Reaction Intermediates Kinetics in Solution Investigated by Electrospray Ionization Mass Spectrometry: Diaurated Complexes

Article abstract of DOI:10.1021/jacs.5b08744

A new method to investigate the reaction kinetics of intermediates in solution by electrospray ionization mass spectrometry is presented. The method, referred to as delayed reactant labeling, allows investigation of a reaction mixture containing isotopically labeled and unlabeled reactants with different reaction times. It is shown that we can extract rate constants for the degradation of reaction intermediates and investigate the effects of various reaction conditions on their half-life. This method directly addresses the problem of the relevance of detected gaseous ions toward the investigated reaction solution. It is demonstrated for geminally diaurated intermediates formed in the gold mediated addition of methanol to alkynes. Delayed reactant labeling allows us to directly link the kinetics of the diaurated intermediates with the overall reaction kinetics determined by NMR spectroscopy. It is shown that the kinetics of protodeauration of these intermediates mirrors the kinetics of the addition of methanol demonstrating they are directly involved in the catalytic cycle. Formation as well as decomposition of diaurated intermediates can be drastically slowed down by employing bulky ancillary ligands at the gold catalyst; the catalytic cycle then proceeds via monoaurated intermediates. The reaction is investigated for 1-phenylpropyne (Ph-CC-CH₃) using [AuCl(PPh₃)]/AgSbF₆ and [AuCl(IPr)]/AgSbF₆ as model catalysts. Delayed reactant labeling is achieved by using a combination of CH₃OH and CD₃OH or Ph-CC-CH₃ and Ph-CC-CD₃.

Full text of DOI:10.1021/jacs.5b08744

Subscriber access provided by NEW YORK MED COLL
Article
Reaction Intermediates Kinetics in Solution Investigated by
Electrospray Ionization Mass Spectrometry: Diaurated Complexes
Lucie Jasikova, Mariarosa Anania, Simona Hybelbauerova, and Jana Roithova
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b08744 • Publication Date (Web): 02 Oct 2015
Downloaded from http://pubs.acs.org on October 11, 2015
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 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Reaction Intermediates Kinetics in Solution
Investigated by Electrospray Ionization Mass
Spectrometry: Diaurated Complexes
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
Lucie Jašíková, Mariarosa Anania, Simona Hybelbauerová, and Jana Roithová*
Department of Organic Chemistry, Faculty of Science, Charles University in Prague, Hlavova 2030/8, 12843
Prague 2 (Czech Republic)
ABSTRACT: A new method to investigate the reaction kinetics of intermediates in solution by electrospray
ionization mass spectrometry is presented. The method, referred to as Delayed Reactant Labeling, allows
investigation of a reaction mixture containing isotopically labeled and unlabeled reactants with different reaction
times. It is shown that we can extract rate constants for the degradation of reaction intermediates and investigate
the effects of various reaction conditions on their halfꢀlife. This method directly addresses the problem of the
relevance of detected gaseous ions towards the investigated reaction solution. It is demonstrated for geminally
diaurated intermediates formed in the gold mediated addition of methanol to alkynes. Delayed Reactant Labeling
allows us to directly link the kinetics of the diaurated intermediates with the overall reaction kinetics determined
by NMR spectroscopy. It is shown that the kinetics of protodeauration of these intermediates mirrors the kinetics
of the addition of methanol demonstrating they are directly involved in the catalytic cycle. Formation as well as
decomposition of diaurated intermediates can be drastically slowed down by employing bulky ancillary ligands
at the gold catalyst; the catalytic cycle then proceeds via monoaurated intermediates. The reaction is investigated
for 1ꢀphenylpropyne (PhꢀCCꢀCH3) using [AuCl(PPh3)]/AgSbF6 and [AuCl(IPr)]/AgSbF6 as model catalysts.
Delayed Reactant Labeling is achieved by using a combination of CH3OH and CD3OH or PhꢀCCꢀCH3 and Phꢀ
CCꢀCD3.
Introduction
-
Mass spectrometry (MS) is an important tool for the investigation of reaction mechanisms.1 5 The
introduction of electrospray ionization (ESI) allowed direct probing of reaction mixtures and ESIꢀMS was
established as a method for “fishing out” reaction intermediates.6,7 A big advantage of using ESIꢀMS is its
sensitivity. Reactive intermediates that are often present at very low concentrations can be detected, isolated in
the gas phase and can have their reactivity probed.8ꢀ10 The structure of the intermediates can be interrogated in
collisional experiments or with ion spectroscopy approaches.11,12
The limitations of ESIꢀMS always have to be considered. For example, the relative intensities of the peaks
in a mass spectrum do not need to correlate with the concentrations in the sprayed solution.13 Indeed the same
reaction mixture can provide completely different ESIꢀMS spectra if the ionization conditions are changed.14ꢀ16
Given common instabilities in the ionization conditions and an unknown response to the concentration of the
ionized species, the use of ESIꢀMS to monitor reaction kinetics is often associated with a large error or even not
applicable. This obstacle can be somewhat circumvented by incorporating a permanently charged group into a
reactant and assuming that the transfer of all ions containing this charged group to the gas phase will be similarly
efficient.17,18 This approach can be rather synthetically demanding and cannot be generally used; for example if
the reaction intermediates are naturally cationic or anionic.
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 19
1
2
3
4
5
6
7
8
9
Another complicating feature of ESIꢀMS is possible artifacts from the electrospray process. Electrospray
ionization is associated with increasing the concentration of the sprayed solution by several orders of magnitude,
so exotic species can be formed although they did not exist in the original solution. The concurrent method for
the investigation of reaction mechanisms and reaction intermediates is nuclear magnetic resonance (NMR)
spectroscopy. This method is largely complementary to ESIꢀMS. Direct examination of a reaction mixture
1
assures the sole detection of relevant species and the most commonly applied H NMR has a linear response to
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 concentration of the given species in solution.19 Nevertheless, all species are investigated at once, so reactive
intermediates that are present in negligible amounts are usually not detected.
Here, we present a new method for the investigation of reaction intermediates using ESIꢀMS. This method
is based on monitoring a reaction mixture containing one of the reactants as a mixture of isotopically labeled and
unlabeled molecules. Either the labeled or unlabeled molecules are added to the reaction mixture with a time
delay. As detailed below, this trick provides a direct link to the condensed phase and helps eliminate artifacts
from the electrospray ionization process. This opens the possibility of investigating the kinetics of all possible
complexes containing this particular reactant during a given reaction.
The new method will be applied to monitor the kinetics of diaurated intermediates20 in the gold(I) mediated
addition of methanol to alkynes (Scheme 1).21ꢀ26 We have chosen 1ꢀphenylpropyne as a model alkyne and
[AuCl(PPh3)] and [AuCl(IPr)] (IPr = 1,3ꢀbis(2,6ꢀdiisopropylphenyl)imidazolꢀ2ꢀylidene) as representative
catalysts. It has been shown that with a gold catalyst bearing a small ligand such as trimethylphosphine or
triphenylphospine the reaction proceeds via the formation of geminally diaurated complexes, which are
subsequently protodeaurated to form the product of the reaction.13,27 Using gold catalysts with bulky ligands,
such as the N-heterocyclic carbene ligand IPr, leads to a suppression of the formation of diaurated intermediates
and renders the reaction faster. If the geminally diaurated complexes are formed with the bulky ligands, then it
was shown that they do not undergo subsequent protodeauration and remain as byproducts of the reaction.27
It was suggested that the geminally diaurated complexes are formed by the trapping of neutral monoaurated
28ꢀ33
intermediates C by another [(L)Au]+ cation (Scheme 1).
The trapping reaction C → E has been suggested to
be reversible for catalysts with small phosphine ligands, or irreversible for catalysts bearing bulky ligands.27 We
have previously suggested that the diaurated complexes are not formed by an offꢀcycle trapping reaction, but
instead are the result of the primary formation of intermediate D, which undergoes fast rearrangement to the
geminally diaurated intermediate E.13 Originally, we proposed [(L)Au(OCH3)] addition to the goldꢀactivated
alkyne, but the exchange of the hydrogen atom of the methanol molecule can also proceed at intermediate B as
depicted.
The presented “delayed reactant labeling” method in ESIꢀMS enables us to study various effects such as the
acidity of the reaction mixture or silver cation concentration on the halfꢀlife of the diaurated intermediates and
relate it to the reaction kinetics obtained by NMR spectroscopy.
2
ACS Paragon Plus Environment

Page 3 of 19
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
Scheme 1. Combination of suggested reaction mechanisms for gold(I) catalyzed addition of methanol
to an alkyne.13,26-33 Complexes in the frames can be detected by ESI-MS.13 The symbol (LAu+)* is used
for an active form of the catalyst.
Experimental and Computational Details
Reaction mixtures were prepared by mixing a filtered solution of AgSbF6 and [AuCl(PPh3)] (or
[AuCl(IPr)], where IPr = 1,3ꢀbis(2,6ꢀdiisopropylphenyl)imidazolꢀ2ꢀylidene) in methanol (either CH3OH or
CD3OH(D)) with 1ꢀphenylpropyne (or D3ꢀ1ꢀphenylpropyne) so that a 0.24 M solution of (D3ꢀ)1ꢀphenylpropyne
with 2.5 mol% of the [AuCl(PPh3)] (or [AuCl(IPr)]) was formed.
The NMR experiments were recorded using a Bruker AVANCE III (600 MHz) and the δ scale was
referenced to the solvent residual peak at δ = 3.31 ppm. The solutions of a catalyst and reactants were mixed and
1
immediately probed by the NMR instrument. The representative H NMR spectrum with peak assignments is
shown in the Supporting Information, Figure S1.
34,35
Mass spectrometry experiments were performed on a TSQ 7000 mass spectrometer
with a
quadrupoleꢀoctopoleꢀquadrupole configuration. The ions were generated by electrospray ionization (ESI) at soft
ionization conditions and the mass spectra were recorded by scanning the first quadrupole (Figures S3ꢀS5 in the
Supporting Information). For the mass spectrometry experiments, the reaction mixtures were diluted 10ꢀtimes
(hence 0.024 M solution of 1ꢀphenylpropyne was used). The exact compositions of all investigated reaction
mixtures are listed in the Supporting Information.
36
37
DFT calculations were performed using the B3LYP density functional and D3 empirical correction
for the dispersion interactions as implemented in the Gaussian program package.38 For the geometry
optimization, the combination of the ccꢀpVTZ basis set for C, H, O, N and P and the LanL2DZ basis set for Au
was used. Frequency calculations were performed in order to control the nature of the stationary points and
obtain thermochemical corrections for enthalpies and free energies at 0 K and 298 K. The solvent effect of
methanol was approximated by singleꢀpoint calculations using the polarized continuum model.39 All optimized
structures and their energies are listed in Table S10 in the Supporting Information. We have also studied the
basis set effect at the gold atom. Results of singleꢀpoint calculations with the LanL2TZ for Au and ccꢀpVTZ for
the remaining atoms show that the effect is negligible (Table S8). The effect of the dispersion interactions was
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 19
1
2
3
4
5
tested by B3LYP/ccꢀpVTZ(LanL2TZ for Au) singleꢀpoint calculations without D3 empirical correction. The
pathways leading via neutral monoaurated intermediates become endothermic (Table S9).
The energetics of protonation/deprotonation steps are strongly influenced by the method used for
approximation of the proton enthalpy. We have tried to apply the published value for the proton enthalpy in
methanol (ꢀ1054 kJ molꢀ1),40 but it renders all reactions where the proton is eliminated about 50 kJ molꢀ1 more
endothermic (see Table S7 in the Supporting Information). Such a potential energy surface would basically
exclude the reaction pathways leading via neutral monoaurated intermediates, therefore we did not apply it here.
Instead, proton was considered in the form of protonꢀbound dimer of methanol. Under this assumption,
formation of a neutral monoaurated intermediate becomes exothermic.
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
Kinetic model.
Scheme 2.
Our kinetic model is derived from the very simple model shown in Scheme 2. We assume a firstꢀorder
reaction between a catalyst (Cat) and a reactant (React) yielding an intermediate (Int). The intermediate then
transforms to a product (Prod). This model can be straightforwardly improved by taking into account
particularities of a given studied system, but to demonstrate our method it will be used in this simplest version.
Such kinetic schemes allow us to express the time evolution of the intermediate concentrations as well as the
concentrations of the reactants and products. It can be easily applied for NMR kinetic experiments, in which we
follow concentration changes of reactants, products, and sometimes also intermediates.
The ESIꢀMS method suffers from a nonlinear response to the concentration of species in the solution as
outlined in the Introduction. The concentration changes have to be therefore approached indirectly. We have
designed a method with delayed labeling of the reaction mixture by an isotopically labeled reactant while the
mixture also contains the corresponding unlabeled reactant. The corresponding mass spectrum shows pairs of
signals of ions containing the reactants. We assume that the ionization efficiencies are not influenced by the
isotopic labeling. The key trick is the time delay in the labeling. The mutual time evolution of the unlabeled and
labeled signal intensities reflects the kinetics of formation and depletion of the given intermediate. The overall
intensities of the signals are therefore not important, because we evaluate the signals only relative to each other.
We demonstrate the utility of this approach for the investigation of intermediates whose formation and depletion
can be described by a steadyꢀstate approximation (Equation 1).
k1[Cat][React] = (kꢀ1 + k2) [Int]eq
(Equation 1)
The reaction mixture is first prepared with unlabeled reactants (React) and allowed to react for time td.
During this time a certain concentration of the intermediate [Int] is achieved. After td elapses, the labeled reactant
(Reactlabel) is added to the reaction mixture, which takes the reaction mixture away from the steady state
conditions. At this time (t0), we start to monitor the reaction mixture by ESIꢀMS and record the abundances of
4
ACS Paragon Plus Environment

Page 5 of 19
Journal of the American Chemical Society
1
2
3
4
5
the labeled and unlabeled ions (Figure 1). The time evolution of the corresponding signals reflects the
reestablishment of the steady state conditions in the solution (Equation 2).
6
7
8
ꢀꢁꢂꢃꢄꢅ ꢆ ꢈꢉꢊꢋꢈꢌ ꢎꢏꢐ ꢑꢒ ꢓ ꢈꢉꢊꢋꢈꢌ ꢎꢏꢐ ꢆ ꢈ′ ꢎꢏꢐ ꢓ ꢎꢏꢐ , where k’ = kꢀ1 + k2.
ꢔꢁ
(Equation 2)
ꢇ
ꢍꢁ
ꢅ
ꢇ
ꢍꢁ
ꢅ
ꢅ
ꢁ
ꢅꢕ
ꢑꢒ
ꢀꢄ
9
At the time of mixing t0 the concentration of the labeled intermediate is zero ([Intlabel]0 = 0). If we
normalize [Int] + [Intlabel] = 1, then the time evolution of the Int and Intlabel signals (blue and red lines in Figure 1)
can be described as
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
ꢘ
ꢘ
ꢎꢏꢐ ꢆ ꢖꢉꢗ ꢄ ꢋ ꢎꢏꢐ ꢑꢒꢇ1 ꢓ ꢖꢉꢗ ꢄ
ꢍ
and
(Equation 3a)
(Equation 3b)
ꢁ
ꢅ
ꢁ
ꢅ
ꢄ
ꢘ
ꢁ
ꢎꢏꢐ
ꢙꢚꢛꢑꢙꢅꢄ
ꢁ
ꢙꢚꢛꢑꢙꢅ
ꢆ
ꢎꢏꢐ
ꢑꢒꢔ1 ꢓ ꢖꢉꢗ ꢄꢕ.
Figure 1. Ideal time evolution of the ESI-MS signals of unlabeled and labeled intermediates with
different half-lives (t1/2). The labeled intermediate (Intlabel) was added to the reaction mixture with a time
delay td. The solid lines show the signals time evolution for intermediates with t1/2 = 14 min and the
dashed lines for those with t1/2 = 0.7 min.
For the given ideal experiment is [Int]eq = [Intlabel]eq = 0.5. In the real experiments, we fit these values and obtain
an accurate ratio of the intermediates in the given reaction mixture at the given reaction time. The halfꢀlife of the
intermediate is determined as ln2/k’. Note that only the decomposition of the intermediates is studied (k’). The
rate of their formation (k1) does not influence the curve shapes, if we assume that both, labeled and unlabeled
intermediates are formed with the same rate constant k1.
While this model allows us to follow the kinetics of intermediates in reaction mixtures, it cannot be
applied for every reaction. For example, if the halfꢀlife of the intermediate is very short, the current technique is
not fast enough to monitor it. The dashed lines in Figure 1 show the steady state conditions establishment for an
intermediate with k’ = 1 minꢀ1 (halfꢀlife ~0.7 min). Clearly, taking into account a time required for the mixing of
the reaction mixtures and setting up the ESIꢀMS experiment, we cannot determine the relevant rate constant. For
such a situation, we only observe the relative concentrations of the unlabeled and labeled intermediates ([Int]eq
and [Intlabel]eq).
Before addressing the details of this method, we should point out that the labeling should be remote so
that it does not affect the rate of the studied reaction. Hence, we assume that the labeled as well as the unlabeled
reactant react with the same rate constants. We also note that it does not matter, whether the reaction time of the
unlabeled or the labeled reactant is longer. For the sake of simplicity, we usually start with unlabeled reactant,
but we have also performed the experiments in the reversed order and obtained the same results.
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 19
1
2
3
4
5
Results
6
7
8
9
NMR experiments. To discuss the role of the diaurated intermediates in the reaction, we have first determined
the kinetics of the formation of the products in the gold(I)ꢀcatalyzed methanol addition to 1ꢀphenylpropyne by
NMR (Scheme 1 and Figure 2). The methanol addition can take place either to the C(1) carbon atom or the C(2)
carbon atom of 1ꢀphenylpropyne and proceeds twice to form a methylvinylether in the first step and a ketal in the
second step (Figure 2). The rate of the individual steps depends critically on the catalyst used. From comparison
of [AuCl(PPh3)]/AgSbF6 and [AuCl(IPr)]/AgSbF6 it is clear that the latter catalyst provides a much faster
conversion with more than 99% reactant consumption after 35 minutes. In a comparable time the conversion in
the [AuCl(PPh3)] catalyzed reaction only reaches about 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
Figure 2. a) Reaction scheme for the kinetic model of gold-mediated methanol addition to 1-
phenylpropyne. b-d) Relative ratios of 1-phenylpropyne and the methanol addition products in
dependence of the reaction time catalyzed by a) 2.5 mol% [AuCl(PPh3)]/3 mol% AgSbF6, b) 2.5 mol%
[AuCl(IPr)]/3 mol% AgSbF6, and c) 1.25 mol% [AuCl(PPh3)]/1.25 mol% [AuCl(IPr)]/3 mol% AgSbF6 as
monitored by NMR spectroscopy. The yield of 1-phenyl-1-methoxypropene and 1-phenyl-2-
methoxypropene is shown in pink and red. Products of the addition of two molecules of methanol (sum
of the corresponding ketals and ketones formed upon hydrolysis with residual moisture) are shown in
green for the C(1) addition and in blue for the C(2) addition. The solid lines correspond to fits obtained
by the Octave program assuming the reaction scheme in a). The corresponding rate constants are
listed in Table 1.
6
ACS Paragon Plus Environment

Page 7 of 19
Journal of the American Chemical Society
1
2
3
Table 1. Rate constants for the gold(I) mediated methanol addition to 1ꢀphenylpropyne determined by the
4
5
6
kinetic modelling of the reaction yields monitored by NMR.
NMR
NMR
NMR
NMR
k1
k2
k11
k22
dm3.molꢀ1.minꢀ1
7
8
9
Catalyst
1.3 ± 0.3
3.4 ± 0.8
0.3 ± 0.1
1.4 ± 0.3
51 ± 14a
56 ± 40a
7.2 ± 0.7
120 ± 4
2.5 mol% [AuCl(PPh3)]/ 3 mol% AgSbF6
0.19 ± 0.03 0.12 ± 0.02
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
2.5 mol% [AuCl(PPh3)]/ 3 mol% AgSbF6/TsOH 0.26 ± 0.01 0.18 ± 0.01
2.5 mol% [AuCl(IPr)]/ 3 mol% AgSbF6
2.5 mol% [AuCl(IPr)]/ 3 mol% AgSbF6/TsOH
1.25 mol% [AuCl(PPh3)]/
5.6 ± 0.3
7.2 ± 0.9
16.9 ± 0.9
30 ± 6
1.0 ± 0.1
2.5 ± 0.3
3 ± 2
27 ± 5
1.25 mol% [AuCl(IPr)]/ 3 mol% AgSbF6
1.25 mol% [AuCl(PPh3)]/
1.1 ± 0.2
0.15b
3.8 ± 0.2
0.42b
2 ± 1
0.85b
81 ± 15
(42)a,b
1.25 mol% [AuCl(IPr)]/ 3 mol% AgSbF6/TsOH
b
2.5 mol% [AuCl(PMe3)]/ 3 mol% AgSbF6
a The large experimental error is given by a negligible abundance of the product of single methanol addition to the C(2) carbon atom. The
second step is two orders of magnitude faster, but the exact number can be determined only with large error bars.
b
Values are derived from previously published kinetics for the very same reaction performed under the same condition with the
[AuCl(PMe3)] catalyst.13
Kinetic modelling of the NMR monitored reactions reveals a fundamental difference between the
catalysts. Using the catalyst with the phosphine ligand renders the first reaction step about 2 orders of magnitude
slower than that with the catalyst with the N-heterocyclic carbene ligand (Table 1). On the other hand, the second
methanol addition is several times faster using the phosphine ligand. Clearly, the first methanol addition is the
rate determining step and the obvious explanation for the retarded rate of the reaction catalyzed by [AuCl(PPh3)]
is the formation of the diaurated intermediates. In the second methanol addition, the diaurated intermediates
cannot be formed and correspondingly, the second methanol addition is even faster with [AuCl(PPh3)] catalysis
than with [AuCl(IPr)] catalysis as could be anticipated based on the larger steric hindrance of the IPr ligand in
the final protodeauration step.
The regioselectivity of the reaction can be controlled electronically or sterically. For the bulky catalyst
[AuCl(IPr)], the addition preferentially proceeds at the C(2) carbon atom. This regioselectivity is
thermodynamically favored (see also results of DFT calculations).13 Using the [AuCl(PPh3)] catalysts, the C(1)
addition is slightly preferred. We have also tested the effect of acid addition. The addition of 4 mol% of p-
toluenesulfonic acid (TsOH) to the reaction mixture of both catalytic systems led to acceleration of the reaction
(Table 1). The first step of the reaction is accelerated 1.3 – 1.8 times. The acceleration is more pronounced for
C2 addition and it is rather similar for both catalysts. The effect in the second step exceeds that in the first step
and it is more pronounced for the [AuCl(IPr)] catalyst. Once again, the effect for C(2) addition is larger.
We have also performed an experiment, in which we have used the same amount of gold(I) catalyst, but
one half was bearing the triphenylphosphine ligand and the other half bears the NHC ligand. This combination
led to a significant drop of the rate of the first reaction step in comparison to the reaction catalyzed by
[AuCl(IPr)] only (to about one seventh) and a slight drop of the second step rate. The effect of the acid addition
was less pronounced than for the reactions with the individual catalysts.
If we compare the present results to the previously published results for the very same reaction
performed under the same conditions, but using the [AuCl(PMe3)] catalyst, we can see that similar to the
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 19
1
2
3
4
5
reaction catalyzed by [AuCl(PPh3)], the second reaction step is faster than the first one. The regioselectivity of
the reaction favors formation of the C(2) products.
6
ESI-MS experiments. The same reaction mixtures investigated by NMR were also probed by ESIꢀMS. The
7
8
9
spectra (Figures S3ꢀS5 in the Supporting Information) contain signals corresponding to the gold complexes
bearing the auxiliary ligand
L
and either methanol ([(L)Au(MeOH)]+), or 1ꢀphenylpropyne
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
([(L)Au(PhCCMe)]+), or the primary reaction product ([(L)Au(PhCCMe,MeOH)]+, observed only for L = IPr).
In addition, signals of the diaurated complexes [(L)2Au2(PhCCMe,OMe)]+ are observed. We have applied
delayed reactant labeling to the reaction mixtures (see the Experimental Details) and monitored the kinetics of
the degradation of the diaurated intermediates. Figures 3a,b show the relevant parts of the ESIꢀMS spectra of the
reaction mixture with the [Au(PPh3)]+ catalyst prepared in CH3OH and labeled with a delay of 5 minutes with
CD3OH. It can be clearly seen that shortly after the labeling of the reaction mixture, the concentration of the
unlabeled diaurated intermediate [(PPh3)2Au2(PhCCMe,OCH3)]+ (EPPh3) was much higher than that of the labeled
intermediate D3ꢀEPPh3 (Figure 3a). After about 60 minutes, the concentration of both, the labeled and unlabeled
intermediates was about the same (Figure 3b). On the contrary, complexes between the catalyst and solvent
molecules ([(PPh3)Au(CH3OH)]+ and [(PPh3)Au(CD3OH)]+) are observed with a constant abundance ratio,
because they are formed in a fast equilibrium that cannot be followed with the current technique.
Figure 3. Delayed reactant labeling method. a) ESI-MS spectrum recorded 2 min after the CD3OH
addition to the reaction mixture of 1-phenylpropyne with 2.5 mol% [AuCl(PPh3)] and 3 mol% AgSbF6 in
CH3OH; time-delay was 5 min. b) The same as a), but the ESI-MS spectrum was recorded 60 minutes
after the labeling of the reaction mixture. c) Mutual time evolution of the EPPh3 and D3-EPPh3 signals in
the experiment described in a) (red symbols). The gray symbols show the effect of the TsOH addition
to the reaction mixture. The dashed lines correspond to the data fits according to Eq. 3a and 3b. d)
Mutual signal evolution of EIPr and D3-EIPr after the CD3OH delayed labeling of the reaction mixture of
1-phenylpropyne with 2.5 mol% [AuCl(IPr)] and 3 mol% AgSbF6 in CH3OH with variable time delays
(color coded). The lines represent fits according to Eq. 4. Mutual time evolutions of the (e) EPPh3 and
D3-EPPh3 and (f) EMix and D3-EMix signals in the experiment with 1.25 mol% [AuCl(PPh3)], 1.25 mol%
[AuCl(IPr)], and 3 mol% AgSbF6 in CH3OH (red symbols). Delayed labeling with CD3OH and td = 5
min. The gray symbols are obtained from the same experiment but with the TsOH addition. The
dashed lines correspond to the data fits according to Eq. 3a and 3b.
8
ACS Paragon Plus Environment

Page 9 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
The time evolution of the relative abundances of EPPh3 and D3ꢀEPPh3 can be seen in Figure 3c (note that
the sum of EPPh3 and D3ꢀEPPh3 was normalized to 1. If we fit the relative abundances using Eq. 3a and 3b (the
dashed lines in Figure 3), we obtain the rate constant k’ ~ 0.063 which corresponds to a halfꢀlife of ~11 minutes
(Table 3). We have examined the effect of the time delay (Table 2). The delay td should be sufficiently long so
that an appreciable amount of the unlabeled product is formed. For very short time delays, the difference
between the EPPh3 and D3ꢀEPPh3 relative abundances would be small and the determined halfꢀlife would be
burdened by a large experimental error. The time delay should also not be too long, because then other effects
come into play (e.g. degradation of catalyst, involvement of byꢀproducts, etc.). If these principles are obeyed,
then the effect of the time delay is rather small as evident from Table 2. We were able to obtain consistent results
over large time delays. For the remaining experiments, we always took a time delay of 5 min.
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
We note in passing that for this series of experiments, a different batch of the gold catalyst was used and
the rate constants were consistently by about 35 % larger than in the remaining experiments listed in Table 3. We
have also monitored the reaction by NMR using this new batch of the catalyst and it was by 37% faster (Figure
S2a). We are showing these results to demonstrate that changes in reaction mixture are directly reflected by our
experiments.
Table 2. The effect of a time delay on the determination of the rate constant k’ for the degradation of the
diaurated intermediate EPPh3 in the reaction catalyzed by [AuCl(PPh3)] and the rate constant k for the degradation
of a reactant yielding the diaurated complex EIPr in the reaction catalyzed by [AuCl(IPr)].
Catalyst
time delay
td [min]
2.5
rate constant
halfꢀlife
t1/2
8.7 ± 1.2
k´ [minꢀ1]
2.5 mol% [AuCl(PPh3)]/ 3 mol% AgSbF6
0.081 ± 0.010
10
30
0.086 ± 0.005
0.083 ± 0.001
8.1 ± 0.4
8.4 ± 0.1
[EIPr]rel,td
[D3ꢀEIPr]rel,∞
Catalyst
time delay
td [min]
2.5
rate constant
kMS [minꢀ1]
0.13 ± 0.06
2.5 mol% [AuCl(IPr)]/ 3 mol% AgSbF6
0.49
0.46
0.44
0.40
0.37
0.01
0.08
0.12
0.20
0.26
5
0.14 ± 0.01
0.13 ± 0.02
0.14 ± 0.01
0.10± 0.02
7.5
10
20
This approach gives us a powerful tool to study the effect of the reaction conditions on the halfꢀlife of
the intermediates (average of several measurements for each experiment together with the derived experimental
error can be found in Table 3). First, we have tested whether the labeling of the reaction mixture with CD3OD
instead of CD3OH would results in significant changes (Entries 1 and 2 in Table 3). The derived halfꢀlives are
slightly longer than those obtained with CD3OH, but the effect is in the range of the experimental error (more
results with CD3OD can be found in the Supporting Information). Secondly, we have investigated the effect of
acid addition (Entry 3). Addition of 4 mol% of p-toluenesulfonic acid (TsOH) results in a decrease of the halfꢀ
life of the [(PPh3)2Au2(PhCCMe,OCH3)]+ intermediate to 7 minutes, hence almost to one half (experimental data
are shown in gray in Figure 3c). This is consistent with a view that the acid catalyzed protodeauration step is
responsible for the depletion of the diaurated intermediates.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 19
1
2
3
4
5
6
7
8
9
We have also tested the effect of silver salt addition on the halfꢀlife of the diaurated intermediates
(Entries 7 – 9). In some of the goldꢀcatalyzed reactions, a peculiar silver effect was observed that accelerated or
even enabled the given reactions.41 We have measured the diaurated intermediate halfꢀlives in dependence of the
silver cations’ concentration. The results show that for the ratios of [AuCl(PPh3)] to [AgSbF6] ranging from 1 :
0.5 via 1 : 1.2 to 1 : 5 the halfꢀlives of the diaurated intermediates are the same within the experimental error. It
means that we do not observe any participation of the silver cations in the depletion of the diaurated
intermediates. We note in passing that we have performed NMR experiments with a large excess of the silver
salt concentration and in agreement, no effect on the overall reactivity was found (Figure S2b in the Supporting
Information).42,43
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
It is interesting to note that the halfꢀlife of the diaurated intermediates did not change even when the
concentration of the silver salt was only half of that of the gold catalyst (Entry 7). Probably, a lower
concentration of the active catalyst in the reaction mixture was investigated in this experiment, but it did not
influence the determined halfꢀlife of the diaurated intermediates.
Table 3. Halfꢀlives determined for the diaurated intermediates EPPh3 in the reaction catalyzed by [AuCl(PPh3)]
and for the diaurated intermediates EPPh3 and EMix in the reaction catalyzed by a mixture of [AuCl(PPh3)] and
[AuCl(IPr)].
EPPh3
1/2 [min]
rate constant
Catalyst
Addition
CD3OH
Additive
t
k´ [minꢀ1]
1
ꢀ
ꢀ
11.3 ± 1.7
12.2 ± 0.8
7.0 ± 1.6
0.063 ± 0.009
0.057 ± 0.004
0.101 ± 0.023
2
3
CD3OD
CD3OH
TsOH
2.5 mol% [AuCl(PPh3)]/3 mol% AgSbF6
4
5
6
PhCCCH3,CD3OH
PhCCCH3,CD3OH
ꢀ
7.3 ± 0.6
0.096 ± 0.009
TsOH
3.7 ± 0.3
4.2 ± 0.4
10.9 ± 1.1
0.188 ± 0.014
0.167 ± 0.018
0.064 ± 0.007
PhCCCD3
CD3OH
ꢀ
ꢀ
7
2.5 mol% [AuCl(PPh3)]/1.5 mol% AgSbF6
2.5 mol% [AuCl(PPh3)]/6.0 mol% AgSbF6
2.5 mol% [AuCl(PPh3)]/15 mol% AgSbF6
8
9
CD3OH
CD3OH
ꢀ
ꢀ
10.3 ± 0.8
0.068 ± 0.005
11.0 ± 1.3
0.063 ± 0.008
EPPh3
EMix
t
1/2 [min]
t1/2 [min]
10
11
12
13
CD3OD
CD3OD
ꢀ
TsOH
ꢀ
7.0 ± 1.0
3.5 ± 0.3
3.2 ± 0.3
19.2 ± 2.5
12.6 ± 0.8
17.7 ± 1.3
1.25 mol% [AuCl(PPh3)]/1.25 mol% [AuCl(IPr)]/
3 mol% AgSbF6
CH3OH, PhCCCD3
CD3OD
2.5 mol% [AuCl(PPh3)]/1.25 mol% [AuCl(IPr)]/
4.5 mol% AgSbF6
ꢀ
ꢀ
8.1 ± 1.4
16.2 ± 1.9
19.9 ± 4.5
14
15
1.25 mol% [AuCl(PPh3)]/2.5 mol% [AuCl(IPr)]/
4.5 mol% AgSbF6
CD3OD
CD3OD
not detected
2.5 mol% [AuCl(PPh3)]/2.5 mol% [AuCl(IPr)]/
6 mol% AgSbF6
ꢀ
8.4 ± 2.4
14.5 ± 1.5
EPMe3
1/2 [min]
rate constant
t
k´ [minꢀ1]
16
CD3OD
a
2.5 mol% [AuCl(PMe3)]/ 3 mol% AgSbF6
4.0 ± 0.3a
0.17 ± 0.1a
a
Values are derived from previously published kinetics for the [(PMe3)2Au2(PhCCMe,OCH3)]+ intermediate detected for the reaction
performed under the same condition with the [AuCl(PMe3)] catalyst.13
10
ACS Paragon Plus Environment

Page 11 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Besides reactionꢀmixture labeling with CD3OH, we can label other reactants, e.g. 1ꢀphenylpropyne as
was previously mentioned. Using PhꢀCCꢀCD3 we have obtained a halfꢀlife for the EPPh3 diaurated intermediates
of 4.2 minutes (Entry 6). As we have added only an equimolar amount of PhꢀCCꢀCD3 to the standard reaction
solution, we were working with a twofold catalyst and fourfold alkyne concentrations in the final reaction
mixture. In order to keep the concentration equal in the remaining experiments and just double the concentration
of the alkyne, we have labeled the reaction mixture with CD3OH and one additional equivalent of PhꢀCCꢀCH3
(Entry 4). This experiment gives a halfꢀlife for the diaurated intermediates of 7.3 min. It clearly shows that the
alkyne plays a role in the decomposition of the diaurated intermediates. Finally, we have repeated the same
experiment with CD3OH/PhꢀCCꢀCH3 labeling, but with addition of 4 mol % of p-toluenesulfonic acid (Entry 5).
As expected, it resulted in a further decrease of the halfꢀlife. Consistently with the results above, the halfꢀlife
dropped to about one half (3.7 min).
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
We have also performed the same experiments with the [AuCl(IPr)] catalyst (Figure 3d). Already at first
sight, it is clear that the shapes of the curves representing the relative abundances of the unlabeled and labeled
signals do not correspond to the establishment of steadyꢀstate conditions. If we assume that the EIPr diaurated
complexes do not decompose at all as suggested earlier27 and that the unlabeled EIPr and labeled D3ꢀEIPr
complexes are formed with the same rate, then we can fit the curve for the labeled product with a simple
exponential function:
ꢢꢣ
ꢙꢚꢛꢑꢙ
ꢇꢄꢤꢄ ꢍ
ꢥ
ꢖꢉꢗ
.
(Equation 4)
ꢁ
ꢜꢝꢞꢟ
ꢅ
ꢁ
ꢆ ꢜꢝꢞꢟ
ꢙꢚꢛꢑꢙꢅ
ꢠꢑꢙ,ꢡ ꢓ
α
ꢠꢑꢙ,ꢄ
This expression assumes normalization of [Prod] + [Prodlabel] = 1. At the end of the reaction, we obtain
the relative yields of the labeled product [Prodlabel
and that of the unlabeled product [Prod]rel,∞ = 1 ꢀ
]
rel,∞
[Prodlabel
]
in that [Prod]rel,td = (1ꢀ2[Prodlabel
]rel,∞) was formed during the delay before the labeling of the
rel,∞
reaction mixture. The exponential function expresses the decay of the reactant from which the diaurated complex
is formed multiplied by the fitting parameter α.
The solid lines in Figure 3d show fits of the results obtained with different timeꢀdelays and Table 2 lists
the fitted parameters. The model correctly shows that with the prolongation of the timeꢀdelay td the relative yield
of the unlabeled product [EIPr]rel,td formed during this time grows. The rate constant determined for the
degradation of the reactant that provides the diaurated complexes (i.e. most probably 1ꢀphenylpropyne) under the
conditions of the experiment is estimated as 0.13 ± 0.02 minꢀ1 (average of the kMS values listed in Table 2). For
comparison, the value obtained from the kinetic modelling of the NMR results is 0.135 minꢀ1 (i.e.
(k1NMR+k2NMR)•[Au(IPr)]+ = 22.5•0.006), which is in very good agreement. As the determination of the kinetics
for the reactants and products is not the aim of this paper, we do not elaborate it further.
The last experiment involves the determination of the halfꢀtimes of the intermediates in the reaction
catalyzed by a mixture of the [AuCl(PPh3)] and [AuCl(IPr)] catalysts. We can not only observe the homoꢀ
diaurated intermediates [(PPh3)2Au2(PhCCMe,OCH3)]+ and [(IPr)2Au2(PhCCMe,OCH3)]+, but also – the
overwhelmingly dominant signal in the ESIꢀMS spectrum of this reaction mixture – the mixed
[(PPh3)(IPr)Au2(PhCCMe,OCH3)]+ intermediate. The diaurated intermediate with two carbene ligands
[(IPr)2Au2(PhCCMe,OCH3)]+ was very low abundant, in some of the experiments it was not even detected (see
the Supporting Information).
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 19
1
2
3
4
5
6
7
The rate constant determined for the degradation of [(PPh3)2Au2(PhCCMe,OCH3)]+ in the reaction
mixture with both catalysts is larger than that determined in the reaction mixture with the
(triphenylphosphino)gold catalyst only. Using delayed labeling by CD3OD, the halfꢀtimes of the intermediates
drop to about one half both without and with the addition of the TsOH acid (7.0 min and 3.5 min, respectively;
Entries 10 and 11 in Table 3). The halfꢀlives of the mixed [(PPh3)(IPr)Au2(PhCCMe,OCH3)]+ (EMix
)
8
9
intermediates are considerably longer. It is important to note that during the equilibration of the concentrations of
the EMix and D3ꢀEMix intermediates, the alkyne reactant is largely depleted (compare with the NMR experiments).
It will surely lead to a decreasing rate of intermediate formation which can lead to the decreasing of their signal
intensities in the MS spectrum. Nevertheless, the rate constant k’ determined according to Eq. 3 should not be
significantly influenced. The EMix halfꢀlife in the reaction mixture is about 19 min and with the addition of 4
mol% of TsOH it drops to about 13 min. Labeling of the reaction mixture with the PhꢀCCꢀCD3 reactant leads
again to a faster degradation of the diaurated intermediates with the halfꢀlife of [(PPh3)2Au2(PhCCMe,OCH3)]+
determined as 3.2 min and that of [(PPh3)(IPr)Au2(PhCCMe,OCH3)]+ as 17.7 min (Entry 12). We have
performed the same experiments with different concentrations and proportions of the catalysts and we have not
determined any significant influence on the determined halfꢀlives (Entries 13 – 15).
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 calculations. In order to aid the discussion of the results, we have performed exploratory DFT calculations
with an implicit methanol solvation model (Table 4). The relative energies are given with respect to the
separated reactant molecules (1ꢀphenylpropyne and methanol) and the methanol adduct of the catalyst (i.e.
[(L)Au(CH3OH)]+, where L = PPh3 or IPr). For the deprotonation steps, we have assumed the formation of
protonꢀbound methanol dimer (see the Computational Details). The energies in Table 4 are calculated on one
potential energy surface and the molecules/ions considered for each minimum are listed in the first two columns.
The initial formation of the complex between the catalyst and 1ꢀphenylpropyne is exothermic by 14 kJ
molꢀ1 for L = PPh3 and 32 kJ molꢀ1 L = IPr. The addition of a methanol molecule to these complexes is
endothermic by 32 and 43 kJ molꢀ1, respectively. We have located this highꢀenergy intermediate only for the
methanol addition to the C(2) position of 1ꢀphenylpropyne. All attempts to localize the C(1) methanol adduct in
the gas phase resulted in its dissociation back to the reactants. It was possible to find these intermediates if we
performed the geometry optimization with the methanol solvation effect. These values are given in italics in
Table 4.
The next step has to be associated with deprotonation. Simple deprotonation is exothermic by about 50
kJ molꢀ1. Formation of the C(2) addition product is slightly more preferred, but the differences are only on the
order of units of kJ molꢀ1. Exchange of the proton by the [(L)Au]+ cation at the oxygen atom is exothermic by
almost 100 kJ molꢀ1 and again formation of the C(2) product is preferred. The alternative deprotonation
associated with the direct formation of the geminally diaurated intermediate is the most exothermic process with
an energy gain of more than 150 kJ molꢀ1. Regioselectivity is again slightly in favor of the formation of the C(2)
product.
The monoaurated intermediate can be protonated, which directly leads to the gold complex of the
product (C + H+ → F). This process is exothermic by more than 100 kJ molꢀ1 and should thus be facile.
Alternatively, the protonation can be associated with deauration, but it is by about 40 kJ molꢀ1 (L = PPh3) or 50
kJ molꢀ1 (L = IPr) less energy favored. The diaurated intermediates can decompose by protodeauration leading to
12
ACS Paragon Plus Environment

Page 13 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
the gold complex of the product (F). This step is endothermic by about 10 – 15 kJ molꢀ1 for L = PPh3 and almost
thermoneutral for L = IPr. If we consider protodeauration associated with the transfer of one of the gold cations
to the alkyne reactant (as suggested by the experimental results), the process becomes thermoneutral for L =
PPh3 and exothermic for L = IPr.
8
9
Table 4. Relative energies (ꢀꢀH298KMeOH and ꢀꢀH298KGP)a in kJ molꢀ1 of different intermediates on the B3LYPꢀ
D3/ccꢀpVTZ:LanL2DZ(Au) potential energy surface for the gold mediated methanol addition to 1ꢀ
phenylpropyne.
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
C(1) addition
L = PPh3
C(2) addition
L = PPh3
C(1) addition
L = IPr
C(2) addition
L = IPr
Reactants/
Catalysts/
Products
Intermediates
ꢀꢀH298KMeOH [kJ molꢀ1] (ꢀꢀH298KGP [kJ molꢀ1])
3x 1ꢀphenylpropyne
2x [(L)Au(CH3OH)]+
2x CH3OH
0 (0)
0 (0)
0 (0)
0 (0)
2x 1ꢀphenylpropyne
[(L)Au(CH3OH)]+
3x CH3OH
AuL
ꢀ14 (ꢀ31)
ꢀ14 (ꢀ31)
ꢀ32 (ꢀ48)
ꢀ32 (ꢀ48)
R
R' (A)
H
2x 1ꢀphenylpropyne
[(L)Au(CH3OH)]+
2x CH3OH
R
O
CH3
(B)
25b
16b, 18 (ꢀ8)
ꢀ31 (33)
not found
ꢀ30 (40)
ꢀ83 (ꢀ64)
11 (ꢀ13)
ꢀ34 (38)
ꢀ88 (ꢀ64)
(L)Au
R
R'
OCH3
2x 1ꢀphenylpropyne
[(L)Au(CH3OH)]+
(CH3OH)2H+
ꢀ31c (24)c
ꢀ23d (38)d
(L)Au
R'
(C)
(L)Au
2x 1ꢀphenylpropyne
CH3OH
R
O
CH3
ꢀ68 (ꢀ56)
ꢀ82 (ꢀ70)
(CH3OH)2H+
(L)Au
R'
(D)
R
OCH3
2x 1ꢀphenylpropyne
CH3OH
ꢀ149 (ꢀ141)
ꢀ153 (ꢀ140)
ꢀ141 (ꢀ165)
ꢀ137 (ꢀ124)
ꢀ141 (ꢀ172)
ꢀ155 (ꢀ135)
ꢀ157 (ꢀ181)
(L)Au
R'
(CH3OH)2H+
Au
(L)
(E)
Au(L)
2x 1ꢀphenylpropyne
[(L)Au(CH3OH)]+
2x CH3OH
ꢀ126 (ꢀ157)
R
H
OCH3
R'
(F)
F + A
1ꢀphenylpropyne
3x CH3OH
ꢀ140 (ꢀ188)
ꢀ97 (ꢀ110)
ꢀ155 (ꢀ197)
ꢀ106 (ꢀ114)
ꢀ173 (ꢀ220)
ꢀ97 (ꢀ110)
ꢀ189 (ꢀ229)
ꢀ106 (ꢀ114)
Product
2x 1ꢀphenylpropyne
2x [(L)Au(CH3OH)]+
CH3OH
Product
1ꢀphenylpropyne
[(L)Au(CH3OH)]+
2x CH3OH
A
ꢀ111 (ꢀ141)
ꢀ126 (ꢀ172)
ꢀ121 (ꢀ145)
ꢀ135 (ꢀ177)
ꢀ129 (ꢀ158)
ꢀ161 (ꢀ206)
ꢀ138 (ꢀ162)
ꢀ170 (ꢀ210)
Product
3x CH3OH
2x A
a The geometries were optimized in the gas phase and the solvation energies in methanol were calculated using PCM method.
bGeometries of the B intermediates were optimized using the PCM solvation model (values are given in italics). The B adducts
corresponding to C(1) addition do not represent minima on the potential energy surface in vacuum.
cSmall imaginary frequency resulted always from the calculations for this stationary point. We have applied a thermal correction to enthalpy
from the other conformer of this isomer.
dA second possible conformer of the C(1) monoaurated intermediate. It is slightly higher in energy, all second derivatives were positive.
The computational results do not give any significant lead as to why the regioselectivity of the first
addition of methanol to 1ꢀphenylpropyne differs on changing the ancillary ligand at the gold catalyst. It also does
not show why the reaction with [(PPh3)Au]+ leads to the diaurated intermediates opposed to the reactions with
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 19
1
2
3
4
5
[(IPr)Au]+. The energy profiles for both reactions are rather similar. Note that solvation is considered by an
implicit model. Using this approach, it is not possible to unravel solvation steric effects that play very significant
role in this reaction as is discussed below.
6
7
8
9
Discussion
The NMR data for the “[(IPr)Au]+” catalyzed reaction shows that the first methanol addition is several
times faster than the second one and the methoxy group addition proceeds preferentially to the C(2) position of
1ꢀphenylpropyne. Both reaction steps are accelerated by acid addition with acceleration of the second methanol
addition more pronounced. The acceleration caused by acid addition (i.e. TsOH here) can either suggest that the
protodeauration step is the rate determining step,44 or that there is active involvement of the anions in the initial
deprotonation of adduct B to form neutral monoaurated intermediates.45 Tosylate was suggested as particularly
favorable anion.46ꢀ48 The ESIꢀMS delayed reactant labeling method shows that the diaurated complexes
[(IPr)2Au2(PhCCMe,OCH3)]+ are not in equilibrium with either the reactants or products, but rather formed as
stable complexes, i.e. byproducts. The reaction therefore does not proceed via their formation.
The reaction catalyzed by “[(PPh3)Au]+” is slower than the “[(IPr)Au]+” catalyzed one due to the
considerable deceleration of the first methanol addition to the alkyne. The second methanol addition is an order
of magnitude faster than the first one and it is also faster than the analogous reaction catalyzed by “[(IPr)Au]+”.
The regioselectivity is switched towards the preferential methoxy addition to the C(1) position of 1ꢀ
phenylpropyne. Both reaction steps are again accelerated by the acid addition. The most obvious explanation of
the deterioration of the “[(PPh3)Au]+” mediated reaction lies in the formation of the diaurated intermediates.
The delayed reactant labeling ESIꢀMS experiments show that the [(PPh3)2Au2(PhCCMe,OCH3)]+
formation and depletion can be approximated by steadyꢀstate conditions. The determined rate constants are
associated with the depletion of the EPPh3 intermediates. It clearly shows that the addition of 4 mol% of TsOH
increases the depletion rate constant by about 50%. In comparison, the NMR experiments show that the reaction
rates for the first methanol addition to the C(1) and C(2) position increase on acid addition by 37% and 50%,
respectively. This nice agreement suggests that depletion of the diaurated intermediates is the rate determining
step in this reaction. Secondly, we have shown that doubling the concentration of the alkyne also leads to an
increase of the rate constant for the depletion of the diaurated intermediates. It means that the alkyne takes an
active part in the process, which might thus correspond to transmetalation.
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
Next to the effect of the acid, which is exposed in the same way to the rate of the diaurated intermediate
depletion as to the overall reaction rate, it is interesting to note similar relationships between the “[(PPh3)Au]+”
catalyzed reaction and the previously published reaction catalyzed by “[(PMe3)Au]+”. The slower rate of
[(PPh3)2Au2(PhCCMe,OCH3)]+ decomposition (k’PPh3 : k’PMe3 = 0.06 : 0.17, see Table 3) translates quite well to
the slower reaction rate of 1ꢀphenylpropyne with methanol determined by the NMR experiments. This is strong
support for the reaction pathway taking place via the diaurated intermediates. An alternative explanation
involves a side equilibrium between the monoaurated and diaurated intermediates. If it would hold true, we
would expect formation as well as decomposition of [(PPh3)2Au2(PhCCMe,OCH3)]+ at a slower rate than the
formation and decomposition of [(PMe3)2Au2(PhCCMe,OCH3)]+ due to steric effects
. The slower formation of
the [(PPh3)2Au2(PhCCMe,OCH3)]+ intermediates would result in a faster reaction cycling via the monoaurated
intermediates. Hence, the overall reaction would be expected to be faster than the reaction catalyzed by
14
ACS Paragon Plus Environment

Page 15 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
“[(PMe3)Au]+”. Such a scenario is observed for the “[(IPr)Au]+” catalyzed reaction, but not for that catalyzed by
“[(PPh3)Au]+”.
The experiment with the mixture of “[(PPh3)Au]+” and “[(IPr)Au]+” catalysts shows that the first
methanol addition is mainly mediated by “[(IPr)Au]+” as it is faster than in the reaction with “[(PPh3)Au]+” alone
and it has the same selectivity towards the C(2) addition as the reaction solely with “[(IPr)Au]+”. The second
methanol addition is mediated by “[(PPh3)Au]+” as the reaction rates are almost the same. The first reaction step
is slowed down about seven times in comparison with the “[(IPr)Au]+” catalyzed reaction. Clearly the reason is
the formation of the mixed diaurated intermediates [(PPh3)(IPr)Au2(PhCCMe,OCH3)]+ (EMix) which fully
decompose but at a very slow 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
What is the origin of these mixed diaurated intermediates? The formation of these intermediates is not
sterically hindered as it was the case for those bearing two IPr ligands ([(IPr)2Au2(PhCCMe,OCH3)]+). From the
drop of the rate we know that about 2/3 of the [(IPr)Au]+ cations are blocked in the diaurated intermediates. It
also suggests that most of the [(PPh3)Au]+ cations are blocked in the EMix intermediates. The prevalence of the
mixed intermediates is associated with their stability which leads to their accumulation with time. The shorter
halfꢀlife of the EPPh3 intermediates in the reaction mixture containing [(IPr)Au]+ cations is probably associated
with the faster formation of the vinylether primary products that can take part in transmetalation of the diaurated
intermediates similar to what was observed for the alkyne reactant. The interaction energy of [(PPh3)Au]+ with
the vinylether products amounts to 29 kJ molꢀ1 for the C(1) product and 35 kJ molꢀ1 for the C(2) product, hence
larger than with the alkyne (14 kJ molꢀ1).
Scheme 3. Suggested reaction mechanism. (a) Mechanism for the reactions catalyzed by “[Au(L)]+”,
where L is bulky. (b) Dual activation mechanism catalyzed by “[Au(L)]+” for which the [Au(L)]+ ⇔ H+
exchange is faster than unassisted deprotonation. (c) Protodeauration of diaurated intermediates
assisted by the alkyne reactant.
Based on all of the experimental data and computational results, we suggest the following reaction
mechanism (Scheme 3). The initial addition of the methanol molecule to the goldꢀactivated alkyne is a reversible
endothermic step. The selectivity towards C(1) or C(2) addition is given by the following irreversible step. It can
be a simple deprotonation and then the selectivity will be tuned towards the formation of the more stable
monoaurated intermediate. In addition, steric effects can also play a role: In the C(1) addition product, the ligated
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 19
1
2
3
4
5
gold cation is bound to the C(2) carbon with a Zꢀconfiguration relative to the reactant phenyl ring. This might
represent a steric hindrance for the formation of the C(1) addition product. The deprotonation can be catalyzed
by the anions in the reaction mixture.
6
The alternative deprotonation pathway is assisted by the intake of another [(L)Au]+ cation. This reaction
is strongly exothermic and is most probably kinetically controlled. The initial B adduct corresponding to the C(2)
addition has the phenyl group in the geminal position with the [(L)Au]+ unit (Figure 4). The phenyl group is
turned out of the plane and sterically shades the carbon to be attacked by the second [(L)Au]+ cation. The
intermediate with the methanol molecule attached to the C(1) carbon atom is less sterically hindered for the
attack of the second [(L)Au]+ cation. Note that for the small [(PMe3)Au]+ catalyst, where steric effects play a less
important role, the thermodynamically preferred C(2) product is formed. The goldꢀassisted deprotonation is
probably much faster than the unassisted or anionꢀassisted deprotonation and that is why the reaction pathway
leads via the formation of the diaurated intermediates. It can only be suppressed, if the [(L)Au]+ cations bear
sufficiently bulky ancillary ligands L (i.e. IPr here).
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. Suggested reaction coordinates for the gold assisted deprotonation of the methanol adduct.
The yellow arrows show the carbon atom attacked by the second gold cation. The attacks would be
probably favorable from the bottom of the complexes. The C(2) intermediate is more sterically
hindered due to the phenyl group. The structures were obtained by the geometry optimization (B3LYP-
D3/cc-pVTZ:LanL2DZ(Au)) in methanol using the PCM model.
As shown by our delayed reactant labeling experiments, the degradation of the diaurated intermediates
is firstꢀorder dependent on the alkyne concentration and it is catalyzed by acid. We suggest that the
protodeauration proceeds to form the goldꢀcomplex of the alkyne and the gold complex of the product (see
Scheme 3c). Clearly, protonation of the neutral monoaurated intermediate will be much faster than that of the
positively charged diaurated intermediate. This explains why the reaction catalyzed by a bulky catalyst, which
blocks the reaction pathway via the diaurated intermediates, is faster than that catalyzed by a sterically less
demanding catalyst. For the second methanol addition, no diaurated intermediates can be formed. The reaction is
therefore faster with the less bulky catalyst, because the formation of the monoaurated intermediate as well as
the protodeauration reaction are less sterically hindered.
Conclusions
We present a new method for extraction of kinetic data for intermediates in reaction mixtures under the
condition that they are observable by ESIꢀMS. We have demonstrated that for intermediates formed in a steadyꢀ
state approximation, the effect of reaction conditions on their halfꢀlife can be extracted. The effect of e.g. acidity,
catalyst concentration, or substrate concentration can be studied. In connection to the classical kinetics obtained
from NMR experiments, this method can bring unprecedented data on key lowꢀabundant intermediates.
16
ACS Paragon Plus Environment

Page 17 of 19
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
The method was applied for the investigation of diaurated intermediates in methanol addition to
alkynes. The diaurated complexes were considered to be offꢀcycle intermediates that slow down the addition
reaction because of the catalyst being retained as diaurated complexes. We have shown here that they are not offꢀ
cycle intermediates. The initial goldꢀmediated methanol addition is an endothermic step. The driving force
towards the products is the exchange of the proton bound to the oxygen atom by a gold cation, leading to the
diaurated intermediates. Their protodeauration is a slow process and therefore the overall reaction is slow. This
reaction pathway can be suppressed, if the gold cations bear sterically demanding ancillary ligands such as a N-
heterocyclic carbene ligand IPr. The reaction then dominantly proceeds by the simple deprotonation step towards
monoaurated intermediates. While the initial deprotonation step is slow, the overall reaction is faster, because
protodeauration of the the monoaurated intermediates is easier than with the diaurated intermediates.
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
ASSOCIATED CONTENT
Supporting Information. Further experimental details, NMR spectra, ESIꢀMS spectra, delayed reactant labeling
experiments, further computational results, and geometries of all optimized structures. “This material is available free of
charge via the Internet at http://pubs.acs.org.”
AUTHOR INFORMATION
Corresponding Authors
* roithova@natur.cuni.cz
ACKNOWLEDGMENT
This work was supported by the Grant Agency of the Czech Republic (14ꢀ20077S) and the European Research
Council (StG ISORI). The authors thank Dr. Andrew Gray for his help in preparing the manuscript.
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 19
1
2
3
References
4
5
6
7
(1)
(2)
(3)
Chen, P. Angew. Chem. Int. Ed. 2003, 42, 2832.
Yunker, L. P. E.; Stoddard, R. L.; McIndoe, J. S. J. Mass Spectrom. 2014, 49, 1.
O'Hair, R. A. J. Int. J. Mass Spectrom. 2015, 377, 121.
8
9
(4)
(5)
(6)
Schröder, D. Acc. Chem. Res. 2012, 45, 1521.
Vikse, K. L.; Ahmadi, Z.; McIndoe, J. S. Coord. Chem. Rev. 2014, 279, 96.
Eberlin, M. N. Eur. J. Mass Spectrom. 2007, 13, 19.
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
(7)
(8)
(9)
(10)
(11)
(12)
(13)
Bock, K.; Feil, J. E.; Karaghiosoff, K.; Koszinowski, K. Chem. Eur. J. 2015, 21, 5548.
Kretschmer, R.; Schlangen, M.; Schwarz, H. Chem. Eur. J. 2012, 18, 40.
Santos, L. S. Eur. J. Org. Chem. 2008, 235.
Ringger, D. H.; Kobylianskii, I. J.; Serra, D.; Chen, P. Chem. Eur. J. 2014, 20, 14270.
Roithová, J. Chem. Soc. Rev. 2012, 41, 547.
Oomens, J.; Sartakov, B. G.; Meijer, G.; Von Helden, G. Int. J. Mass Spectrom. 2006, 254, 1.
Roithova, J.; Jankova, S.; Jasikova, L.; Vana, J.; Hybelbauerova, S. Angew. Chem. Int. Ed. 2012, 51,
8378.
Kohler, M.; Leary, J. A. Int. J. Mass Spectrom. Ion Process. 1997, 162, 17.
Schröder, D.; Weiske, T.; Schwarz, H. Int. J. Mass Spectrom. 2002, 219, 729.
Schröder, D. Phys. Chem. Chem. Phys. 2012, 14, 6382.
(14)
(15)
(16)
(17)
Vikse, K. L.; Ahmadi, Z.; Manning, C. C.; Harrington, D. A.; McIndoe, J. S. Angew. Chem. Int. Ed.
2011, 50, 8304.
(18)
(19)
(20)
(21)
(22)
(23)
(24)
(25)
(26)
(27)
(28)
(29)
(30)
(31)
(32)
Limberger, J.; Leal, B. C.; Monteiro, A. L.; Dupont, J. Chem. Sci. 2015, 6, 77.
von Rekowski, F.; Koch, C.; Gschwind, R. M. J. Am. Chem. Soc. 2014, 136, 11389.
For the review on diaurated intermediates see: Hashmi, A. S. K. Acc. Chem. Res. 2014, 47, 864.
Lein, M.; Rudolph, M.; Hashmi, A. S. K.; Schwerdtfeger, P. Organometallics 2010, 29, 2206.
Mizushima, E.; Sato, K.; Hayashi, T.; Tanaka, M. Angew. Chem. Int. Ed. 2002, 41, 4563.
Kovács, G.; Ujaque, G.; Lledós, A. J. Am. Chem. Soc. 2008, 130, 853.
Nolan, S. P. Acc. Chem. Res. 2011, 44, 91.
Obradors, C.; Echavarren, A. M. Chem. Comm. 2014, 50, 16.
Roithova, J; Hrusak, J.; Schröder, D.; Schwarz, H. Inorg. Chim. Acta 2005, 358, 4287.
Zhdanko, A.; Maier, M. E. Chem. Eur. J. 2014, 20, 1918.
Hashmi, A. S. K.; Schuster, A. M.; Rominger, F. Angew. Chem. Int. Ed. 2009, 48, 8247.
Liu, L. R.; Hammond, G. B. Chem. Soc. Rev. 2012, 41, 3129.
Seidel, G.; Lehmann, C. W.; Fürstner, A. Angew. Chem. Int. Ed. 2010, 49, 8466.
Weber, D.; Tarselli, M. A.; Gagne, M. R. Angew. Chem. Int. Ed. 2009, 48, 5733.
Simonneau, A.; Jaroschik, F.; Lesage, D.; Karanik, M.; Guillot, R.; Malacria, M.; Tabet, J. C.; Goddard,
J. P.; Fensterbank, L.; Gandon, V.; Gimbert, Y. Chem. Sci. 2011, 2, 2417.
Weber, D.; Jones, T. D.; Adduci, L. L.; Gagne, M. R. Angew. Chem. Int. Ed. 2012, 51, 2452.
Ducháčková, L.; Roithová, J. Chem. Eur. J. 2009, 15, 13399.
(33)
(34)
(35)
(36)
Jašíková, L.; Roithová, J. Organometallics 2012, 31, 1935.
(a) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200. (b) Lee, C.; Yang, W.; Parr, R. G.
Phys. Rev. B 1988, 37, 785. (c) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (d) Stephens, P. J.; Devlin,
F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98, 11623.
Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104.
Frisch M. J. et al., Gaussian 09, Revision A.02, Gaussian, Inc., Wallingford CT, 2009.
Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105, 2999.
Fifen, J. J.; Nsangou, M.; Dhaouadi, Z.; Motapon, O.; Jaidane, N. E. J. Chem. Theory Comput. 2013, 9,
1173.
Wang, D.; Cai, R.; Sharma, S.; Jirak, J.; Thummanapelli, S. K.; Akhmedov, N. G.; Zhang, H.; Liu, X.;
Petersen J. L.; Shi, X. J. Am. Chem. Soc. 2012, 134, 9012.
Ranieri, B.; Escofet, I.; Echavarren, A. M. Org. Biomol. Chem. 2015, 13, 7103.
Zhdanko, A.; Maier, M. E. ACS Catal. 2015, 5, 5994.
Brown, T. J.; Weber, D.; Gagne, M. R.; Widenhoefer, R. A. J. Am. Chem. Soc. 2012, 134, 9134.
Jia, M.; Bandini, M. ACS Catal. 2015, 5 1638.
Biasiolo, L.; Trinchillo, M.; Belanzoni, P.; Belpassi, L.; Busico, V.; Ciancaleoni, G.; D’Amora, A.;
Macchioni, A.; Tarantelli, F.; Zuccaccia, D. Chem. Eur. J. 2014, 20, 14594.
Biasiolo, L.; Del Zotto, A.; Zuccaccia, D. Organometallics 2015, 34, 1759.
Ciancaleoni, G.; Belpassi, L.; Zuccaccia, D.; Tarantelli, F.; Belanzoni, P. ACS Catal. 2015, 5, 803.
(37)
(38)
(39)
(40)
(41)
(42)
(43)
(44)
(45)
(46)
(47)
(48)
18
ACS Paragon Plus Environment

Page 19 of 19
Journal of the American Chemical Society
1
2
3
4
5
TOC Graphics
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
19
ACS Paragon Plus Environment