier of ~11 kcal/mol, i.e., in between the barriers of its vinyl and methyl analogues,
with the vinyl complex having the lower barrier.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ariafard and Yates reported details of both electronic and steric effects of phosphine ligands
on the reductive elimination of ethane and butadiene from cis-[Pd(PR3)2R2] (R = Me, vinyl) at the
B3LYP/LANL2DZ/6-31g* level of theory.24a The authors concluded that both factors play a role in
the reductive elimination process for the parent four-coordinate complex, whereas only the electronics
dominated in the three-coordinate complex cis-[Pd(PR3)R2], formed from the parent complex via
dissociation of a PR3 ligand. Notably, steric factors in the starting four-coordinate complex cis-
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 40
1
2
3
4
5
6
7
8
9
[Pd(PR3)2R2] destabilizes it, while both factors stabilized the reductive elimination transition state.
These factors led to a reduced reductive elimination barrier for this step. However, this reduced
barrier is still high relative to that of the respective three-coordinate complex. Overall, from all these
computational studies, the reductive elimination barrier for C–C coupling from the four- and three-
coordinate Pd(II) complexes generally increases in the order Csp–Csp < Csp2–Csp2 < Csp3–Csp3. In 2010,
Korenaga et al. reported a combined experimental and theoretical study (theory level:
B3LYP/LANL2DZ/6-31g*) of the electronic and steric effects of various diphosphine ligands on
reductive elimination of biphenyl from the Pt(II) complex cis-[Pt(diphosphine)(Ph)2].32 The authors
suggested that the electronic effects of diphosphine ligands contribute mainly to the reductive
elimination rates. In particular, electron-poor diphosphine ligands that have the ability to act as a
weak σ-donor significantly reduced the reductive elimination barrier in comparison to the ones which
are strongly σ donating.32 It was reasoned that electron-poor diphosphine ligands decrease the d-
orbital energy gap between the important high-lying molecular orbital of the starting four-coordinate
complex and its reductive elimination transition state, and thereby reduce the reductive elimination
barrier. Thus far, much theoretical focus was placed on the reductive elimination of different C-C
single bonds and, to the best of our knowledge, there was only one report on a computational study of
Csp2–Csp2 reductive elimination from [PdL2(Ar)2] complexes containing fluorinated aryl ligands.3h
Alcarazo et al. reported a combined experimental and computational study of the reductive
elimination from symmetrical and dicationic unsymmetrical diphosphine ligand-containing Pd(II)
complexes of type cis-[Pd(diphosphine)(C6F5)2]. Their computations at the BP86-D3/def2-SVP level
of theory revealed that the dicationic unsymmetrical diphosphine ligand makes the coupling process
not only exergonic but also, remarkably, reduces the reductive elimination barrier to ~23 kcal/mol,
whereas employing a symmetrical ligand makes the process highly endergonic with a barrier of ~32
kcal/mol. By virtue of its strong π-accepting character, the cationic phosphine component of the
unsymmetrical diphosphine ligand was reasoned to be responsible for the reduced reductive
elimination barrier in the former case, as the cationic phosphine unit readily accepts the electron
density that is accumulated on the metal during the transition state for the reductive elimination
process.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
16
ACS Paragon Plus Environment

Page 17 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Computational Results and Discussion. It was previously demonstrated that the Pd-catalyzed
Suzuki-Miyaura cross-coupling of C6F5-boronates with aryl iodides or bromides requires a
stoichiometric amount of silver oxide (Ag2O) to accelerate the transmetalation between C6F5-
boronates and Pd(II) complexes. Ag2O has not been considered to play a role in the reductive
elimination from a diaryl palladium complex.5b,20 Other previous studies have indicated that the
acceleration of transmetalation with Ag2O in the presence of H2O led to the generation of an hydroxy
complex that shows a higher reactivity for transmetalation with arylboronates.5b,20c In our case, the
formation of a stoichiometric amount of the stable diaryl palladium(II) intermediate cis-
[Pd(MeCN)2(C6F5)2] (3a), cis-[Pd(MeCN)2(2,4,6-C6F3H2)2] (3b), and cis-[Pd(MeCN)2(2,6-C6F2H3)2]
(3e) from the homocoupling reactions conducted in acetonitrile solvent with excess Ag2O present
implies that Ag2O is ineffective in accelerating the reductive elimination process. Moreover, treatment
of 3e in the non-polar solvent toluene without addition of Ag2O led to decomposition and gave the
reductive elimination product 2e quantitatively, implying again that Ag2O does not accelerate
reductive elimination. Hence, our computational emphasis was placed exclusively on the reductive
elimination step without the engagement of Ag2O. As the reaction conducted in acetonitrile led to
intermediate 3a, two possible mechanistic routes from this intermediate, as shown in Figure 4, were
conceived for the coupling step. The first route involves a concerted reductive elimination pathway
whereby reductive elimination takes place directly from complex 3a, and the second route, a
dissociative pathway, involves an initial dissociation of an MeCN ligand,19 leading to the three-
coordinate Pd(II) complex cis-[Pd(MeCN)(C6F5)2], followed by reductive elimination. Computations
at the B3LYP-D3/TZVP level of theory show that this is a slightly endergonic reaction and the barrier
for reductive elimination directly from cis-[Pd(MeCN)2(C6F5)2] (3a) is 34.3 kcal/mol (pathway A,
Figure 4). The second route (pathway B, Figure 4) did not bring down the overall reductive
elimination barrier significantly with respect to that of pathway A, although the barrier of the actual
reductive elimination step itself via this route was reduced by 8.9 kcal/mol. As this coupling reaction
is already slightly endergonic and possess a high barrier achieving this reaction catalytically is
unlikely. Like acetonitrile, THF can also form the bis(solvent) complex cis-[Pd(THF)2(C6F5)2] (7).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 40
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Computed mechanistic pathways (DFT, B3LYP-D3/TZVP) for the C–C reductive elimination step
from the bis-perfluoroaryl complex ligated with acetonitrile. Gibbs free energy values (kcal/mol) relative to
reactant (3a) are given in parentheses.
Figure 5. Computed mechanistic pathways (DFT, B3LYP-D3/TZVP) for the C–C reductive elimination step
from the bis-perfluoroaryl complex ligated with THF. Gibbs free energy values (kcal/mol) relative to reactant
(7) are given in parentheses.
While the THF complex 7 was not isolated or observed in our reactions, it is a known compound
which was structurally characterized by X-ray crystallography.22a Therefore, for reactions conducted
in THF, the involvement of two routes resembling that of acetonitrile can be envisioned. For the first
route (Pathway A, Figure 5), computations predict a high barrier comparable to that of the acetonitrile
complex cis-[Pd(MeCN)2(C6F5)2] (3a). Even via the second route, i.e., the dissociative pathway, no
reduction in the overall barrier was obtained (pathway B, Figure 5). Furthermore, this homocoupling
reaction is highly endergonic. Overall, such catalytic homocoupling reactions attempted in THF will
likely not be rewarding. While this was the exact outcome we obtained from the experiments
conducted in THF, the recent demonstration by Espinet et al.12 of successful reductive coupling of
18
ACS Paragon Plus Environment

Page 19 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
perfluorobiphenyl using the synthesized THF complex cis-[Pd(THF)2(C6F5)2] (7) in toluene suggests
that either a THF ligand dissociates or both THF ligands were displaced by a toluene solvent prior to
the reductive elimination step. Only under one of these conditions could their reported experimental
barrier of 23.1 kcal/mol be obtained. Though not experimentally explored in the current work, we
have computationally explored the possibilities of reductive elimination from the
bis(perfluorophenyl)palladium complexes containing two other simple ligands, namely
dimethylsulfide and trimethylphosphine. Two mechanistic pathways, as discussed above for complex
cis-[Pd(MeCN)2(C6F5)2] (3a), are also possible with these ligands. Calculations show that, with these
two ligands, the reductive elimination barriers remain high for the first route (pathway A, Figures 6
and 7), and the overall barrier is even higher for the second route (pathway B, Figures 6 and 7).
Overall, our computations on reductive elimination from fluorinated bis-phenyl square planar
complexes of the type cis-[PdL2(C6F5)2] suggest that the barrier for reductive elimination from the
four-coordinate complex is too high to overcome under the reaction conditions we employed, and the
T-shaped tricoordinate complexes of type [PdL(C6F5)2] that result from the dissociation of L in cis-
[PdL2(C6F5)2], while possessing a lower barrier for the actual reductive elimination step, still place the
overall reductive elimination barrier at similar or higher energies than those of cis-[PdL2(C6F5)2].
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Computed mechanistic pathways (DFT, B3LYP-D3/TZVP) for the C-C reductive elimination step
from the bis-perfluoroaryl complex ligated with dimethylsulfide. Gibbs free energy values (kcal/mol) relative to
reactant (12) are given in parentheses.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 40
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Computed mechanistic pathways (DFT, B3LYP-D3/TZVP) for the C-C reductive elimination step
from the bis-perfluoroaryl complex ligated with PMe3. Gibbs free energy values (kcal/mol) relative to reactant
(16) are given in parentheses.
To understand the feasibility of the catalytic reactions in aromatic solvents, computations
were performed on two mechanistic pathways for reductive elimination: (i) with an aromatic solvent
coordinated to the metal; and (ii) without any arene coordination to the metal. Benzene, toluene, m-
xylene, and perfluorobenzene were explored as solvents. For the metal complexed with an aromatic
solvent, a slightly distorted η6-interaction was calculated for the ground state structure of [Pd(η6-
arene)(C6F5)2]. Ring slippage of the arene ligand along the reaction coordinate was calculated to occur
and the hapticity change may best be described as going from 6 in the reactant to n (n ≤ 3) in the
transition state and the products. The optimized structures of the benzene complex [Pd(6-
C6H6)(C6F5)2] 20, the primary elimination product 21 and of the transition state TS(20-21) (see also
Figure 9) are shown in Figure 8.
20
ACS Paragon Plus Environment

Page 21 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Side-view (a) and view along the arene-Pd vector (b) of the DFT-optimized structures (B3LYP-
D3/TZVP) of the benzene complexes [Pd(6-C6H6)(C6F5)2] 20, the primary elimination product 21 and the
transition state TS(20-21). Calculated Pd-Carene distances [Å]: 20: 2.61735, 2.61795, 2.64026, 2.64086, 2.66559,
2.66676, TS(20-21): 2.27466, 2.60792, 2.60934, 3.20505, 3.20613, 3.47771, 21: 2.23648, 2.56489, 2.63107,
3.20371, 3.25464, 3.51379.
For [Pd(ηn-arene)(C6F5)2] complexes 20, 25, 30 and 35 the overall reductive elimination process in all
of the aromatic solvents is exergonic. In benzene, a reductive elimination barrier of only 20.4
kcal/mol (ΔG‡) was found for the first route (pathway A, Figure 9). This overall barrier, in
comparison with that of the two routes for complexes that involve MeCN or THF solvent, is
exceptionally low and can be readily overcome under the reaction conditions employed
experimentally. The examination of a second route for reductive elimination led only to the
observation of an elevated overall barrier of 29.8 kcal/mol (pathway B, Figure 9). Thus, due to the
disparity in the barriers, it can be suggested that reductive elimination via the first route is
predominant for this reductive elimination process. The computed results for reductive elimination of
the two mechanistic routes involving the solvents toluene, m-xylene, and perfluorobenzene are
displayed in Figures 10, 11, and 12, respectively. The computed free energy profiles in toluene
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 40
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. Computed mechanistic pathways (DFT, B3LYP-D3/TZVP) for the C-C reductive elimination step
from the bis-perfluoroaryl complex ligated with benzene. Gibbs free energy values (kcal/mol) relative to
reactant (20) are given in parentheses.
and m-xylene follows a similar trend to that in benzene. However, unlike in benzene, toluene and m-
xylene, in perfluorobenzene the dissociation of the perfluorobenzene from the postulated reactant 35
was found to be exergonic. Therefore, the overall barriers for both mechanistic routes which were
found to be similar to each other come from the dissociated Pd(C6F5)2 complex 38. As a result, both
pathways might be possible. Despite the lowest reductive elimination barrier predicted in this solvent,
catalytic experiments provided lower conversions in comparison with the outcomes of catalytic
experiments carried out in other aromatic solvents in the same reaction times (Table 2, entry 2).
Prolonging the reaction time did improve the desired outcome of this reaction and afforded the
product in a good yield (75%). One reason for such a slow conversion may be the deactivation of the
catalyst, possibly via oxidative addition of a C–F bond of the coordinated perfluorobenzene solvent
after the reductive elimination step, or the poor ability of the oxidizing reagent Ag2O to oxidize the
electron-poor Pd(0) complex 36 that has formed after the reductive elimination step. The computed
barrier for a C-F oxidative addition process was found to be only 24.8 kcal/mol (TS(36-40), Figure
12), which is accessible under our reaction conditions. However, treating the cis-[Pd(MeCN)2(2,6-
C6F2H3)2] complex (3e) in toluene with 3 equiv of perfluorobenzene at 75 ˚C for a day did not show
any indication of C-F oxidative addition to Pd(0) even though we obtained the coupled product 2e in
>99% yield (see Supporting Information).
22
ACS Paragon Plus Environment

Page 23 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Computed mechanistic pathways (DFT, B3LYP-D3/TZVP) for the C-C reductive elimination step
from the bis-perfluoroaryl complex ligated with toluene. Gibbs free energy values (kcal/mol) relative to reactant
(25) are given in parentheses.
Figure 11. Computed mechanistic pathways (DFT, B3LYP-D3/TZVP) for the C-C reductive elimination step
from the bis-perfluoroaryl complex ligated with m-xylene. Gibbs free energies (kcal/mol) relative to reactant
(30) are given in parentheses.
Figure 12. Computed mechanistic pathways (DFT, B3LYP-D3/TZVP) for the C-C reductive elimination step
from the bis-perfluoroaryl complex ligated with perfluorobenzene and a possible mode of deactivation of the
catalyst. Gibbs free energy values (kcal/mol) relative to reactant (35) are given in parentheses.
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 40
1
2
3
4
As the reductive elimination process generates a lone pair on the transition metal, the level of
5
6
2
2
stabilization of this lone pair (dx -y ) was examined by analyzing the interactions between the ligands
7
8
L2 and n-Ar and the bent (C2v) fragment [Pd(C6F5)2] of cis-[PdL2(C6F5)2] (L = MeCN, THF, SMe2,
9
PMe3) and [Pd(6-Ar)(C6F5)2] (Ar = C6H6, C7H8, C8H10, C6F6) and their corresponding reductive
elimination transition states. Thus, EDA-NOCV calculations,33 performed at the B3LYP-D3/TZVP
level of theory, offer detailed information on the interactions in terms of Pauli repulsion (ΔEPauli),
attractive electrostatic interaction (ΔEelstat) and orbital interaction (ΔEorb) and also the attractive
dispersion interaction (ΔEdisp) (Tables S3 and S4 in the Supporting Information). More explanations
of these terms are provided in the EDA-NOCV analysis section in the Supporting Information. Table
4 summarizes the results of our calculations. The contribution to the overall barrier by distortion
energy (ΔΔEdist), which is the difference in energy penalty linked with the distortion of the interacting
fragments from their equilibrium geometries to the frozen geometries of the reactant and transition
state, was found to be similar in all cases with calculated ΔΔEdist values between +17.5 and +20.7
kcal/mol (Table 4, column a). Therefore, the differences observed for the barriers of the reductive
elimination step for different ligands can be explained by contributions from the intrinsic interaction
energies (ΔEint = ΔEPauli + ΔEelstat + ΔEorb + ΔEdisp). The interaction energy ΔEint (Table 4, column b) of
the reactants decreases in the order [Pd(PMe3)2(C6F5)2] 16 (-102.0 kcal/mol) > [Pd(SMe2)2(C6F5)2] 12
(-70.8 kcal/mol) > [Pd(THF)2(C6F5)2] 7 (-58.1 kcal/mol) > [Pd(MeCN)2(C6F5)2] 3a (-55.3 kcal/mol) >
[Pd(n-C8H10)(C6F5)2] 30 (-32.6 kcal/mol) ≈ [Pd(n-C7H8)(C6F5)2] 25 (-30.3 kcal/mol) ≈ [Pd(n-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
C6H6)(C6F5)2] 20 (-27.4 kcal/mol) > [Pd(n-C6F6)(C6F5)2] 35 (-12.2 kcal/mol), which reflects the
different behavior of the ligands.
ΔEint was reduced significantly for cis-[PdL2(C6F5)2] when moving from the reactant to the
reductive elimination transition state as indicated by the differences Eint. The interaction energy
term between the ligands L2 and [Pd(C6F5)2] decreases for L = MeCN (3a, Eint= +19.9 kcal/mol),
THF (7, Eint= +21.3 kcal/mol)), SMe2 (12, Eint= +18.7 kcal/mol)) and PMe3 (16, Eint= +25.0
kcal/mol) (Table 4, column b), whereas Eint remains almost constant for [Pd(n-Ar)(C6F5)2] (+1.2
to +4.3 kcal/mol) and even increases for [Pd(n-C6F6)(C6F5)2] (-4.7 kcal/mol). A closer inspection of
24
ACS Paragon Plus Environment

Page 25 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
the sum of the electrostatic contributions ΔEPauli and ΔEelstat reveals that it varies little, and remains
almost constant going from the ground state to the transition state. The sum of ΔEPauli and ΔEelstat is
+3.2 kcal/mol for [Pd(MeCN)2(C6F5)2], +4.8 kcal/mol for [Pd(THF)2(C6F5)2], +1.5 kcal/mol for
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4. Results of EDA-NOCV analysis (B3LYP-D3/TZVP level of theory) on the interaction
between the ligands L2 and n-Ar and the bent fragment (C2v) [Pd(C6F5)2] of various reactants [Pd
L2(C6F5)2] (L = MeCN, THF, SMe2, PMe3) and [Pd(n-Ar)(C6F5)2] (Ar = C6H6, C7H8, C8H10, C6F6)
and their corresponding transition states of the reductive elimination of perfluorobiphenyl from these
complexes. Energies are given in kcal/mol.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
ΔEdist ΔEint
ΔEPauli ΔEelstat ΔEorb ΔEσ
ΔEπ
ΔEorb(rest) ΔEdisp
L/Ar
MeCN (3a)
TS(3a-4)
1.2
-55.3
-35.4
130.6
106.5
-117.1
-89.8
-60.6
-45.1
-10.3
-9.9
-8.2
-7.0
-41.7
-21.1
-8.6
17.5
-14.2
TS(3a-4)-3a
(ΔΔE)
16.3
1.8
19.9
-24.1
102.5
80.0
27.3
15.5
20.6
-5.6
0.4
1.2
a
THF (7)
-58.1
-36.8
-95.6
-68.3
-44.2
-27.8
-15.3
-9.7
-20.8
-20.7
-29.1
-18.1
–
a
TS(7-8)
17.9
–
TS(7-8)-7
(ΔΔE)
16.1
2.7
21.3
-22.5
162.7
163.9
27.3
16.4
11.0
–
5.6
0.1
SMe2 (12)
-70.8
-52.1
-141.2
-141.0
-70.6
-54.5
-19.3
-12.8
-21.7
-20.6
-47.4
-37.7
-3.9
-4.0
TS(12-13)
17.8
TS(12-13)-12
(ΔΔE)
15.1
5.6
18.7
1.2
0.3
16.1
9.7
-0.1
6.5
1.1
PMe3 (16)
-102.0
-77.0
280.0
281.1
-250.1
-252.6
-104.7
-81.4
-15.0
-11.9
-27.2
-24.1
-79.2
-57.6
-10.5
-11.9
TS(16-17)
18.5
TS(16-17)-16
(ΔΔE)
12.9
1.6
25.0
1.1
-2.5
23.2
21.6
-1.4
3.1
3.1
C6H6 (20)
-27.4
64.5
-45.5
-35.0
-9.2
-11.3
-22.8
-14.2
-3.0
TS(20-21)
19.2
-26.2
66.8
-53.1
-31.3
-4.7
-8.6
-12.4
TS(20-21)-20
(ΔΔE)
17.6
1.8
1.2
2.3
-7.6
3.7
8.6
-9.4
4.5
2.7
C7H8 (25)
-30.3
-27.5
67.8
68.0
-48.5
-54.6
-36.6
-31.8
-10.0
-4.8
-12.9
-9.2
-23.4
-14.4
-3.2
TS(25-26)
19.2
-12.5
TS(25-26)-25
(ΔΔE)
17.4
1.9
2.8
0.2
-6.0
4.8
9.0
-9.3
5.2
3.8
C8H10 (30)
-32.6
-28.4
69.1
68.3
-50.5
-55.2
-37.6
-30.7
-13.5
-5.0
-13.6
-10.7
-20.8
-14.1
-3.3
TS(30-31)
19.1
-11.5
TS(30-31)-30
(ΔΔE)
17.2
4.3
-0.8
-4.7
6.9
6.7
-8.2
8.5
2.9
C6F6 (35)
1.3
-12.2
-16.9
53.6
68.9
-25.4
-41.5
-28.9
-30.1
-8.4
-4.8
-11.6
-14.2
-18.4
-9.7
-2.1
TS(35-36)
20.7
-15.5
TS(35-36)-35
(ΔΔE)
19.4
-4.7
15.2
-16.2
-1.1
8.7
-13.4
3.6
-2.6
a
No π back-donation interaction was found. Details about deriving different orbital interaction components
(ΔEσ, ΔEπ and ΔEorb(rest)) from the EDA-NOCV analysis are provided in the Supporting Information.
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 40
1
2
3
4
5
6
7
8
9
[Pd(SMe2)2(C6F5)2], -1,4 kcal/mol for [Pd(C6F5)2(PMe3)2], -5.3 kcal/mol for [Pd(n-C6H6)(C6F5)2], -
5.8 kcal/mol for [Pd(n-C7H8)(C6F5)2], -5.5 kcal/mol for [Pd(n-C8H10)(C6F5)2], and -1.0 kcal/mol for
[Pd(n-C6F6)(C6F5)2] (Table 4, columns d and e). The change in Eint is mainly attributable to a
change in the orbital interaction energies ΔEorb on going from the ground state to the transition state
(Table 4, column e).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The orbital interactions (ΔEorb) in our systems are comprised of three different components: (i) σ
donation interaction (ΔEσ) from neutral ligand(s) to the Pd(ArF)2 fragment (column f, Table 4); (ii) π
back-donation interaction (ΔEπ) from the Pd(ArF)2 fragment to neutral ligand(s) (column g, Table 4);
and (iii) the rest of the orbital interactions (ΔEorb(rest)), which mostly entails the electronic polarization
within each fragment. The difference in the orbital interactions Eorb of cis-[PdL2(C6F5)2] is +15.5
kcal/mol for L = MeCN, +16.4 kcal/mol for L= THF, +16.1 kcal/mol for L = SMe2 and +23.2
kcal/mol for L = PMe3 (Table 4, column e). This significant reduction in ΔEorb in the transition states
is mainly attributable to a decreased -interaction (Table 4, column f): (3a, E = +20.6 kcal/mol),
THF (7, E = +11.0 kcal/mol)), SMe2 (12, E = +9.7 kcal/mol)) and PMe3 (16, E = +21.6
kcal/mol). Eorb increases by only a small amount (-1.1 - +6.9 kcal/mol) for [Pd(n-Ar)(C6F5)2], as a
result of the ring slippage and an increase in -acceptance in the transition state, as the -contribution
becomes favorable by -9.4 kcal/mol for [Pd(n-C6H6)(C6F5)2], -9.3 kcal/mol for [Pd(n-C7H8)(C6F5)2],
-8.2 kcal/mol for [Pd(n-C8H10)(C6F5)2], and -13.4 kcal/mol for [Pd(n-C6F6)(C6F5)2] (Table 4,
columns e and g). Although the difference in orbital contributions ΔΔEorb in all cases can be explained
using mainly the contributions from σ donation (ΔΔEσ) and π back-donation (ΔΔEπ) interactions,
impact from the rest of the orbital interactions (ΔΔEorb(rest)) is also sizable (Table 4, column h),
suggesting that there exist considerable polarization effects. The attractive energy contribution due to
dispersion interactions, which arises from the attractive forces between the induced dipoles of the
interacting fragments, in both reactant and transition state of all cases was found to be substantial
albeit slightly less pronounced in the transition states (Table 4, column i). This contribution to the
overall barrier is contained in ΔEint. Changes in the dispersion interaction Edisp play only a minor
26
ACS Paragon Plus Environment

Page 27 of 40
Journal of the American Chemical Society
1
2
3
4
5
6
role in the increase in the reductive elimination barrier, as the maximum difference between reactant
and transition state is only +3.8 kcal/mol (Table 4, column i).
7
8
9
According to our calculations, the lower barrier for the complexes [Pd(n-Ar)(C6F5)2] is due
to: (i) ring slippage of the arene ligand and the adjustment of the arene ring to the electronic situation
at the metal which leads to less -repulsion in the reductive elimination transition state; and (ii) to
more favorable π back-bonding from Pd(ArF)2 to the arene fragment in the transition state. These
factors lead to a lower lying transition state for [Pd(n-Ar)(C6F5)2]. Furthermore, the 2e- donor ligands
of [Pd L2(C6F5)2] stabilize the ground state much better than n-Ar ligand in [Pd(n-Ar)(C6F5)2] and
both factors, destabilization of the ground state of Pd(C6F5)2 and stabilization of the reductive
elimination transition state of Pd(C6F5)2 for n-Ar in comparison to L, lead to a significantly reduced
barrier for reductive elimination.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Overall, insights gained into the bonding interactions between various ligands and the
Pd(ArF)2 fragment using the EDA-NOCV method suggest that it is the weakly donating and strongly
accepting characters of arene ligands that minimize the reductive elimination barrier compared to
stronger donor and weaker acceptor ligands L. Ligands such as MeCN, THF, SMe2 and PMe3
stabilize the (electron poor) Pd(II) ground state more efficiently than an arene ligand, whereas the
arenes stabilize the developing charge in the transition state better than L. Both, destabilization of the
ground state and stabilization of the transition state leads to lower barriers for the reductive
elimination process.
Catalytic cycle of the homocoupling reaction. A plausible mechanism for the homocoupling
reaction of fluorinated arylboronates conducted in arene solvent is depicted in Figure 13. Ag2O likely
plays two roles. First, in the presence of trace moisture, it assists in the transmetalation process,5b,20c
generating species C and two equivalents of HOBpin. The latter can condense yielding pinBOBpin,
regenerating water. Second, following reductive elimination of the product biaryl from C, Ag2O
oxidizes Pd(0) to Pd(II) to restart the cycle. Indeed, a silver mirror forms during the reaction.
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 40
1
2
3
4
Pd(OAc)2
5
6
Ag2O + benzene
7
8
+ H2O
2 AgOAc
OH
PdII
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
A
2 Ag0 + benzene
Ag2O
ArFBpin
ArF
PdII
H2O
+
O(Bpin)2
HOBpin
+
HOBpin
Pd0
OH
D
B
ArFBpin
ArF-ArF
ArF
PdII
ArF
C
Figure 13. Proposed mechanism for the palladium-catalyzed homocoupling of fluorinated
arylboronates.
28
ACS Paragon Plus Environment

Page 29 of 40
Journal of the American Chemical Society
1
2
3
4
CONCLUSIONS
5
6
7
8
9
In summary, we demonstrated that the palladium-catalyzed homocoupling of highly fluorinated aryl-
boronate esters containing two ortho-fluorine substituents does not proceed in coordinating solvents
because the reaction stalls at the cis-[PdL2(ArF)2] stage, as demonstrated by the isolation of the stable
complexes cis-[Pd(MeCN)2(C6F5)2] (3a), cis-[Pd(MeCN)2(2,4,6-C6F3H2)2] (3b), and cis-
[Pd(MeCN)2(2,6-C6F2H3)2] (3e). This investigation also introduces the use of arylboronates instead of
aryl Grignard or lithium substrates3,9 to synthesize [PdL2(ArF)2] complexes. Avoiding ligands and
employing very weakly coordinating solvents (“non-coordinating”) such as toluene, benzene or m-
xylene allows the catalytic homocoupling reaction to proceed smoothly, providing fluorinated
biphenyl products.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DFT computations predict that catalytic homocoupling reactions of highly fluorinated
arylboronate esters conducted in acetonitrile, THF, SMe2 and PMe3 are unfeasible, as the overall
reductive elimination process is endergonic, and the reductive elimination step has a high barrier.
While this is the case with strong σ donating solvents and ligands, computations indicate that coupling
reactions carried out in aromatic solvents, which are, in general, weak σ donors and π acceptors,
should be highly feasible, as the reductive elimination process is not only sufficiently exergonic, but
also the overall barrier for reductive elimination was found to be <21 kcal/mol. The lower barrier for
the reductive elimination from [(n-Ar)Pd(C6F5)2] is due to: (i) ring slippage of the arene ligand and
the adjustment of the arene ring to the electronic situation at the metal which leads to less -repulsion
in the reductive elimination transition state; and (ii) more favorable π back-bonding from Pd(ArF)2 to
the arene fragment in the transition state. Inspection of the reaction coordinates of the homocoupling
process conducted in different solvents nicely explains our experimental results which showed that no
catalysis takes place in acetonitrile and THF solvents, whereas catalytic homocoupling is efficient in
aromatic solvents.
29
ACS Paragon Plus Environment