A Role for Pd(IV) in Catalytic Enantioselective C-H Functionalization with Monoprotected Amino Acid Ligands under Mild Conditions

Article abstract of DOI:10.1021/jacs.7b03716

Kinetic and mechanistic studies of the desymmetrization of benzhydrylamine using Pd/monoprotected amino acid ligands (Pd/MPAA) via C-H functionalization with molecular iodine provide mechanistic insight into the rate-determining step and the oxidation state of Pd in the C-H functionalization step. Enantiomeric excess is strikingly insensitive to temperature from ambient temperature up to over 70 °C, and reaction rate is insensitive to the electronic characteristics of the ligand's benzoyl protecting group. The reaction is highly robust with no evidence of catalyst deactivation. Intriguingly, C-H bond breaking does not occur prior to the addition of I₂ to the reaction mixture. Electrochemical experiments demonstrate the viability of oxidative addition of I₂ to Pd(II). Together with ¹⁹F NMR studies, these observations suggest that iodine oxidizes Pd prior to addition of the amine substrate. This work may lead to a better general understanding of the subtle variations in the reaction mechanisms for C-H functionalization reactions that may be extant for this ligand class depending on substrate, amino acid ligand and protecting group, and reaction conditions.

Full text of DOI:10.1021/jacs.7b03716

Subscriber access provided by UNIV LAVAL
Article
A Role for Pd(IV) in Catalytic Enantioselective C-H Functionalization
With Monoprotected Amino Acid Ligands Under Mild Conditions
R. Erik Plata, David E. Hill, Brandon E. Haines, Djamaladdin G Musaev, Ling Chu,
David P. Hickey, Matthew S. Sigman, Jin-Quan Yu, and Donna G Blackmond
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 12 Jun 2017
Downloaded from http://pubs.acs.org on June 12, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
A Role for Pd(IV) in Catalytic Enantioselective C-H Functionaliza-
tion With Monoprotected Amino Acid Ligands Under Mild Condi-
tions
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
R. Erik Plata,1 David E. Hill,1 Brandon E. Haines,2 Djamaladdin G. Musaev,2 Ling Chu,1 David P.
Hickey,3 Matthew S. Sigman,3 JinꢀQuan Yu,1 and Donna G. Blackmond1*
1Department of Chemistry, The Scripps Research Institute, La Jolla, California 92037 USA ; 2Cherry L. Emerson Center for
Scientific Computation, Emory University, Atlanta, Georgia 30322, USA, 3Department of Chemistry, University of Utah,
Salt Lake City, Utah 84112
ABSTRACT: Kinetic and mechanistic studies of the desymmetrization of benzhydrylamine using Pd/monoprotected amino acid
ligands (Pd/MPAA) via CꢀH functionalization with molecular iodine provide mechanistic insight into the rateꢀdetermining step and
the oxidation state of Pd in the CꢀH functionalization step. Enantiomeric excess is strikingly insensitive to temperature from ambiꢀ
ent temperature up to over 70 oC, and reaction rate is insensitive to the electronic characteristics of the ligand’s benzoyl protecting
group. The reaction is highly robust, with no evidence of catalyst deactivation. Intriguingly, CꢀH bond breaking does not occur
prior to the addition of I2 to the reaction mixture. Electrochemical experiments demonstrate the viability of oxidative addition of I2
to Pd(II). Together with 19F NMR studies, these observations suggest that iodine oxidizes Pd prior to addition of the amine subꢀ
strate. This work may lead to a better general understanding of the subtle variations in reaction mechanism for CꢀH functionalizaꢀ
tion reactions that may be extant for this ligand class depending on substrate, lamino aicd igand and protecting group, and reaction
conditions.
INTRODUCTION
the most efficient for the reaction of Scheme 1, our work foꢀ
cused on this MPAA system. NMR studies of the interaction
of Pd with the ligand in the presence of base revealed that
deprotonation of the amino acid NꢀH occurred rapidly. The
substrate 1 is also deprotonated in the presence of base even in
the absence of catalyst. Intriguingly, however, we were unable
to detect interactions between catalyst and substrate. Substrate
titration showed no changes in the 1HꢀNMR spectrum. In addiꢀ
tion, after mixing deutero 1-D with PdꢀBzꢀLeuꢀOH and base in
CꢀH activation reactions employing broadly useful subꢀ
strates and the concept of weak coordination have gained atꢀ
tention in recent years.1 In particular, triflylꢀprotected diarylaꢀ
mines have demonstrated broad scope as substrates in enantiꢀ
oselective iodination via both desymmetrization (Scheme 1)2
and kinetic resolution.3 Those reports followed on from the
finding that a weaklyꢀcoordinating amide auxiliary could efꢀ
fect the orthoꢀiodination of benzamides using I2 as the sole
oxidant.4 The mild conditions of these reactions stand out in
comparison to many of the other reactions employing monoꢀ
protected amino acid (MPAA) ligands, prompting us to study
the reaction in Scheme 1 in detail to probe for deeper underꢀ
standing. The present investigation reports mechanistic studies
that suggest a a novel role for Pd(IV) in the CꢀH activation
process. This finding should enable further development of
similar reactions.
the DMSO/protio solvent mixture, or protio
1 from the
DMSO/deuteron solvent mixture, recovery of 1 was quantitaꢀ
tive and no isotopic scrambling was observed, indicating both
that substrate binding is very weak and that reversible CꢀH
cleavage had not occurred (Scheme 2).
Scheme 2. Substrate Interaction with Catalyst
Scheme 1. Iodination of Diarylmethylamines
10 mol% Pd(OAc)2
20 mol% Bz-Leu-OH
NHTf
NHTf
+ I2
CsOAc (3 equiv)
Na2CO3 (3 equiv)
DMSO
H
I
1
2
3
t-AmylOH
50 °C
Further NMR studies revealed that interaction between the
Pd salt and the ligand led to a rapid and quantitative intramo-
lecular CꢀH activation process between the benzoyl protecting
group of the bound ligand and Pd to form palladacycle 4, as
RESULTS AND DISCUSSION
Because optimization studies identified the leucine amino
acid/benzoyl protecting group combination (LeuꢀBzꢀOH) as
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 10
1
2
3
4
5
6
7
8
9
confirmed by 1H and 13CꢀDEPTꢀQ experiments. Palladacycle 4
Figure 2 shows temporal product concentration profiles for
“different excess” experiments. Figure 2a reveals that rate
depends on [1] in a manner that is between 0 and 1st order,
characteristic of saturation kinetics in [1]. The upper left inset
plots the data as MichaelisꢀMenten kinetics giving a binding
equilibrium constant of Keq ≈ 10 Mꢀ1. The right inset confirms
this by plotting the data according to the Bures method7 of
determinging power law order in [1] to give n = 0.8. Figure
2b shows that the reaction exhibits zero order kinetics in [2].
represents a stable catalyst resting state, and we postulated that
the lack of substrate binding and CꢀH activation noted above
could be due to the inertness of 4. Addition of I2, may iodinate
the benzoyl group to form the active catalyst prior to binding
and reaction of substrate 1. When the reaction in Scheme 1
was carried out using 2,6ꢀdifluorobenzoylꢀprotected leucine as
ligand, where the ortho sites of the ligand are blocked, the
reaction proceeds at the same rate and with the same enantiꢀ
oselectivity as the BzꢀLeuꢀOH ligand. Thus both catalyst sysꢀ
tems are employed in these mechanistic studies.
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
a)
Reactions aimed at establishing catalyst robustness and conꢀ
centration dependences of substrates 1 and 2 were carried out
according to the “same excess” and “different excess” protoꢀ
cols of Reaction Progress Kinetic Analysis (RPKA),5 with
excess defined in eq 1.
(1)
b)
excess = 2 − 1
[
]
[
]
[ ]0
0
Figure 1 shows the kinetic profiles for experiments carried
out using the “same excess” protocol. The timeꢀadjusting
curves, as marked by the arrows in Figure 1, allows compariꢀ
son of the kinetic profiles for the two reactions as they react
under identical substrate concentrations onwards from the
point of intersection of the arrows. The fact that the profiles
overlay indicates that reaction rate is not influenced by either
the additional catalyst turnover or the presence of a higher
concentration of product in the reaction vial for the case of run
a) compared to run b). This confirms that neither catalyst deꢀ
activation nor product inhibition is present in this robust cataꢀ
lyst system.
Figure 2. Different excess experiments to determine reaction
orders in [1] and [2]. Catalyst/ligand system 5 with Pd:L = 2; all
other conditions are given in Scheme 1. Top: concentration of
product 3 plotted as a function of time with [2]0 = 0.4 M and [1]0
= 0.157 M (blue diamonds); [1]0 = 0.125 M red squares); [1]0
=
0.051 M (green triangles); [1]0 = 0.016 M (purple circles). Inset
left: initial rates vs. [1]0 fit to saturation kinetics with Keq = 10 Mꢀ1
Inset right: power law reaction order n = 0.8 calculated by the
method of Bures (Ref. xx). bottom: concentration of product 3
plotted as a function of time with [1]0 = 0.13 M and [2]0 = 0.3 M
([excess] = 0.17 M, red squares); [2]0 = 0.0.39 M and ([exꢀ
cess=0.26 M, black crosses).
Figure 1. Kinetic profile for the CꢀH Iodination reaction shown in
Scheme 1 carried out using the “same excess” protocol with [exꢀ
cess] = 0.2 M. Both reactions with 0.01 M Pd(OAc)2 and 0.02 M
LꢀBzꢀLeuOH ligand with 3 equiv. each Na2(CO2)3 and CsOAc in
The initial rate data shown in the inset of Figure 2a are fit to
the simple rate expression shown in eq 2. The catalyst concenꢀ
tration is lumped into kkin, and value of the binding constant
Keq = 10 Mꢀ1 suggests that at [1] = 0.1 M, ca. 50% of the cataꢀ
lyst is bound to 1.
o
DMSO/tꢀamyl alcohol at 50 C. a) [1]0 = 0.1 M; [2]0 = 0.3 M. b)
1]0 = 0.1=08 M; [2]0 = 0.28 M.
kkin
1
[ ]
rate =
(2)
1+ Keq
1
[ ]
ACS Paragon Plus Environment

Page 3 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
A rate enhancement due to increased stirring/grinding couꢀ
pled with the observation of overall zero order kinetics is often
attributed to that of a solidꢀsolution mass transfer step is rateꢀ
Reactions carried out using deuterated substrate 1 in global
reactions gave lower rates, confirming a normal kinetic isoꢀ
8
9
determining; however, the observation of first order kinetics in
[Pd] precludes this interpretation, because the catalyst should
not be a factor in the solidꢀtoꢀsolution mass transfer of base
and the low solubility of the inorganic bases means that only a
few percent conversion could be achieved from the quantity of
base held in solution at saturation. A more likely rationalizaꢀ
tion of the stirring/grinding effect on rate is that the decrease
in solid particle size results in an increase in the sparing soluꢀ
bility of the base due to the GibbsꢀThomson effect. Thus rapid
equilibrium in the solidꢀsolution mass transfer step maintains a
constant solution concentration of base that is higher under
aggressive stirring/grinding conditions. The further implicaꢀ
tion of these results is that the base is involved in a kinetically
meaningful step in the catalytic cycle.
tope effect (Figure 3). Fitting the initial rates as a function of
[1ꢀH] and [1ꢀD] to the simple saturation kinetics model used
for the data in Figure 3 revealed that the protioꢀ and deuterated
substrates exhibit the same binding constant and differ in their
elementary rateꢀdetermining step rate constant to give a value
of kH/kD = 1.73.The magnitude of this value supports the kinetꢀ
ic observation that both substrate binding and the subsequent
CꢀH cleavage step contribute to the observed rate.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.4
kH
1.2
=
1.73
kD
1-H
1
0.8
0.6
0.4
0.2
0
1-D
Reactions began to stall typically around 75ꢀ80% converꢀ
sion of 1, and the linear profiles indicative of overall zeroꢀ
order kinetics in [1] and [2] altered at higher conversions as
the reactions slowed. Because the same excess protocol had
indicated a robust catalyst, we probed reasons other than cataꢀ
lyst deactivation or product inhibition for this failure to
achieve complete conversion. We found that addition of furꢀ
ther equivalents of I2 allowed the reaction to progress further.
Figure 5 shows the results of a reaction carried out initially
with equimolar concentrations of [1] and [2], with an additionꢀ
al equivalent of 2 added three further times until complete
conversion of 1 was observed. The reaction cleanly afforded
the product in 96% ee. This suggests that lower yields and
stalling of the reaction is due to consumption of 2 in side reacꢀ
tions, while the catalyst maintains its activity over the entire
course of the reaction. Consumption of I2 was confirmed by
the formation of green crystals in the crude reaction mixture at
the end of the reaction which were identified by single crystal
Xꢀray analysis to be Na4(I3)3I(DMSO)15 (Figure 6). The reꢀ
quirement for multiple equivalents of I2 to achieve high prodꢀ
uct yield in this reaction is rationalized by these results.
0
0.04
0.08
0.12
0.16
substrate concentration (M)
Figure 3. Reaction rate as a function of initial substrate concenꢀ
tration for the reaction in Scheme 1 carried out using cataꢀ
lyst/ligand system 5, 1ꢀH and 1ꢀD. [2]0 = 0.5 M; [Pd] = 0.025 M;
[Pd]/[ligand] = 1:2; 87 mg CsOAc; 50 mg Na2CO3; reaction volꢀ
ume 1.2 ml; T = 50 oC.
Reactions carried out using different concentrations of
Pd(OAc)2 showed that the reaction is first order in [Pd]. The
base employed in these reactions appears to play a complex
role, as the original studies found that a combination of two
bases was most effective. The use of solid base suggests that
mass transfer limitations must be considered. Indeed, we
found that the reaction rate increased with increased stirring
speed. Grinding the Na2CO3 into fine particles prior to use in
the reaction increased the rate further, at which point either
low or high stirring speed gave the same rate (Figure 4). Kiꢀ
netic profiles were linear in all cases. Both rapid stirring and
preꢀgrinding serve to increase the surface area for mass transꢀ
fer between solid base and the reaction solution.
2
1.5
1
0.5
0
Figure 5. Kinetic profile for the CꢀH Iodination reaction shown in
Scheme 1 carried out with initial concentrations [1]0 = [2
] = 0.1
0
600
900
1200
900*
1200*
M. An additional equivalent of 2 (0.1 M) was added three times
as indicated in the figure.
stirring conditions (rpm)
*ground Na2CO3
Figure 4. Reaction rates for the reaction in Scheme 1 carried out
under identical conditions except for stirring speed and preꢀ
grinding of the Na2CO3 base. [1]0 = 0.1 M, [2]0 = 0.3 M, all other
conditions as in Scheme 1.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 10
1
2
3
4
5
6
7
8
Figure 7. Eyring plot for each product enantiomer in the reaction
9
of Scheme 1.
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
These observations appear to support a mechanism where
moderately strong substrateꢀcatalyst binding is followed by
irreversible CꢀH activation and addition of the reaction couꢀ
pling partner, in this case I2, as has been documented in a
number of different Pd(II)ꢀcatalyzed reactions involving these
MPAA ligands. However, several additional features of the
reaction invoke a more complex mechanistic picture. First, the
mildness of the reaction conditions is in contrast with other
cases where significantly higher temperatures are required to
undergo CꢀH activation. Indeed, we found that much harsher
conditions were required in our synthetic Pdꢀcatalyzed reacꢀ
tions to produce the deuterated substrate for the KIE experiꢀ
ments.
Second, we found that we remain unable to detect substrate
binding commensurate with the binding constant predicted by
Figure 6. Unit cell of crystals formed in the crude reaction mixꢀ
ture at the end of the reaction in Scheme 1 with unit cell formula
Na4(I3)3I (DMSO)15
the reaction kinetics using catalyst 5 formed with the oꢀ
difluoro benzoyl ligand system. The kinetic fit of eq 2 predicts
that 1 should ultimately partition to form Pdꢀ1 in an amount
equal to the total [Pd] during reaction. However, a mass balꢀ
ance showed that recovered starting material was nearly quanꢀ
titative, indicating much weaker binding between substrate 1
and catalyst 5 than is predicted from the reaction kinetics.
Further, we found that recovered 1ꢀD in interaction with 5
showed no isotopic scrambling in experiments mixing deuterꢀ
ated 1ꢀD and protic solvent, just as was found for the benzoyl
ligand. These results confirm that neither reversible nor irreꢀ
versible CꢀH cleavage occurs with catalyst 5 prior to addition
of iodine. Intriguingly, given this apparent inertness of the
substrate to CꢀH bond activation, the kinetic data clearly show
that reaction proceeds smoothly immediately upon addition of
I2 with either the original PdꢀLeuꢀBzꢀOH or catalyst 5. This
observed requirement for the presence of I2 is all the more
puzzling given that palladacycle 4 forms readily via CꢀH actiꢀ
vation of the benzoyl protecting group of the ligand in the
absence of I2.
.
Activation parameters were determined by measuring reacꢀ
tion rate over a range of temperature from 25 oC to 72 oC. (Taꢀ
ble 1, Figure 7). Remarkably, product ee was maintained
above 96% ee over a range of nearly 50 oC, decreasing only to
just below 95% ee at 84 oC. The reaction in the absence of
ligand is tenꢀfold slower at 40 oC and exhibits a higher activaꢀ
tion enthalpy and a lower activation entropy. Similar rates,
activation parameters, and enantioselectivities were obtained
in reactions using substituted benzoyl protecting groups OMeꢀ
BzꢀLeuꢀOH and CF3ꢀBzꢀOH, indicating the lack of a strong
electronic influence on the course of the reaction.
Table 1. Activation Parameters for the Reaction of Scheme
1.
parameter
No ligand
25.2
BzꢀLeuꢀOH
16.8
Further studies thus focused on probing the role of I2 in trigꢀ
gering CꢀH activation in this system. The conventional mechꢀ
anistic hypothesis for similar reactions obeying similar kinetꢀ
ics6 involves substrate 1 binding to the Pd(II) center preceding
addition of substrate 2, an alternative hypothesis may be inꢀ
voked that exhibits the same kinetic features. Positive order
kinetics in 1 and zero order kinetics in 2 could result if 2 enꢀ
ters the cycle prior to 1 and forms the saturated catalyst resting
state. Oxidative addition of iodine to the catalyst to form the
species Pd(IV)ꢀ(2), occupying all of the Pd to form a comꢀ
pletely saturated intermediate, would satisfy the observed zero
order kinetics in [2]. Substrate 1 binding reversibly to Pd(IV)ꢀ
(2) followed by rateꢀdetermining CꢀH activation gives the
observed saturation kinetics in [1] and normal deuterium isoꢀ
tope effect. If the CꢀH activation step proceeds via such a
Pd(IV) species, with a role for I2 in oxidizing Pd(II) to Pd(IV),
a rationalization for the lack of CꢀH activation in the absence
of iodine is provided.
ꢀH‡ (kcal/mol)
ꢀS‡ (cal/mol/K)
3.6
ꢀ18
-9.5
-11
R-enantiomer
R² = 0.998
-12.5
-14
-15.05.0028
0.003
0.0032
0.0034
S-enantiomer
R² = 0.920
Although reaction mechanisms proceeding through high oxꢀ
idation state Pd(IV) have been proposed, and organopalladium
Pd(IV) complexes have been isolated,8ꢀ10 their reactivity is less
well known than Pd(0) and Pd(II). In particular, oxidation of
-17
-18.5
1/T (1/K)
ACS Paragon Plus Environment

Page 5 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
Pd is typically effected via hypervalent iodine reagents. Deꢀ
tailed studies of stoichiometric CꢀH activation via Pd(IV)
lines). These peaks are unchanged when substrate 1 is added
to the solution, supporting other evidence that the substrate
does not interact strongly with the catalyst. However, when
iodine is mixed with Pd and ligand, the peaks attributed to
interaction between Pd and the ligand show a significant shift
that is also observed under reaction conditions with both subꢀ
strate 1 and I2 (Figure 9, red dashed lines). These spectroscopꢀ
ic results demonstrate a change in the catalyst upon interaction
with I2 that is consistent with the proposal of oxidation to
Pd(IV) from the electrochemical studies.
8
9
complexes were carried out in the paraꢀselective CꢀH arylation
of monosubstituted arenes, where an [ArPd(IV)F] intermediate
was invoked.8c The proposal of a Pd(IV) species in the current
mechanistic scenario relies on the oxidizing capability of I2.
Cyclic voltammetry experiments carried out using catalyst 5
and increasing amounts of I2 (Figure 8) reveal an irreversible
oxidation peak at 0.95 V (all potentials are given versus SCE)
that corresponds to the oxidation of I2 to I+ and surfaceꢀ
adsorbed I.11 A peak at 1.4 V corresponding to Pd(II)/Pd(IV)
oxidation shows decreasing concentration of Pd(II) as the conꢀ
centration of I2 increases. These experiments suggest that the
2,6ꢀF,FꢀBzꢀLeuꢀOH ligand provides for sufficient overlap in
the 5 (Pd(II/IV)) and I2 oxidation potentials to enable oxidative
addition of I2 directly to 5. A more detailed description of the
electrochemical results is provided in the Supporting Inforꢀ
mation and suggests the oxidative addition of I2 to 5 is reversiꢀ
ble and provides an equilibrium of Pd(II) and Pd(IV) that,
under reaction concentrations of I2, is shifted strongly towards
Pd(IV).. Because substrate 1 is not involved in the Pd(IV)ꢀ
forming process, this system appears to be different from those
of previous studies where Pd(II)ꢀmediated CꢀH cleavage formꢀ
ing a Pd(II) aryl or alkyl species precedes Pd oxidation and a
second CꢀH cleavage by Pd(IV).
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
Further support for the requirement of Pd(IV) in such a
mechanism comes from the inertness of a Pd(II)I2 complex to
reaction with amine 1, as shown in Scheme 3. Although PdI2 is
a viable precatalyst in the reaction between 1 when I2 is preꢀ
sent, no product is formed under conditions of
(Pd(II)I2+1+bases) prior to addition of molecular I2.
Scheme 3. Reaction of PdI2 and amine 1 in the absence and
presence of molecular I2.
2, 6-F,F-Bz-Leu-OH (1 equiv)
Na2CO3, CsOAc (3 equiv)
PdI2
+
1
no product
DMSO, t-AmylOH (50 oC)
0.5 equiv
0.125 M
I2 (3 equiv)
product 3
reaction
conditions
Pd + ligand
+ I2
Pd + ligand
+ substrate
Pd + ligand
ligand
-112.7 -112.9 -113.1 -113.3 -113.5 -113.7 -113.9 -114.1 -114.3
f1 (ppm)
Figure 9. 19FꢀNMR studies of the role of iodine. a) free 2,6ꢀ
difluorobenzoyl ligand; b) ligand +Pd; c) ligand + Pd + substrate
1; d) ligand + Pd + I2; e) system with both substrate 1 and iodine 2
under reaction turnover. All spectra taken with Pd:L = 2:1 after
heating to 50 oC in d6ꢀDMSO/tꢀAmylꢀOH with CsOAc and
Na2(CO3)2.
Figure 8. Cyclic voltammetry with variable concentration of I2
alone (top) and with 1 mM Pd(OAc)2 and 2.5 mM 2,6ꢀF,FꢀBzꢀ
LeuꢀOH The peak at 0.95 V corresponds to oxidation of I2. The
disappearance of a peak at 1.4 V corresponds to Pd(II)/Pd(IV)
oxidation.
A reaction mechanism accounting for all of these experimental
observations is proposed in Scheme 4. Conversion of Pd(II) to
Pd(IV) via oxidative addition of iodine to produce intermediꢀ
ate A precedes reversible substrate addition. Irreversible CꢀH
activation of intermediate C is followed by transfer of iodine
to the palladacycle intermediate D and reductive elimination to
regenerate catalyst 5. The observed saturation kinetics in [1]
along with the magnitude of the observed H/D KIE suggests
that the resting state shifts between species A, and C as a funcꢀ
tion of substrate concentration [1]. The cycle may be written
as the elementary steps given in eqs 3ꢀ5, with the kinetically
19F NMR studies using catalyst 5 shed further light on this
mechanistic proposal as shown in Figure 9. The leucine
MPAA ligand containing the 2,6ꢀdifluorobenzoyl protecting
group gives a strong signal at ꢀ112.7 ppm (referenced to triꢀ
fluorotoluene). Mixing the ligand with Pd(OAc)2 produces a
pair of peaks at ꢀ113.4 and 114.1 ppm (Figure 9, blue dashed
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 10
1
2
3
4
5
6
7
8
9
meaningful steps above the dashed line. Intermediates D and E
are kinetically indistinguishable from B, since these species
involve only internal transformations of the catalytic complexꢀ
dative addition product [MPAA]Pd(IV)I2, (III). It is exergonic
by 10.0 kcal/mol, and is driven further by another 6.8 kcal/mol
via extraction of an iodide from Pd(IV) by I2 to form sodium
triiodide, IV. Thus, overall the oxidative addition of I2 to the
Pd(II)ꢀcenter of the [MPAA]/Pd(II) catalyst and subsequent
triiodide formation, i.e. I → IV, is 16.8 kcal/mol exergonic.
This result is in accordance with the experimental formation of
Na4(I3)3I (DMSO)15 crystals and supports the observation that
excess I2 is required even for a reaction exhibiting zero order
kinetics in [I2].
es.
Keq →
B krds →3 + 5
A +1
B
←
(3)
(4)
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
--------------
(5)
5 + 2 →A
As noted above, this mechanism
differs from previous proposals in which the substrate is inꢀ
volved in two CꢀH cleavage events, where Pd(IV) palladacyꢀ
cle formation is preceded by aryl or alkyl CꢀH cleavage by
Pd(II). In the present case, experimental evidence suggests that
the Pd(IV) species is formed without substrate involvement.
Such a species represents a different class of precatalyst for Cꢀ
H functionalization reactions compared to the “double CꢀH
cleavage” systems in which Pd(IV) has previously been impliꢀ
cated in the second step.
Scheme 4. Proposed Reaction Mechanism.
Figure 10. Important structural parameters of the model catalyst
I, oxidative addition transition state TS-OA and oxidative addiꢀ
tion product [MPAA]Pd(IV)I2, III, of the [MPAA]Pd(II) catalyst
insertion into the IꢀI bond of I2. Bond distances are in Å.
The next step of the reaction is coordination of the substrate
and expulsion of MeOH to form intermediate V. This process
is exergonic by 5.1 kcal/mol. The finding that substrate coorꢀ
dination is favorable is generally consistent with the experiꢀ
mental data presented above. Substrate C–H bond activation in
complex V, required for CꢀH iodination, can occur through the
following concertedꢀmetallationꢀdeprotonation (CMD) pathꢀ
ways13ꢀ15 a) “internalꢀligandꢀassisted”, where the protecting
group of the MPAA ligand is the base, or b) “internalꢀ
carbonateꢀassisted”, where the coordinated carbonate is the
base. For Pd(II) reactions, the “internalꢀligandꢀassisted” mechꢀ
anism is widely studied and accepted,16ꢀ20 but this pathway has
not been investigated for Pd(IV) species.
We investigated both CMD pathways to explore the role of
the MPAA ligand in C–H activation on Pd(IV) systems (see
Figure 11). Calculations show that the “internalꢀligandꢀ
assisted” C–H activation occurs through the TS-CH-L transiꢀ
tion state with a free energy barrier of 26.4 kcal/mol, and the
“internalꢀcarbonateꢀassisted” C–H activation occurs through
the transition state TS-CH-C with a free energy barrier of 23.3
kcal/mol (both calculated relative to intermediate V), demonꢀ
strating that the “internalꢀcarbonateꢀassisted” C–H activation
is kinetically favored. In addition, the “internalꢀcarbonateꢀ
assisted” pathway is exergonic (V → VII-C, ꢀG = –4.7
kcal/mol), whereas the “internalꢀligandꢀassisted” pathway is
To probe this mechanistic hypothesis further, DFT calculaꢀ
tions were carried out using LꢀLeuꢀBzꢀOH as the MPAA ligꢀ
and and a model Pd catalyst I (Figure 10) derived from experꢀ
imental findings, comprised of the diꢀanionic MPAA ligand,
one sodium carbonate molecule (required for the reaction) and
one methanol molecule (model for tꢀAmyl alcohol).12
The proposed first step of the catalytic cycle is oxidative
addition of I2 to the Pd(II)ꢀcenter of the catalyst. As seen in
Figure 11, this step of the reaction requires 16.9 kcal/mol free
energy barrier at transition state TS-OA and results in the oxiꢀ
ACS Paragon Plus Environment

Page 7 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
endergonic (V → VII-L, ꢀG = 4.9 kcal/mol). Thus the “interꢀ
nalꢀcarbonateꢀassisted” pathway is kinetically and thermodyꢀ
namically preferred for C–H activation by Pd(IV)/MPAA sysꢀ
tems, and the MPAA ligand does not directly participate in the
rateꢀlimiting C–H activation transition state as observed for
the Pd(II)/MPAA systems. The lack of a Hammett effect of
the ligand benzoyl group is in accordance with this finding.
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 11. Calculated free energy surface for the full catalytic cycle for C–H iodination by [MPAA]/Pd(II) catalyst with I2. The “internalꢀ
ligandꢀassisted” pathway is shown in blue and the “internalꢀcarbonateꢀassisted” pathway is shown in black. Energies are reported as
ꢀ
G/ꢀH in kcal/mol.
Close examination of the transition state TS-CH-C shows
that the “internalꢀcarbonateꢀassisted” C–H activation occurs
trans to the apical iodide ligand (see Figure 12), and isomeriꢀ
zation is required to position the substrate aryl group and ioꢀ
dide in close vicinity for reductive elimination. This isomeriꢀ
zation process, i.e. VII-L → VIII, requires only 5.0 kcal/mol
free energy. In the resulting intermediate VIII, the bicarbonate
ligand forms a hydrogen bonding interaction with the carbonyl
of the MPAA benzoyl protecting group.
Figure 12. Important structural parameters for the
[MPAA]/Pd(IV)ꢀsubstrate intermediate , and “internalꢀligandꢀ
assisted” and “internalꢀcarbonateꢀassisted” CꢀH activation transiꢀ
tion states TS-CH-L and TS-CH-C, respectively. Bond distances
are in Å.
V
The C–I reductive elimination occurs through the transition
state TS-RE with a free energy barrier of 20.1 kcal/mol (relaꢀ
tive to VII-C). Formation of the reductive elimination product,
IX, is exergonic by 11.8 kcal/mol relative to intermediate VII-
C. Careful analysis of the geometries reveals that the bicarꢀ
bonate ligand plays an important role in the reductive eliminaꢀ
tion step. First, the ligand dissociates from the Pd(IV) center,
as evidenced by lengthening of the Pd–O distance upon going
from VIII (Pd–O = 2.11 Å) to TS-RE (Pd–O = 2.29 Å). Disꢀ
sociation of a ligand prior to reductive elimination on Pd(IV)
has been described previously.21,22 Next, bicarbonate particiꢀ
pates in the transition state TS-RE to facilitate displacement
of the aryl ligand as the C–I bond forms. As such, the Pdꢀ
carbonate bond is fully formed (Pd–O = 2.07 Å) in IX, as conꢀ
firmed by IRC from TS-RE (see Figure 13). Subsequent
deprotonation and product (XI) dissociation steps to regenerꢀ
ate catalyst I are exergonic and the calculated reaction free
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 10
1
2
3
4
5
6
7
8
CONCLUSIONS
energy for the entire reaction, i.e., I + 10 + 2 I2 + Na2CO3 → I
+ 3 + NaI3 + NaHCO3, is ꢀG = –42.7 kcal/mol.
An exhaustive mechanistic study of the enantioselective
desymmetrization of diarylmethylamines using catalyzed by
Pd with MPAA ligands sheds new light on how the concept of
weak coordination in CꢀH functionalization affords a rich varꢀ
iation in reaction pathways available to these systems. The
mild reaction conditions in the present work steer CꢀH the
activation step to a Pd(IV) species formed from oxidative adꢀ
dition of iodine to the Pd(II) catalyst. The reaction remains
highly enantioselective over a wide temperature range and
strikingly insensitive to the electronic characteristics of the
ligand, and computational provide insight into these findings.
This work may lead to a better general understanding of the
subtle variations in reaction mechanism for CꢀH functionalizaꢀ
tion reactions that may be extant for this ligand class dependꢀ
ing on substrate, amino acid ligand and protecting group, and
reaction conditions.
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 13. Important structural parameters for the C–I reductive
elimination transition state TS-RE and product IX. Bond distancꢀ
es are in Å.
The observed high enantioselectivity observed in this reacꢀ
tion is supported by transition state calculations comparing
TS-CH-C leading to the R vs. the S product. As shown in
Figure 14, the calculated energy difference between these
transition states produces the correct stereoisomer, but it is
much larger than that expected based on the experimental ee
value. This overestimation of the enantioselectivity suggests
the possibility of additional complexity in the mechanism of
asymmetric induction with the likelihood of several diastereꢀ
omeric transition states that contribute to the observed enantiꢀ
oselectivity. Further studies of enantioinduction including the
relative insensitivity to temperature are ongoing.
ASSOCIATED CONTENT
Supporting Information includes: Details of kinetic and
mechanistic experiments. Computational details. Optimized
structures along the full catalytic cycle for C–H iodination by
[MPAA]Pd(II) catalyst with I2 (Figure S1); Calculated enerꢀ
gies for coordination of the substrate to model Pd(II) and
Pd(IV) complexes on the reaction pathway (Figure S2); Comꢀ
puted energies (in hartree) for dimerization and ligand exꢀ
change (Table S1), and Cartesian (xyz) coordinates of all reꢀ
ported structures. These materials are available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
Blackmond@scripps.edu
NOTES
The authors declare no competing financial interests.
Figure 14. Transition state energy (ꢀG/ꢀH in kcal per mol) of the
CꢀH cleavage step for the minor enantiomer compared to the maꢀ
jor enantiomer (see Figure 11).
ACKNOWLEDGMENT
This work was supported by the National Science Foundation
under the CCI Center for Selective C–H Functionalization
(CHEꢀ1205646). D.G.M. and B.E.H. gratefully acknowledge
NSF MRIꢀR2 grant (CHEꢀ0958205 for D.G.M.) and the use of
the resources of the Cherry Emerson Center for Scientific
Computation. Helpful advice on the NMR spectroscopy
from D. Huang and L. Paternack (TSRI) is acknowledged.
Curtis Moore and Arnold Rheinhold (UCSD) and Oana
Luca (TSRI) are acknowledged for crystallography studies.
These computational results are fully consistent with the exꢀ
perimental findings and indicate that I2 oxidative addition to
Pd(II) precedes rateꢀlimiting C–H activation at Pd(IV). Strikꢀ
ingly, this proposed mechanism differs from the previously
reported computational findings on the Pd(II)ꢀcatalyzed ioꢀ
dination by I2 both in systems with23 and without24 MPAA
ligands. It is important to note, however, that the substrates
involved in the experimental and computational studies of
iodination without MPAA ligands bear greater similarity to
these ligands, both of which contain amide groups, than they
do to the diarylmethylamine substrate 1 used in our work,
which lacks the C=O functionality. Our finding of facile CꢀH
functionalization to form palladacycle 4 between Pd(II) and
the benzoyl group, coupled with the inertness of substrate 1 to
this reaction with Pd(II) under mild conditions, suggests that
the amide group plays an important role in the CꢀH activation
process mediated by Pd(II).
REFERENCES
1
2
Engle. K.E.; Mei, T.S.; Wasa, M.; Yu, J.ꢀQ.; Acc. Chem. Res.
2012, 45, 788ꢀ802.
Chu, L.; Wang, Y.ꢀC.; Moore, C.E.; Rheingold, A.L.; Yu, J.ꢀQ. J.
Am. Chem. Soc. 2013, 135, 16344ꢀ16347.
3
4
Chu. L.; Xiao, K.ꢀJ.; Yu, J.ꢀQ.; Science, 2014, 346, 451ꢀ455.
Wang, X.ꢀC.; Hu, Y.; Bonacorsi, S.; Hong, Y.; Burrell, R.; Yu, J.ꢀ
Q.; J. Am. Chem. Soc., 2013, 135, 10326ꢀ10329.
5
a) Blackmond, D.G.; Angew. Chemie Int. Ed. 2005, 44, 4032; b)
Mathew, J.S.; Klussmann, M.; Iwamura, H.; Valera, F.; Futran,
ACS Paragon Plus Environment

Page 9 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
A.; Emanuelsson, E.A.C.; Blackmond, D.G. J. Org. Chem. 2006,
71, 4711.
7
8
9
6
Baxter, R.D.; Sale, D.; Engle, K. M.; Yu, J.ꢀQ.; Blackmond, D.G.
J. Am.Chem. Soc. 2012, 134, 4600.
Bures, J.; Angew. Chemie Int. Ed., 2016, 55, 16084.
Xu, LꢀM.; Li, B.ꢀJ.; Yang, Z.; Shi, Z.ꢀJ., Chem. Soc. Rev. 2016,
39, 712ꢀ733.
7
8
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
9
a) Hull, K.L.; Lanni, E.L.; Sanford, M.S. J. Am. Chem. Soc.,
2006, 128, 14047ꢀ14049; b) Dick, A.R.; Kampf, J.; Sanford, M.S.
J. Am. Chem. Soc. 2007, 129, 15142; c) Racowksi, J.M.; Ball.,
N.D.; Sanford, M.S. J. Am. Chem. Soc., 2011, 133, 18022ꢀ18025
10 a) Rosewall, C.F.; Sibbald, P.A.; Liskin, D.V.; Michael, F.E. J.
Am. Chem. Soc. 2009, 131, 9488ꢀ9489; b) Sibbald, P.A.; Roseꢀ
wall, C.F.; Swartz, R.D.; Michael, F.E. J. Am. Chem. Soc. 2009,
131, 15945ꢀ15951.
11 Geissler, W.; Nitzsche, R.; Landsberg, R. Electrochim. Acta,
1966, 11, 389ꢀ400.
12 DFT calculations were performed at the [B3LYPꢀD3/6ꢀ
311+g(2d,p) and SDD for Pd, I]//[B3LYPꢀD3/6ꢀ31g(d,p) and
Lanl2dz for Pd, I] level of theory. Bulk solvent effects
were included in all calculations with an implicit solvation modꢀ
el (IEFꢀPCM) for methanol. See SI for full computational details.
13 Biswas, B.; Sugimoto, M.; Sakaki, S. Organometallics 2000, 19,
3895ꢀ3908.
14 Davies, D. L.; Donald, S. M. A.; Macgregor, S. A. J. Am. Chem.
Soc. 2005, 127, 13754ꢀ13755.
15 Gorelsky, S. I.; Lapointe, D.; Fagnou, K. J. Am. Chem. Soc. 2008,
130, 10848ꢀ10849.
16 Cheng, G. J.; Chen, P.; Sun, T. Y.; Zhang, X. H.; Yu, J. Q.; Wu,
Y. D. Chem. Eur. J. 2015, 21, 11180ꢀ11188.
17 Cheng, G. J.; Yang, Y. F.; Liu, P.; Chen, P.; Sun, T. Y.; Li, G.;
Zhang, X. H.; Houk, K. N.; Yu, J. Q.; Wu, Y. D. J. Am. Chem.
Soc. 2014, 136, 894ꢀ897.
18 Haines, B. E.; Musaev, D. G. ACS Catal. 2015, 5, 830ꢀ840.
19 Musaev, D. G.; Figg, T. M.; Kaledin, A. L. Chem. Soc. Rev. 2014,
43, 5009ꢀ5031.
20 Engle, K. M. Pure Appl. Chem. 2016, 88, 119ꢀ138.
21 Camasso, N. M.; PerezꢀTemprano, M. H.; Sanford, M. S. J. Am.
Chem. Soc. 2014, 136, 12771ꢀ12775.
22 Furuya, T.; Benitez, D.; Tkatchouk, E.; Strom, A. E.; Tang, P. P.;
Goddard, W. A.; Ritter, T. J. Am. Chem. Soc. 2010, 132, 3793ꢀ
3807.
23 Zhou, M. J.; Yang, T. L.; Dang, L. J. Org. Chem. 2016, 81, 1006ꢀ
1020.
24 Haines, B. E.; Xu, H. Y.; Verma, P.; Wang, X.; Yu, J. Q.;
Musaev, D. G. J. Am. Chem. Soc. 2015, 137, 9022ꢀ9031.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 10
1
2
3
4
5
6
7
8
9
Formatted: Justified
Insert Table of Contents artwork here
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
Pd
+ ligand
+ I2
Pd
+ ligand
+ substrate
Pd
+ ligand
-112.7 -112.9 -113.1 -113.3 -113.5 -113.7 -113.9 -114.1 -114.3
f1 (ppm)
ACS Paragon Plus Environment