View Online
Green Chemistry
Dynamic Article Links
Cite this: Green Chem., 2011, 13, 2806
PAPER
www.rsc.org/greenchem
Simple sonochemical protocols for fast and reproducible Grignard reactions
Downloaded by University of Oxford on 11 October 2011
Published on 16 August 2011 on http://pubs.rsc.org | doi:10.1039/C1GC15756F
Giancarlo Cravotto,*a Antonio Procopio,b Manuela Oliverio,b Laura Orioa and Diego Carnaroglioa
Received 27th June 2011, Accepted 12th July 2011
DOI: 10.1039/c1gc15756f
The Grignard reaction is one of the most common synthetic pathways in organic chemistry and
has widely been used in the preparation of fine chemicals and pharmaceutical intermediates. The
reaction’s reproducibility can be affected by several factors (metal purity, degree of association in
solution, Schlenk equilibrium) and its activation with chemical, mechanical, sonochemical or
electromagnetic treatments is practically essential. In this study we demonstrate that
ultrasound-assisted Grignard reactions are safe, efficient and reproducible. The activation of the
Mg surface, the single electron transfer (SET) from the metal to organic halide, and the
subsequent nucleophilic addition of ArMgX to carbonyl substrates, fall into the larger domain of
sonochemistry. As expected for a radical mechanism the best sonochemical conditions were
observed at 300 kHz (cup horn). An efficient one-pot protocol with poorly-reactive aryl halides
and the subsequent quenching with a hindered ketone are reported.
Introduction
The reaction discovered by Victor Grignard at the beginning
of the XXth century still plays a relevant role in organic and
organometallic synthetic chemistry. The activation of magnesium turnings for the preparation of organomagnesium halides
(the Grignard reagent) and the subsequent initiation of the
Grignard reaction still makes for a tricky procedure during
which chemists moderately heat iodine, 1,2-dibromomethane or
TMSCl in freshly distilled ethereal solvents1 or toluene2 under
dry argon or nitrogen. Although several investigations have
tried to replace tetrahydrofuran (THF),3 it remains the solvent
of choice because of its high Mg2+ coordinating capacity and
ability to stabilize the products. Practical limitations and safety
concerns have emerged from its use in industrial applications due
to the large amount of heat released during the initiation step in
low-boiling point media. Traces of impurities can react strongly
with the Grignard reagents, and process control under reflux
conditions is difficult to maintain.4 Different metallic magnesium chip sizes and forms have been tested,5 under chemical,5,6
mechanical7 and sonochemical8,9 activation. Hulshof et al. have
studied the influence of microwave (MW) irradiation on metallic
magnesium,10,11 and showed that this method can produce a
surface which is just as accessible as the one obtained via the
use of an initiator. Mass-transport and kinetics are strongly
favored by MW irradiation at these interfaces between solid and
a
Dipartimento di Scienza e Tecnologia del Farmaco, via P. Giuria 9,
10125, Torino, Italy. E-mail: [email protected]
b
Dipartimento di Scienze Farmacobiologiche, Viale Europa, Loc.
Germaneto, 88100, Catanzaro, Italy
2806 | Green Chem., 2011, 13, 2806–2809
liquid, although this interaction generates violent electrostatic
discharge phenomena between the Mg particles (arcing).11
More recently Kappe et al.12 have shown that the preparation
of the same Grignard reagent in different professional MW
instruments (mono- or multimodal) can lead to entirely different
results, even at the same macroscopically determined reaction
temperature. In other words, the authors have demonstrated
that Grignard reagent formation under MW irradiation is either
accelerated or slowed down as a result of field density conditions.
In the search for a simple and better reproducible general
procedure, we have investigated both the reaction of magnesium
with organic halides which leads, via oxidative addition, to the
Grignard reagent and the alkylation of carbonyl derivatives,
in a one-pot sonochemical protocol. High-intensity ultrasound
(US) proved to be effective in metal surface activation.13 Sonochemical reactors, and even simple cleaning baths,14 facilitate the
removal of the oxide film on the metal surface and counteract
the depositing of Grignard reaction products onto the new
surface.8 Despite the leading role of US in metal-assisted
synthetic chemistry (i.e. Barbier reaction),14 this technique has
so far only been considered as an alternative way to activate
the magnesium metal.15 All the studies that have described the
promoting effect of US on the Grignard reaction have only
covered the initiation process and not the whole synthesis which
is usually performed under conductive heating and stirring.8,16
The optimization of a Grignard reaction, from the preparation
of organomagnesium halides to its scaling up for industrial
purposes, is hampered by several factors and intrinsic dangers.
The spontaneous heat release during the initiation step,4 the
induction of very exothermic radical side reactions1 and the
risk of explosive reactant accumulation17 are the main safety
This journal is © The Royal Society of Chemistry 2011
Downloaded by University of Oxford on 11 October 2011
Published on 16 August 2011 on http://pubs.rsc.org | doi:10.1039/C1GC15756F
View Online
problems associated with the reaction. Efficient MW-activation
might unfortunately, on a larger scale, result in violent arcing and
the danger of explosion. Simultaneous US/MW irradiation18 is
a promising innovation in metal-assisted synthesis, although it is
not much safer for this application and in fact cannot overcome
the massive arcing risk.13
Sonication is a milder technique for activating metal surfaces
in organometallic chemistry and has been widely used for Mg
activation in Grignard reagent formation.13–15
Our aim was to design a safe and reproducible general
procedure for Grignard synthesis which could be suited for
scaling up processes.
poorly reactive aryl halides under controlled temperature and
atmosphere (Scheme 2 and Fig. 1) conditions.
Scheme 2 One-pot power US-assisted Grignard formation and reaction with benzophenone.
Results and discussion
In this study, various conditions and US reactors have been
used to prepare several Grignard reagents and to carry out the
quenching of the corresponding adducts with a ketone. After a
preliminary screening of the experimental conditions, we focused
our investigation on four US instruments: a common cleaning
bath (35 kHz), a high power US bath (19.6 kHz) and two cup
horns (20 kHz and 300 kHz).
We started our experiments using 2-bromopyridine (1) as
our model halide, magnesium chips in dry THF and an inert
atmosphere.
The results of the three different non-conventional irradiation
techniques, MW, US and MW/US (Scheme 1), have been
compared.
Scheme 1 Comparison of MW and US irradiation on 2-pyridinyl
magnesium bromide (1a) formation.
At MW experiments, in our hands, substrate 1 showed exactly
the same reaction rate, yield and arcing as had been reported in
the literature under MW (path a, Scheme 1).10 The reaction
under simultaneous MW/US irradiation increased the amount
of electrical discharge caused by violent particle collisions. The
best result was achieved under sonication in a common cleaning
US bath (path b, Scheme 1). Total conversion was obtained in
a few minutes at a milder temperature. Data on US-assisted
Grignard formation had been reported by Tuulmets et al.2,8,16
that observed an accelerating effect on all steps of the reaction,
but the final yield of the Grignard reagent is not any higher
than when conventional heating is used.8 They reported only
moderate yields for bromoalkyl derivatives which were reacted
without any Mg pre-activation step,8 whereas good yields were
recorded for chloroalkyl derivatives when a catalytic amount
of iodine was used to activate Mg turnings before sonication.16
All experiments were performed using a titanium immersion
horn with which an anhydrous atmosphere is difficult achieve.
To the best of our knowledge, no sonochemical procedure for a
whole Grignard reaction has been described in the literature. A
series of one-pot Grignard reactions were carried out on a few
This journal is © The Royal Society of Chemistry 2011
Fig. 1 From the left: a solution of bromobenzene and Mg chips in
THF; the solution after sonication (step 1); the solution after reaction
with benzophenone (step 2).
In the first step, 2-bromopyridine (1), 2-chloropyridine
(2), bromobenzene (3), chlorobenzene (4) and 3,5dimethoxybromobenzene (5), were reacted with 1.5 eq of
Mg chips in dry THF without any pre-activation (step 1,
Scheme 2). The formation of Grignard reagents 1a–5a was
monitored by GC-MS analysis after quenching with a proton
donor. The sonication was performed using four different
devices: namely a common cleaning bath (40 W), a high-power
US bath (200 W) and 20 and 300 kHz cup horns (100 W and
200 W respectively).
The reaction conditions, times and yields for all the substrates
and all the US equipment employed are shown in Table 1
together with the results obtained under MW irradiation for
comparison.
In almost all US-assisted reactions, with the exception of
chlorobenzene (4), complete conversion was observed when
working with a 300 kHz cup horn at mild temperatures (entry 4,
Table 1). The most reactive substrates, such as 2-bromopyridine
(1) and bromobenzene (3), showed equivalent results for US
and MW assisted irradiation and for the various US devices
used (entries 1 to 5, Table 1), giving complete conversion
in all cases. Substrates with moderate reactivity, such as 2chloropyridine (2) and 3,5-dimethoxy-bromobenzene (5), gave
a more interesting proof of the promoting effect of US on
Grignard reagent formation. In these cases, only the instruments
which generate more intensive cavitation energy gave good yields
of Grignard reagents 2a and 5a (entries 3 and 4, Table 1).
The experiments performed using the 20 kHz cup horn, which
allows the reaction temperature to be strictly regulated, showed
a direct yield/temperature dependence (entry 3, Table 1). In
Green Chem., 2011, 13, 2806–2809 | 2807
View Online
Table 1 Power US-assisted Grignard formation (step1, Scheme 1)
1 US cleaning bath
(40 W, 35 kHz) T =
35 ◦ C
Downloaded by University of Oxford on 11 October 2011
Published on 16 August 2011 on http://pubs.rsc.org | doi:10.1039/C1GC15756F
Entry
2 Power US bath
(200 W, 19.6 kHz)
T = 40 ◦ C
3 US Cup Horn
(100 W, 20 kHz) T =
45 ◦ C
4 US Cup Horn
(200 W, 300 kHz)
T = 45 ◦ C
5 Multimode MW
(165 W) T = 65 ◦ C
Time
(min)a
Conv Yield Time
(%)
(%)
(min)a
Conv
(%)
Yield Time
(%)
(min)a
Conv
(%)
Yield
(%)
Time
(min)a
Conv Yield
(%)
(%)
Time
(min)a
Conv Yield
(%)
(%)
20 (5)
100
100
20 (10)
100
100
20 (10)
100
100
20 (7)
100
100
10 (2)
100
100
240
(120)
2
2
285 (45)
1
1
240b
90 (30)
30 (10)d
—
50c
100
—
15
100
60 (30)
100
100
15 (5)
100
71c
25 (10)
100
100
20 (5)
100
100
20 (5)
100
100
20 (5)
100
100
10 (2)
30 (5)
100
100
80c
95d
180
—
—
180
—
—
120e , f , g
1c
<1
180
—
—
60 (5)
2
2
180
—
—
180
—
—
180
30 (5)e
—
65
—
50
10 (5)
100
100
60 (30)
18
17
Total reaction time. Activation time is reported in parentheses. b Reaction carried out at 10 ◦ C. c A significant percentage of Wurtz-tipe side products
were detected. d Reaction carried out in standard conditions (reflux) without MW irradiation. e Reaction carried out at 80 ◦ C. f Reaction carried out
in solvent-free conditions and in the presence of 2.5 eq of Mg. g Reaction carried out directly in the cup horn cavity.
a
fact no conversion was observed at 45 ◦ C, while 100% and 65%
conversions were achieved at 80 ◦ C for 2 and 5 respectively.
Prolonged exposure to US did not improve the reaction yield and
only served to increase the amount of Wurz-type side products
generated from radical coupling or from the reaction of the
initially formed organometallic with more organic halide.
It was interesting to find that the optimal US frequency was
300 kHz which gave complete conversion for both substrates
2 and 5 (entry 4, Table 1) even at low temperature (45 ◦ C).
Only chlorobenzene (4) failed to react; no organomagnesium
halides were detected while a significant amount of Wurz-type
side products was formed (entries 1 to 5, Table 1). The low
reactivity of chlorobenzene in the Grignard formation has been
extensively described in the literature.2,6 When comparing USand MW-assisted reactions it can be seen that reaction times
and yields were comparable. However, US provides the great
advantage of working at milder and safer conditions and avoids
the risk of arcing.
In the second step, a solution of 1.5 eq of benzophenone in
dry THF was added to each of the Grignard reagent mixtures
which had been formed in the previous step and the resulting
system was irradiated in the 300 kHz cup horn. The reaction
2808 | Green Chem., 2011, 13, 2806–2809
yields were determined by GC-MS analysis after work-up (step
2, Scheme 2). All the tertiary alcohols obtained were identified
after chromatographic purification by 1 H and 13 C NMR and
compared with the data reported in the literature.19–21
As shown in Table 2, all the Grignard reagents formed in
the previous step efficiently reacted with benzophenone in a few
minutes giving products 1b, 3b and 5b (entries 1–3, Table 2)
in good-to-excellent yields. These results have confirmed that
the 300 kHz cup horn is the best US device for a fast, safe
and efficient one-pot US-assisted Grignard reaction over both
steps.
Table 2 Power US-assisted Grignard reaction with benzophenone
(step 2, Scheme 2)
Entry
Product
Time (min)
Conv (%)
Yield (%)
1
2
3
1ba
3b
5b
5
10
10
98
100
92
94
98
84
a
Product 1b is obtained under the same reaction conditions by both
2-pyridinyl magnesium chloride (2a) and bromide (1a).
This journal is © The Royal Society of Chemistry 2011
View Online
Downloaded by University of Oxford on 11 October 2011
Published on 16 August 2011 on http://pubs.rsc.org | doi:10.1039/C1GC15756F
Conclusions
In this work we have reported the first comparative study of
Grignard reactions using several physical activation techniques;
high-intensity US using different frequencies and reactors and
MW irradiation.
The reactions were monitored following the formation of
the organomagnesium halides and their subsequent addition
to benzophenone. Under such sonochemical conditions we
optimised yield, reaction times and process reproducibility. The
milder, safer conditions and its ability to be scaled up were the
main advantages of this efficient 300 kHz protocol over the MWassisted procedure.
the corresponding quenching product (i.e. benzene, pyridine,
dimetoxybenzene).
A solution of benzophenone (2.73 g, 15 mmol, 1.5 eq) in
3 mL of dry THF was then added dropwise in 10 min under
sonication (water bath temperature: 55 ◦ C). After cooling, 10 mL
of a NH4 Cl saturated solution was added and the product was
extracted with 5 ¥ 3 mL of EtOAc. The combined organic layers
were dried over Na2 SO4 , and the solvent was removed under
reduced pressure.
Yield and conversion were calculated using GC-MS. Products
were purified using column chromatography and characterized
according to literature procedures.19–21
Acknowledgements
Experimental
Materials
All the chemicals were obtained from commercial sources and
used without further purification. THF was freshly distilled
before each experiment. Reactions were monitored by TLC
on Merck 60 F254 (0.25 mm) plates, which were visualized by
UV inspection and/or staining with 5% H2 SO4 in ethanol and
heating.
MW-promoted reactions were carried out in a professional
oven, Microsynth - Milestone (Italy); this was also used for
combined MW/US irradiation after a probe equipped with a
R horn was inserted into it.
Pyrex
Gas chromatography–mass spectrometry (GC-MS) analyses
were carried out with a gas chromatograph Agilent 6890
(Agilent Technologies - USA) fitted with a mass detector Agilent
Network 5973, using a 30 m long capillary column, i.d of 0.25
mm and film thickness 0.25 mm. NMR spectra were performed
on Bruker 300 Advance spectrometer (300 MHz and 75 MHz
for 1 H and 13 C, respectively).
Description of the sonochemical instruments
The high-power US-bath (19.6 kHz)22 and cup horns for
irradiation at 2023 and 300 kHz23 were made by Danacamerini
(Torino) in collaboration with the corresponding author’s group
and are commercially available. An Elma TS540 cleaner bath
operating at 35 kHz was used to provide some comparison of
sonication effects.
US 300 kHz general procedure
366 mg (15 mmol, 1.5 eq) of Mg chips, 3 mL of dry THF
and 1 mmol (0.1 eq) of aryl halide were added to a dry,
one neck, round bottom flask. The mixture was sonicated
(water bath temperature: 55 ◦ C) under a nitrogen atmosphere
until activation, then another 9 mmol (0.9 eq) of aryl halide
were added and the mixture was sonicated until completion
(reaction monitored by TLC, work up with NH4 Cl/diethyl
ether). For product 3a, bromobenzene was added dropwise to
avoid the formation of Wurz-type side products. A sample of the
solution was quenched with saturated NH4 Cl solution, extracted
with diethyl ether and analyzed by GC-MS. We calculated
the ratio starting material/Grignard reagent on the base of
This journal is © The Royal Society of Chemistry 2011
Financial support from MIUR PRIN 2008 “A Green Approach
to Process Intensification in Organic Synthesis” is gratefully
acknowledged.
References
1 J. F. Garst and M. P. Soriaga, Coord. Chem. Rev., 2004, 248, 623.
2 H. Simuste, D. Panov, A. Tuulmets and B. T. Nguyen, J. Organomet.
Chem., 2005, 690, 3061.
3 R. T. Patterson, Chem. Today, 2000, 18, 14.
4 H. Kryk, G. Hessel, W. Schmitt and N. Tefera, Chem. Eng. Sci., 2007,
62, 5198.
5 U. Tilstam and H. Weinmann, Org. Process Res. Dev., 2002, 6, 906.
6 (a) R. C. Smith, S. Shah and J. D. Protasiewicz, J. Organomet. Chem.,
2002, 646, 255; (b) E. E. Grinberg, V. I. Rakhlin, A. A. Petrova, M.
G. Voronkov, I. E. Strel’nikova and S. G. D’yachkova, Dokl. Chem.,
2007, 412, 22; (c) L. M. Fleury and B. L. Ashfeld, Tetrahedron Lett.,
2010, 51, 2427.
7 K. V. Baker, J. M. Brown, N. Hughes, A. J. Skarnulis and A. Sexton,
J. Org. Chem., 1991, 56, 698.
8 A. Tuulmets, K. Kaubi and K. Heinoja, Ultrason. Sonochem., 1995,
2, S75.
9 (a) J. P. Lorimer and T. J. Mason, Chem. Soc. Rev., 1987, 16, 239;
(b) J. Lindley and T. J. Mason, Chem. Soc. Rev., 1987, 16, 275; (c) B.
H. Han, J. Korean Chem. Soc., 1985, 29, 557.
10 M. H. C. L. Dressen, B. H. P. van de Kruijs, J. Meuldijk, J. A. J. M.
Vekemans and L. A. Hulshof, Org. Process Res. Dev., 2007, 11, 865.
11 H. P. van de Kruijs, M. H. C. L. Dressen, J. Meuldijk, J. A. J. M.
Vekemans and L. A. Hulshof, Org. Biomol. Chem., 2010, 8, 1688.
12 Gutmann, A. M. Schwan, B. Reichart, C. Gspan, F. Hofer and C. O.
Kappe, Angew. Chem., Int. Ed., 2011, DOI: 10.1002/anie.201100856.
13 (a) P. Cintas, G. Palmisano and G. Cravotto, Ultrason. Sonochem.,
2011, 18, 836; (b) J. D. Sprich and G. S. Lewandos, Inorg. Chim. Acta,
1983, 76, L241.
14 (a) J. L. Luche and J. C. Damiano, J. Am. Chem. Soc., 1980, 102, 7926;
(b) P. Cintas, G. Palmisano and G. Cravotto, Ultrason. Sonochem.,
2011, 18, 836.
15 H. Hagiwara and H. Uda, J. Chem. Soc., Chem. Commun., 1988, 815.
16 A. Tuulmets and D. Panov, J. Organomet. Chem., 1999, 575, 182.
17 J. Wiss, M. Lanzlinger and M. Wermuth, Org. Process Res. Dev.,
2005, 9, 365.
18 G. Cravotto and P. Cintas, Chem. Soc. Rev., 2006, 35, 180.
19 P. A. van der Schaaf, R. A. T. M. Abbenhuis, W. P. A. van der Noort,
R. de Graaf, D. M. Grove, W. J. J. Smeets, A. L. Spek and G. van
Koten, Organometallics, 1994, 13, 1433.
20 M. C. C. Ng, D. J. Craig, J. B. Harper, L. van-Eijck and J. A. Stride,
Chem.–Eur. J., 2009, 15, 6569.
21 P. Wang, L. Zhou, X. Zhang and X. Liang, Chem. Commun., 2010,
46, 1514.
22 G. Cravotto, L. Boffa, S. Mantegna, M. Avogadro, P. Perego and P.
Cintas, Ultrason. Sonochem., 2008, 15, 898.
23 G. Cravotto, D. Garella, E. Calcio Gaudino, F. Turci, S. Bertarione,
G. Agostini and F. Cesano, New J. Chem., 2011, 35, 915.
Green Chem., 2011, 13, 2806–2809 | 2809