A Joint Experimental-Computational Comparative Study of the Pd0-Catalysed Reactions of Aryl Iodides and Aldehydes with N, O, and S Tethers

Article abstract of DOI:10.1002/ejoc.201500393

The influence of the heteroatom (nitrogen, oxygen, and sulfur) on the course of the palladium-catalysed intramolecular reactions of aryl iodides and aldehydes having heteroatom-containing tethers has been explored by an extensive experimental-computationa

Full text of DOI:10.1002/ejoc.201500393

FULL PAPER
DOI: 10.1002/ejoc.201500393
A Joint Experimental–Computational Comparative Study of the Pd0-Catalysed
Reactions of Aryl Iodides and Aldehydes with N, O, and S Tethers
Daniel Solé,*[a] Francesco Mariani,[a] and Israel Fernández*[b]
Keywords: Homogeneous catalysis / Palladium / Cyclization / Heterocycles / Density functional calculations
The influence of the heteroatom (nitrogen, oxygen, and
sulfur) on the course of the palladium-catalysed intramolec-
ular reactions of aryl iodides and aldehydes having hetero-
atom-containing tethers has been explored by an extensive
experimental–computational (DFT) study. Two series of sub-
strates were considered, namely aldehydes bearing either
the α-(2-iodobenzylheteroatom) or β-(2-iodophenylhetero-
atom) moieties. While some experimental differences were
observed when changing from nitrogen to oxygen or sulfur
in the 2-iodobenzyl series, the aldehydes in which the
heteroatom is directly bonded to the aromatic ring showed
common chemical behaviour regardless of the nature of the
heteroatom. The different reaction pathways leading to the
experimentally observed reaction products were studied by
computational means. Our calculations suggest that in all
cases the initial nucleophilic addition involving a σ-aryl–PdII
intermediate is preferred over the competing concerted
metallation–deprotonation (CMD) process.
over the possible C(=O)–H activation mechanism was as-
cribed to the high nucleophilicity of the carbon atom di-
rectly attached to the metal, which results from the π-donor
effect of the ortho-nitrogen atom.[4a]
Introduction
The ubiquity of heterocycles has resulted in the organic
chemistry community working continually to develop meth-
ods for their synthesis. In the last decade, the demand for
more efficient syntheses of heterocycles has directed this re-
search towards the development of catalytic processes based
on transition metals.[1] Besides its synthetic applications,
transition-metal chemistry involving heteroatom-containing
compounds is of mechanistic interest due to its unique
characteristics stemming from the heteroatom present in the
substrates. Thus, in recent years, a number of examples have
appeared that highlight the controlling or modifying role of
heteroatoms in transition-metal-promoted reactions.
As part of our ongoing program on the synthesis of
nitrogen heterocycles,[2] we have been studying the palla-
dium-catalysed intramolecular coupling of amine-tethered
aryl iodides and aldehydes. In this context, we recently de-
scribed the palladium-catalysed intramolecular acylation[3]
of (2-iodoanilino) aldehydes (Scheme 1). It has been sug-
gested that this reaction proceeds through the nucleophilic
addition of a readily formed σ-aryl–PdII intermediate to the
carbonyl group.[4,5] The preference for this reaction pathway
Scheme 1. Pd0-catalysed intramolecular acylation of (2-iodoanil-
ino) aldehydes.
We have also recently reported palladium-catalysed reac-
tions of α-(2-iodobenzylamino) aldehydes, in which the nu-
cleophilicity of the corresponding σ-aryl–PdII intermediates
are manifestly diminished. As a consequence, starting from
these substrates, two competitive reaction pathways can be
promoted by Pd0, and the selectively of the process is con-
trolled by the choice of the base (Scheme 2).[6] The use of
Et3N results only in nucleophilic addition of the aryl–PdII
intermediate to the carbonyl group to give a tetrahydroiso-
quinolin-4-ol derivative. In contrast, the use of Cs2CO3 as
the base results in competition between a CO32–-mediated
C–H-bond-activation process and the nucleophilic addition,
which thereby generates mixtures of addition products and
tridentate [C,N,O]–PdII complexes. Our density functional
theory (DFT) calculations showed that as well as the effect
[a] Laboratori de Química Orgànica, Facultat de Farmàcia,
Universitat de Barcelona,
Av. Joan XXIII s/n, 08028 Barcelona, Spain
E-mail: dsole@ub.edu
http://www.ub.edu/farmaco/ca/quimica/llistat_recerca
[b] Departamento de Química Orgánica I, Facultad de Ciencias
Químicas, Centro de Innovación en Química Avanzada
(ORFEO-CINQA), Universidad Complutense de Madrid,
Ciudad Universitaria, 28040 Madrid, Spain
E-mail: israel@quim.ucm.es
http://www.ucm.es/info/biorgmet/espanol/isra.html
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201500393.
Eur. J. Org. Chem. 2015, 3935–3942
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3935

D. Solé, F. Mariani, I. Fernández
FULL PAPER
of the base, coordination of the nitrogen atom to the metal oxygen analogues, respectively, of α-(2-iodobenzylamino)
centre during the reaction is also important for the control aldehydes. For the sake of comparison, a selection of the
of the reaction pathway.[6]
reaction conditions previously explored with the nitrogen-
containing series was used with aldehydes 1 and 2. The re-
sults of these reactions are summarised in Table 1. Treat-
ment of sulfide 1 with an equimolar amount of both Pd0
and ligand in the presence of Cs2CO3 resulted in the decom-
position of the starting material (Table 1, entry 1), while the
use of catalytic amounts of Pd(OAc)2 and dppe [1,2-bis(di-
phenylphosphino)ethane],[9] and still using Cs2CO3 as the
base, exclusively promoted the reduction of the aryl iodide
of sulfide 1 to give 3 (Table 1, entry 2, and Figure 1). In
contrast, the latter reaction conditions with ether 2 led to
the acylated ketone (i.e., 5), which was isolated in 30% yield
(Table 1, entry 5). These results are very different from
those obtained with α-(2-iodobenzylamino) aldehydes,
which, in the presence of Cs2CO3, led mainly to the forma-
tion of stable [C,N,O]–PdII complexes that divert the metal
from the productive cycle, hindering the development of a
catalytic cycle for the acylation reaction (Scheme 2).[6]
Scheme 2. Pd0-promoted reactions of α-(2-iodobenzylamino) alde-
hydes.
The remarkable structure-dependent reactivity of these
nitrogen-containing aldehydes is in sharp contrast to the
reactivity of the carbocyclic series,[7] in which only the acyl-
ation reaction takes place.
Continuing our research in this area, we were interested
in exploring the influence of the heteroatom (changing from
nitrogen to oxygen or sulfur) on the course of the palla-
dium-catalysed reactions of these substrates. We hypoth-
esised that the lesser extent of lone-pair delocalisation when
the heteroatom directly bonded to the aromatic ring is oxy-
gen or sulfur could hinder the acylation reaction. Addition-
ally, we wondered whether the strong coordination ability
of sulfur and the low oxophilicity of palladium could be-
come obstacles to the reactions of aldehydes bearing a 2-
iodobenzyl-heteroatom moiety.
In this paper, we report the results of our studies on the
palladium-catalysed intramolecular coupling reactions of
aryl iodides and aldehydes in substrates with oxygen or
sulfur tethers. The joint experimental–computational study
of these reactions revealed both similarities and some differ-
ences compared to their nitrogen-based counterparts. This
work also opens new synthetic routes to the chromane and
isochromane nuclei, as well as their sulfur analogues
(Scheme 3).[8]
Table 1. Pd0-catalysed coupling reactions of aldehydes 1 and 2.[a]
Entry Ald. [Pd] (mol-%)
ligand (mol-%)
Base
(equiv.)
t [h]
Products
(yield [%])[b]
1
2
1
1
Pd2(dba)3 (55)
PPh3 (120)
Pd(OAc)2 (20)
dppe (25)
Cs2CO3 (3)
Cs2CO3 (3)
24
72
–
3[c]
3
4
1
1
Pd(PPh3)4 (5)
Pd2(dba)3 (5)
PPh3 (11)
Pd(OAc)2 (20)
dppe (25)
Et3N (6)
Et3N (6)
72
40
3 (25), 4 (64)
3/4 (1:2)[d]
5
2
Cs2CO3 (3)
72
5 (30)[e]
6
7
2
2
Pd(PPh3)4 (5)
Pd2(dba)3 (5)
PPh3 (11)
Et3N (6)
Et3N (6)
72
48
5 (31), 6 (42)
6 (42)[f]
8
2
Pd(PPh3)4 (5)
Et3N (6)
72
5 (17), 6 (69)
KOAc (1)
[a] Reaction conditions: [Pd], ligand, and base in toluene at 120 °C
in a sealed tube. [b] Isolated yields. [c] Yield not quantified. [d] Ra-
tio by 1H NMR spectroscopy; yield not quantified. [e] Trace
amounts of the hydrodehalogenation product were also observed
1
in the crude reaction mixture. [f] Ratio by H NMR spectroscopy,
5/6 (1:1.5).
Scheme 3. Synthesis of oxygen and sulfur heterocycles.
Figure 1. Reduced compounds 3 and 11.
Results and Discussion
On the other hand, treatment of sulfide 1 with a catalytic
We began our investigation by focussing on the intra- amount of Pd(PPh3)4 and Et3N as the base promoted the
molecular coupling of sulfide 1 and ether 2, the sulfur and nucleophilic addition of the corresponding σ-aryl–PdII in-
3936
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 3935–3942

Pd0-Catalysed Reactions of Aryl Iodides and Aldehydes
termediate to the carbonyl group to give alcohol 4 in 64% forming the reaction in THF (Table 2, entry 5) or changing
yield (Table 1, entry 3). When submitted to the same reac- the ligand to dppe (Table 2, entry 6) gave poor results.
tion conditions, ether 2 gave a 1:1.4 mixture of ketone 5 Interestingly, the use of the more polar solvent DMF to-
and alcohol 6 (Table 1, entry 6). Similar results were ob- gether with a combination of Cs2CO3 and Et3N as the base
tained using Pd2(dba)3/PPh3 (dba = dibenzylideneacetone) led to alcohol 10 as the major reaction product (Table 2,
as the catalyst (Table 1, entries 4 and 7). Interestingly, when entry 7). Changing the salt from Cs2CO3 to KOAc in-
starting from ether 2, the addition of KOAc to a reaction creased the formation of 10, which was isolated in 42%
run under otherwise identical reaction conditions resulted yield when dtpf was used as the ligand (Table 2, entry 8).
in an increase of the alcohol-to-ketone ratio, thus allowing
6 to be isolated in 69% yield (Table 1, entry 8).
When ether 8 was submitted to the conditions optimised
for the reaction with sulfide 7, acylation was the only reac-
Continuing with our comparative study, we then ex- tion pathway observed, which allowed the preparation of
plored the palladium-catalysed reactions of aldehydes 7 and ketone 12 in 89% yield (Table 2, entry 10). Also, starting
8, in which the heteroatom is directly bonded to the aro- from ether 8 and performing the reaction in DMF in the
matic ring (Table 2). As mentioned above, the nitrogen- presence of KOAc reproduced the behaviour previously ob-
based counterparts of these compounds, i.e., β-(2-iodoanil- served with sulfide 7. Under these conditions, 8 gave
ino) aldehydes, selectively undergo acylation through a alcohol 13 as the major reaction product (Table 2, entry 11).
mechanism involving carbopalladation between the aryl– Changing the ligand to PPh3 resulted in a 1:2.6 alcohol-to-
PdII intermediate and the carbonyl group. The preference ketone ratio (Table 2, entry 12). Finally, alcohol 13 was still
for this process was ascribed to the high nucleophilicity of obtained as the major product when the reactions of 8 were
the carbon atom directly attached to the metal because of carried out using Et3N as the base in toluene, either with
the π-donor effect of the ortho-nitrogen atom.[4a]
or without KOAc (Table 2, entries 13–14).
Under the reaction conditions developed for this intra-
In this context, it should be noted that the palladium-
molecular acylation (Scheme 1), using Cs2CO3/Et3N as the catalysed reaction of 2-iodoaniline 14 either in DMF in the
base in toluene,[4] sulfide 7 underwent the same reaction to presence of KOAc or using Et3N as the base in toluene also
give ketone 9, although significant amounts of the hydrode- led to mixtures of ketone 15 and alcohol 16 (Scheme 4).
halogenation product (i.e., 11; Figure 1) were also invariably Leaving aside minor differences, these results confirm that
isolated (Table 2, entries 1–3). The best results for the acyl- sulfide 7, ether 8, and aniline 14 show the same behaviour
ation of 7 were obtained using dtpf [1,1Ј-bis(di-tert-butyl- in their palladium-catalysed reactions.
phosphino)ferrocene] as the ligand, which gave 9 in 59%
At this point, we decided to explore these processes using
yield (Table 2, entry 3). Other combinations of ligand, base, computational tools[10] in order to understand the effect of
and solvent were also explored. The acylation reaction was the heteroatom on the outcome of the process. The main
still the major process when 7 was treated with Pd(PPh3)4 differences from the amine-tethered series were observed
as the catalyst and K3PO4 as the base, allowing 9 to be when Cs2CO3 was used as the base. Thus, we began our
isolated in 52% yield (Table 2, entry 4). However, per- investigations using density functional theory (DFT) calcu-
Table 2. Pd0-catalysed coupling reactions of aldehydes 7 and 8.[a]
Entry Ald. [Pd] (mol-%)/ligand (mol-%)
Base (equiv.)
Cs2CO3 (3)/Et3N (6) toluene
Pd2(dba)3 (7)/(tBu)3PH·BF4 (14) Cs2CO3 (3)/Et3N (6) toluene
Solvent
t [h]
1H NMR ratio
Products (yield [%])[b]
1
7
7
7
7
7
7
7
7
7
8
8
8
8
8
Pd2(dba)3 (7.5)/(o-tolyl)3P (15)
72
72
72
72
72
72
0.5
0.5
0.5
45
48
60
72
72
9/11 (9:1)
9/11 (2.6:1)
9/11 (8:1)
9/11 (3:1)
9/11 (3.5:1)
–
9/10/11 (2:3:1)
9/10/11 (1:7:1.7)
9/10/11 (1:2.8:1)
–
9 (43)
2
3
4
9 (35)
Pd2(dba)3 (7.5)/dtpf (15)
Pd(PPh3)4 (10)
Cs2CO3 (3)/Et3N (6) toluene
9 (59)
K3PO4 (3)
K3PO4 (3)
Cs2CO3 (3)
toluene
THF
9 (52), 11 (15)
9 (36)
5
6
7
8
Pd2(dba)3 (7.5)/dtpf (15)
Pd(OAc)2 (10)/dppe (25)
Pd(OAc)2 (10)/dtpf (11)
Pd(OAc)2 (10)/dtpf (11)
Pd(OAc)2 (10)/(o-tolyl)3P (11)
Pd2(dba)3 (7.5)/dtpf (15)
Pd(OAc)2 (10)/dtpf (11)
Pd2(dba)3 (5)/PPh3 (11)
Pd2(dba)3 (5)/PPh3 (11)
Pd2(dba)3 (5)/PPh3 (11)
[c]
toluene
–
Cs2CO3 (1)/Et3N (3) DMF[d]
9 (20), 10 (30)
KOAc (1)/Et3N (3)
KOAc (1)/Et3N (3)
DMF[d]
10 (42)
[e]
9
DMF[d]
–
10
11
12
13
14
Cs2CO3 (3)/Et3N (6) toluene
12 (89)
KOAc (1)/Et3N (3)
KOAc (1)/Et3N (3)
KOAc (1)/Et3N (3)
Et3N (3)
DMF
DMF
toluene
toluene
–
12 (20), 13 (55)
[e]
12/13 (1:2.6)
12/13 (1:3.8)
12/13 (1:1.3)
–
13 (51)
[e]
–
[a] Reaction conditions: [Pd], ligand, base, and solvent at 120 °C in a sealed tube. [b] Isolated yields. [c] Decomposition. [d] The reaction
was carried out in a sealed vessel in a microwave reactor (fixed temperature and variable pressure). [e] Yields not quantified.
Eur. J. Org. Chem. 2015, 3935–3942
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3937

D. Solé, F. Mariani, I. Fernández
FULL PAPER
Our calculations indicate that the reaction profiles in-
volving the different heteroatoms (X = N, O, S) are quite
similar. Of course, differences in the calculated relative free
energies are found as a consequence of the distinct donor/
coordination abilities of the different heteroatoms. How-
ever, it is clear that in each case the concerted metallation–
deprotonation (CMD) process, mediated by carbonate and
leading to C–H activation product INT2, has quite a high
activation barrier (ΔG‡ = 39.9, 38.1, and 37.6 kcal/mol, for
X = S, NMe, and O, respectively). In contrast, the nucleo-
Scheme 4. Reagents and conditions: (a) Pd2(dba)3 (0.05 equiv.), philic addition reaction, which begins with an initial X to
PPh3 (0.1 equiv.), Et3N (6 equiv.), toluene, 120 °C, 52 h, 17% (15),
C=O ligand exchange via TS2, proceeds with a much lower
32% (16); (b) Pd(OAc)2 (0.1 equiv.), dtpf (0.11 equiv.), KOAc
energetic cost. As a consequence, starting from ether 2, the
(1 equiv.), Et3N (3 equiv.), DMF, 120 °C, 52 h, 45% (15), 20% (16).
formation of ketone 5 via INT2 by a reductive elimination
reaction[6] seems clearly unfeasible. Moreover, the forma-
tion of the corresponding tridentate [C,X,O]–PdII complex
from INT2 is not expected for X = O, and only products
derived from the nucleophilic addition should be found ex-
perimentally. This computational prediction is in nice agree-
ment with the experimental findings, as ketone 5 was iso-
lated in the process involving ether 2 when Cs2CO3 was
used as the base (Table 1, entry 5). This reaction product
derives therefore from a nucleophilic addition reaction
(which leads to alkoxide–palladium intermediate INT4 via
TS3, with an activation barrier of 12.5 kcal/mol) and subse-
quent β-elimination reaction via TS4 (ΔG‡ = 5.2 kcal/mol).
Our calculations also show that the reactions involving
lations by focussing on the palladium-catalysed reactions of
sulfide 1 and ether 2 in the presence of this base. As pre-
viously reported by us,[6] the observed ketones (i.e., 5) may
derive from two alternative reaction pathways, namely C–H
activation vs. nucleophilic attack followed by β-elimination.
Figure 2 shows the calculated reaction profiles for the com-
petitive nucleophilic addition and C–H activation processes
starting from INT1, the intermediate formed upon the ini-
tial oxidative addition of aryl iodides 1 and 2 and subse-
quent iodide ligand displacement by CO32–. For compari-
son, the calculated profile already reported for the analo-
gous α-(2-iodobenzylamino) aldehyde[6] is also shown in
Figure 2, and the corresponding free energies (ΔG393, at sulfide 1 have reaction profiles much more similar to those
393.15 K) in toluene as solvent are also shown. of its nitrogen counterpart. However, only reduction prod-
Figure 2. Calculated reaction profiles starting from INT1-X (X = NMe, O, and S). Relative free energy values (ΔG393, calculated at
120 °C) are given in kcal/mol. All data were calculated at the PCM(toluene)-B3LYP/def2-TZVP//B3LYP/def2-SVP level.
3938
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 3935–3942

Pd0-Catalysed Reactions of Aryl Iodides and Aldehydes
uct 3 was isolated experimentally (Table 1, entry 2). This to the corresponding isoquinolone.[6] Our calculations indi-
suggests that the reduction process[11] from intermediates cate that a similar coordination from the sulfur atom in the
INT1 or INT3 or even from the initially formed aryl-palla- related INT6-S also occurs after the corresponding nucleo-
dium intermediate, which has a Pd-iodide bond, is faster philic addition step (Figure 3, a). Therefore, alcohol 4
than the nucleophilic addition reaction. This is compatible should be exclusively formed from INT6-S by a well-known
with the high calculated free-energy difference between the Et3N-mediated reductive protonation process,[2a,12] as was
corresponding transition state (i.e., TS3) and the initial observed experimentally.
complex (i.e., INT1) (ΔG = 33.3 kcal/mol; see Figure 2).
In contrast, the nucleophilic addition step from oxygen
In addition, we also calculated the reaction profiles derivative INT5-O leads to the formation of INT6-O, which
shown in Figure 2 at the PCM(toluene)-B3LYP-D3/def2- has the cisoid conformation required for the β-elimination
TZVP//B3LYP/def2-SVP level to account for dispersion ef- step to form ketone 5 (via TS5-O, calculated activation bar-
fects. From the data in Figure S1 (see Supporting Infor- rier of only 0.2 kcal/mol, see Figure 3, b). Interestingly,
mation), it is evident that the inclusion of dispersion in the INT6-O is in equilibrium with INT6-O-B, an intermediate
calculations does not modify the scenario described above, that resembles INT6-S, with Pd–O coordination and lead-
i.e., the nucleophilic addition/β-elimination pathway is pre- ing to alcohol 6. The coexistence of these two intermediates
ferred over the demanding C–H activation pathway (acti- (free-energy difference of only 2.4 kcal/mol) is responsible
vation barriers ca. 38 kcal/mol).
for the observed mixture of reaction products, and has its
On the other hand, when Et3N is used as base instead origin in the low coordination ability of the oxygen atom
of Cs2CO3, sulfide 1 gives alcohol 4 (Table 1, entries 3–4), as compared to nitrogen or sulfur. Indeed, this is reflected
whereas ether 2 leads to a mixture of alcohol 6 and ketone in the much lower calculated Pd–X Wiberg bond order in
5 (Table 1, entries 6–7). Thus, the behaviour of the sulfide INT6-O-B as compared to INT6-S (0.12 vs. 0.43 au, for X
resembles, once again, that of its nitrogen counterpart, = O and S, respectively), and in the lower exergonicity cal-
which exclusively produces the corresponding isoquinolin- culated for the INT5-XǞINT6-X transformation (–7.4 and
4-ol derivative when the reaction is conducted in the pres- –15.1 kcal/mol for X = O and S, respectively).
ence of Et3N.[6] This was ascribed to coordination of the
DFT calculations were also carried out to explore the
nitrogen atom to the palladium, which hinders the cisoid effect of the heteroatom in this series of compounds in
conformation required for the β-elimination step that leads which the heteroatom is directly attached to the ortho posi-
Figure 3. Calculated reaction profiles starting from INT5-S (a) and INT5-O (b). Relative free energy values (ΔG393, calculated at 120 °C)
and bond lengths are given in kcal/mol and Å, respectively. All data were calculated at the PCM(toluene)-B3LYP/def2-TZVP//B3LYP/
def2-SVP level.
Eur. J. Org. Chem. 2015, 3935–3942
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3939

D. Solé, F. Mariani, I. Fernández
FULL PAPER
tion of the aryl group. Again, the reaction profiles shown ural bond orbital (NBO) method, the SOPT energy associ-
in Figure 4 reveal the possible competitive reaction path- ated with such a delocalisation in INT10-X varies as ΔE(2)
ways for the conversion of INT7 (the species formed upon = –33.2 Ͼ –21.5 Ͼ –14.6 kcal/mol, for X = NMe, O, and
initial oxidative addition of 7, 8, and the analogous aniline S, respectively. Thus, it parallels the calculated trend in acti-
derivative) into the experimentally observed ketones (i.e., 9, vation barriers.
12, and 9N), respectively.
The influence of the heteroatom is further reflected in
Our calculations indicate that the formation of the re- the HOMO–LUMO interaction. As shown in Figure 5, the
spective ketones do not arise from the carbonate-mediated HOMO of INT10-N can be viewed as a π molecular orbital
(CMD) process that leads to C–H activation product INT9. that involves both the LP of the nitrogen atom and the
This is again due to the quite high activation barrier calcu- carbon atom attached to palladium. The LUMO can be
lated for this transformation, which involves transition considered to be the π*(C=O) orbital involving the coordi-
states TS6 (ΔG‡ = 38.8, 35.4, and 31.9 kcal/mol, for X = O, nated carbonyl group.
S, and NMe, respectively). Therefore, chromanone 9, thio-
Similar molecular orbitals were calculated for INT10-O
chromanone 12, and dihydroquinolinone 9N result again (and INT10-S), with the exception that in these species the
from a nucleophilic addition reaction followed by a β-elimi- HOMO of INT10-N corresponds to the HOMO–1. Despite
nation step, which proceed with a comparatively much that, the HOMO–LUMO gap in the nitrogen derivative is
lower energetic cost.
significantly reduced (3.90 eV) as compared to the
Interestingly, the influence of the heteroatom is signifi- (HOMO–1)–LUMO gap calculated for its oxygen counter-
cant in the initial intramolecular nucleophilic addition reac- part (4.26 eV). As a result of a more favourable HOMO–
tion, as confirmed by the different calculated activation LUMO interaction, a lower reaction barrier should be ex-
barriers involving TS8 (ΔG‡ = 2.0, 5.6, and 7.7 kcal/mol, pected for INT10-N, which agrees with the values calculated
for X = NMe, O, and S, respectively). This can be ascribed for the nucleophilic addition reaction.
to the extent of the delocalisation of the corresponding het-
Interestingly, both sulfide 7 and ether 8 give the corre-
eroatom lone-pair (LP) into the aryl moiety, which controls sponding alcohols (i.e., 10 and 13, respectively) in addition
the relative nucleophilicity of the carbon atom directly at- to the ketones (i.e., 9 and 12, respectively) when the reaction
tached to the transition metal in INT10. Indeed, according is conducted in DMF as solvent (Table 2, entries 7–9 and
to the second-order perturbation theory (SOPT) of the nat- 11–12). Our calculations indicate that the activation barrier
Figure 4. Calculated reaction profiles starting from INT7-X (X = NMe, O, and S). Relative free energy values (ΔG393, calculated at
120 °C) are given in kcal/mol. All data were calculated at the PCM(toluene)-B3LYP/def2-TZVP//B3LYP/def2-SVP level.
3940
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 3935–3942

Pd0-Catalysed Reactions of Aryl Iodides and Aldehydes
corded using Me4Si as the internal standard, with a Varian Gemini
300 or Varian Mercury 400 instrument. 1H and 13C chemical shifts
are reported in ppm downfield (δ) from Me4Si. TLC was carried
out on SiO2 (silica gel 60 F254, Merck), and the spots were located
with UV light or KMnO4 (1% aq.). Flash chromatography was
carried out on SiO2 (silica gel 60, SDS, 230–400 mesh ASTM). Un-
less otherwise noted, all experiments were carried out under an
argon atmosphere. During the work-up of reactions, organic ex-
tracts were dried with anhydrous Na2SO4.
Typical Method for Pd0-Catalysed Reactions: A mixture of aldehyde
1 (70 mg, 0.22 mmol), Et3N (0.18 mL, 1.32 mmol), and Pd(PPh3)4
(13 mg, 0.011 mmol) in toluene (7 mL) was stirred at 120 °C in a
sealed tube for 72 h. The reaction mixture was partitioned between
saturated aqueous NaHCO3 and Et2O. The organic extracts were
washed with brine, dried, and concentrated. The residue was puri-
fied by chromatography (hexanes to hexanes/EtOAc, 4:1) to give 3
(10 mg, 25%) and 4 (27 mg, 64%).
Figure 5. Calculated molecular orbitals involved in the intramolec-
ular nucleophilic addition reaction of INT10. Orbital energies are
given in eV. All data were calculated at the B3LYP/def2-SVP level.
Computational Details: All the calculations reported in this paper
were carried out with the GAUSSIAN 09 suite of programs.[13]
Electron correlation was partially taken into account using the
hybrid functional usually denoted as B3LYP[14] using the double-ζ
quality plus the polarisation def2-SVP basis set[15] for all atoms.
Reactants and products were characterised by frequency calcula-
tions,[16] and have positive definite Hessian matrices. Transition
structures (TS’s) show only one negative eigenvalue in their diago-
nalised force constant matrices, and it was confirmed that their
associated eigenvectors correspond to motion along the reaction
coordinate under consideration by using the Intrinsic Reaction Co-
ordinate (IRC) method.[17] Solvent effects were taken into account
by using the polarisable continuum model (PCM).[18] Single-point
calculations (PCM-B3LYP/def2-TZVP) on the gas-phase-op-
timised geometries were performed to estimate the change in the
Gibbs energies (at 393.15 K) in the presence of toluene or DMF as
solvent using the triple-ζ quality plus the polarisation def2-TZVP
basis set.[15] This level is denoted PCM-B3LYP/def2-TZVP//
B3LYP/def2-SVP. For the reaction profiles shown in Figure 2, dis-
persion effects were taken into account using the D3-correction
described by Grimme and coworkers.[19]
of the β-elimination step leading to ketones 9 and 12 in-
creases in the solvent DMF (9.2 kcal/mol for INT11-S, and
8.9 kcal/mol for INT11-O) compared to toluene (4.9 kcal/
mol for INT11-S, and 8.4 kcal/mol for INT11-O). Ad-
ditionally, the highly solvating/coordinating ability of the
DMF (as compared to toluene) would lead to a more facile
transmetallation of the PdII alkoxide INT11 to give the cor-
responding cesium or potassium alkoxide, which, after pro-
tonolysis during the work-up, would give the alcohol.[2a] As
a consequence of both of these effects, significant amounts
of the corresponding alcohols are observed experimentally
(see Table 2).
Conclusions
This joint experimental–computational comparative
study has clarified the influence of the heteroatom (chang-
ing from nitrogen to oxygen and sulfur) on the course of
the palladium-catalysed intramolecular reactions of aryl
iodides and aldehydes tethered by heteroatom-containing
chains. Two alternative reaction pathways can be envisaged
to explain the formation of the observed ketone derivatives
when the reaction is conducted in the presence of Cs2CO3,
namely C–H activation vs. nucleophilic attack followed by
β-elimination. Although some differences in the calculated
relative free energies were found as a consequence of either
the distinct coordination ability of the heteroatom or the
delocalisation of the heteroatom lone-pair into the aryl
fragment, our calculations indicate that, in all cases, the nu-
cleophilic addition reaction followed by β-elimination pro-
ceeds with a much lower energetic cost than the concerted
metallation–deprotonation (CMD) process. Therefore, it
can be concluded that the nature of the heteroatom in the
initial aryl iodide is not decisive for the outcome of the
reaction.
Wiberg bond indices (WBIs) and donor–acceptor interactions were
calculated using the natural bond orbital (NBO) method.[20] The
energies associated with these two-electron interactions were calcu-
lated according to the equation shown below.
φ and φ* are two filled and unfilled natural bond orbitals having
energies εφ and εφ*, respectively; nφ stands for the occupation
number of the filled orbital.
Acknowledgments
The authors gratefully acknowledge financial support for this work
from the Spanish Ministerio de Economía y Competitividad,
Fondos Europeos para el Desarrollo Regional (MINECO-
FEDER) (grant numbers CTQ2012-31391, CTQ2013-44303-P, and
Red de Excelencia Consolider CTQ2014-51912-REDC).
Experimental Section
[1] For recent reviews, see: a) J. J. Li, G. W. Gribble, Palladium in
Heterocyclic Chemistry: A Guide for the Synthetic Chemist, Per-
gamon, Amsterdam, 2000; b) E. Negishi (Ed.), Handbook of
General Remarks: All commercially sourced reagents were used
without further purification. 1H and 13C NMR spectra were re-
Eur. J. Org. Chem. 2015, 3935–3942
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3941

D. Solé, F. Mariani, I. Fernández
See the section Computational Details of this article.
FULL PAPER
Organopalladium Chemistry for Organic Synthesis Wiley-VCH, [10]
New York, 2002, vol. I and II; c) J. Tsuji (Ed.), Palladium in
Organic Synthesis (Topics in Organometallic Chemistry),
Springer-Verlag, Berlin, 2005; d) G. Zeni, R. C. Larock, Chem.
Rev. 2006, 106, 4644; e) I. V. Seregin, V. Gevorgyan, Chem. Soc.
Rev. 2007, 36, 1173; f) F. Bellina, R. Rossi, Tetrahedron 2009,
65, 10269; g) M. Platon, R. Amardeil, L. Djakovitch, J.-C.
Hierso, Chem. Soc. Rev. 2012, 41, 3929; h) B. Alcaide, P. Al-
mendros, C. Aragoncillos, Chem. Soc. Rev. 2014, 43, 3106; see
also: i) E. Soriano, I. Fernández, Chem. Soc. Rev. 2014, 43,
3041, and references cited therein.
[11]
The formation of hydrodehalogenation compounds has been
reported in many different Pd0-catalysed reactions starting
from aryl halides. The exact mechanism of this reduction, as
well as the identity of the reducing agent could be diverse. See,
for example: a) K. H. Shaughnessy, B. C. Hamann, J. F. Hart-
wig, J. Org. Chem. 1998, 63, 6546; b) L. G. Quan, M. Lamrani,
Y. Yamamoto, J. Am. Chem. Soc. 2000, 122, 4827; c) K.
Högenauer, J. Mulzer, Org. Lett. 2001, 3, 1495; d) A. A. Plet-
nev, R. C. Larock, J. Org. Chem. 2002, 67, 9428.
[12]
[13]
C. C. Lu, J. C. Peters, J. Am. Chem. Soc. 2004, 126, 15818.
[2] a) D. Solé, L. Vallverdú, X. Solans, M. Font-Bardia, J. Bon-
joch, J. Am. Chem. Soc. 2003, 125, 1587; b) D. Solé, X. Ur-
baneja, J. Bonjoch, Tetrahedron Lett. 2004, 45, 3131; c) D. Solé,
O. Serrano, J. Org. Chem. 2008, 73, 2476; d) D. Solé, O. Ser-
rano, J. Org. Chem. 2010, 75, 6267; e) D. Solé, M.-L. Bennasar,
I. Jiménez, Org. Biomol. Chem. 2011, 9, 4535; f) D. Solé, I.
Fernández, M. A. Sierra, Chem. Eur. J. 2012, 18, 6950; g) D.
Solé, I. Fernández, Acc. Chem. Res. 2014, 47, 168.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starov-
erov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M.
Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Ad-
amo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Mar-
tin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,
revision D.01, Gaussian, Inc., Wallingford CT, 2009.
[3] For the Pd0-catalysed intermolecular acylation of aryl brom-
ides with aldehydes, see: P. Nareddy, C. Mazet, Chem. Asian J.
2013, 8, 2579.
[4] a) D. Solé, F. Mariani, I. Fernández, M. A. Sierra, J. Org.
Chem. 2012, 77, 10272; b) D. Solé, F. Mariani, J. Org. Chem.
2013, 78, 8136. See also ref.[2g]
[5] For the Pd0-catalysed intramolecular acylation with esters and
amides, see: a) D. Solé, O. Serrano, Angew. Chem. Int. Ed.
2007, 46, 7270; Angew. Chem. 2007, 119, 7408; b) D. Solé, O.
Serrano, J. Org. Chem. 2008, 73, 9372.
[14]
[15]
a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648; b) C. Lee, W.
Yang, R. G. Parr, Phys. Rev. B 1998, 37, 785; c) S. H. Vosko,
L. Wilk, M. Nusair, Can. J. Phys. 1980, 58, 1200.
F. Weigend, R. Alhrichs, Phys. Chem. Chem. Phys. 2005, 7,
3297.
[6] D. Solé, F. Mariani, I. Fernández, Adv. Synth. Catal. 2014, 356,
3237.
[7] a) P. Álvarez-Bercedo, A. Flores-Gaspar, A. Correa, R. Martin,
J. Am. Chem. Soc. 2010, 132, 466; b) A. Flores-Gaspar, A.
Gutiérrez-Bonet, R. Martin, Org. Lett. 2012, 14, 5234.
[8] For recent examples of the metal-catalysed synthesis of these
heterocyclic systems, see: a) P. W. Davies, S. J.-C. Albrecht, An-
[16] J. W. McIver, A. K. Komornicki, J. Am. Chem. Soc. 1972, 94,
2625.
gew. Chem. Int. Ed. 2009, 48, 8372; Angew. Chem. 2009, 121, [17] C. González, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523.
8522; b) J. Zhang, L. Wei, Chem. Commun. 2012, 48, 2636; c)
Q. Liu, K. Wen, Z. Zhang, Z. Wu, Y. J. Zhang, W. Zhang,
Tetrahedron 2012, 68, 5209; d) K. R. Prasad, C. Nagaraju, Org.
Lett. 2013, 15, 2778; e) E. Coutant, P. C. Young, G. Barker, A.-
L. Lee, Beilstein J. Org. Chem. 2013, 9, 1797; f) T.-S. Zhu, J.-P.
Chen, M.-H. Xu, Chem. Eur. J. 2013, 19, 865; g) S. E. Am-
mann, G. T. Rice, M. C. White, J. Am. Chem. Soc. 2014, 136,
10834; h) H. Murase, K. Senda, M. Senoo, T. Hata, H. Urabe,
Chem. Eur. J. 2014, 20, 317.
[18] a) S. Miertusˇ, E. Scrocco, J. Tomasi, Chem. Phys. 1981, 55, 117;
b) J. L. Pascual-Ahuir, E. Silla, I. Tuñón, J. Comput. Chem.
1994, 15, 1127; c) V. Barone, M. Cossi, J. Phys. Chem. A 1998,
102, 1995.
[19] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys.
2010, 132, 154104.
[20] a) J. P. Foster, F. Weinhold, J. Am. Chem. Soc. 1980, 102, 7211;
b) A. E. Reed, F. Weinhold, J. Chem. Phys. 1985, 83, 1736; c)
A. E. Reed, R. B. Weinstock, F. Weinhold, J. Chem. Phys. 1985,
83, 735; d) A. E. Reed, L. A. Curtiss, F. Weinhold, Chem. Rev.
1988, 88, 899.
[9] It is well known that Pd(OAc)2 is reduced to Pd0 in the pres-
ence of phosphines. See, for example: C. Amatore, L. El Kaïm,
L. Grimaud, A. Jutand, A. Meignié, G. Romanov, Eur. J. Org.
Chem. 2014, 4709.
Received: August 24, 2014
Published Online: May 19, 2015
3942
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 3935–3942