The Heck reaction under ball-milling conditions

Article abstract of DOI:10.1016/j.jorganchem.2004.06.045

The synthesis of unsaturated unnatural amino acids according to the Heck-Jeffery protocol using a ball-milling procedure is presented. NOE data recorded on the products confirm that the Z-isomer is formed. This type of mechano-chemistry provides an efficient, "solvent free" and reliable method for the synthesis of substituted dehydroalanines. These conditions provide patterns of reactivity complementary to conventional procedures.

Full text of DOI:10.1016/j.jorganchem.2004.06.045

Journal of Organometallic Chemistry 689 (2004) 3778–3781
www.elsevier.com/locate/jorganchem
The Heck reaction under ball-milling conditions
a
Erik Tullberg , Dan Peters , Torbjorn Frejd
b
a,
*
¨
a
Department of Organic and Bioorganic Chemistry 1, Lund University, P.O. Box 124, Lund SE-221 00, Sweden
b
NeuroSearch A/S, 93 Pederstrupvej, DK-2750 Ballerup, Denmark
Received 31 March 2004; accepted 24 June 2004
Available online 6 August 2004
Abstract
The synthesis of unsaturated unnatural amino acids according to the Heck–Jeffery protocol using a ball-milling procedure is pre-
sented. NOE data recorded on the products confirm that the Z-isomer is formed. This type of mechano-chemistry provides an effi-
cient, ‘‘solvent free’’ and reliable method for the synthesis of substituted dehydroalanines. These conditions provide patterns of
reactivity complementary to conventional procedures.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Palladium catalysed Heck-couplings using ball-mill grinding
1. Introduction
dure [10,12]. In order to find a more efficient methodol-
ogy, we decided to investigate non-solvent ball-mill
chemistry. The first investigations in this field were made
in the synthesis of diphenols using oxidative coupling [5]
and in the synthesis of biaryls using Suzuki conditions
[6]. Ball-milling chemistry is of interest because of the
mild conditions under which it operates and it has al-
ready been applied to reactions such as cyclopropana-
tions of fullerenes [7] and the Wittig reaction [8].
The Heck reaction is one of the most useful palladi-
um-catalysed coupling reactions for the formation of
C–C bonds between an aryl halide and an olefin to form
styrenic derivatives. This, together with other palladi-
um-catalysed coupling reactions, have been reviewed
[1,2]. The wide functional group tolerance of this reac-
tion has allowed rapid syntheses to a large variety of
interesting compounds including unsaturated unnatural
amino acids using a procedure developed by Jeffery et al.
[3]. The methodology for Heck–Jeffery couplings in-
volves extensive reaction times in hot DMF followed
by an extractive workup and a chromatographic separa-
tion to remove excess olefin [4]. Finally, the remaining
DMF is usually removed by co-evaporation with tolu-
ene or by lengthy evaporation in vacuo. Hitherto, the
Heck–Jeffery conditions involving tetrabutylammonium
chloride as phase transfer agent have been the preferred
method for coupling of protected amino acrylates,
which behaved sluggishly in the standard Heck proce-
2. Results and discussion
Our results are described and shown in Scheme 1 and
Table 1. The protected amino acrylates used in this
study were selected among those that were the most suc-
cessfully used in previous work [4].
All aryl halides are commercially available except for
1,3,5-triiodobenzene, which was synthesized according
to the literature procedure [9]. Also the amino acrylates
were synthesized according to the literature procedure
[10]. The yields given in Table 1 are the optimum result
based on three separate reactions for each case carried
out in a planetary ball mill.
*
Corresponding author. Fax: +46-46-222-4119.
E-mail address: torbjorn.frejd@orgk1.lu.se (T. Frejd).
0022-328X/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.06.045

E. Tullberg et al. / Journal of Organometallic Chemistry 689 (2004) 3778–3781
3779
tion of the steel balls, certainly create a high pressure in
contact with the walls of the container. Thus, high pres-
sure conditions would be possible to test using an ordi-
nary hydraulic press used for making IR-tablets. The
reaction between iodobenzene and methyl-2-[(tertbut-
oxycarbonyl)-amino] acrylate was chosen as a model
reaction. Indeed, at 200 kg/cm2 pressure for 60 min
using a preheated anvil (80 °C), the expected coupling
product was formed but only in 13% yield. The pressure
in the rotating ball-mill is difficult to estimate and it is
possible that higher pressure than 200 kg/cm2 is neces-
sary to reach higher yields.
We also tested if the reaction mixture would give the
products on heating to 80 °C for 60 min with and with-
out stirring in ordinary test tubes. This resulted in yields
of 33% and 18%, respectively. The reaction temperature
was chosen to fit that generated by the ball mill when
run at full speed for an hour.
Further, the effect of catalyst loading was probed and
it was found that only a 5% yield of coupling product was
obtained in the absence of the catalyst (Pd(OAc)2) but al-
ready at a catalyst loading of 2 mol% the yield reached
73%. The yield was much the same for catalyst concentra-
tions up to 10 mol%. For practical reasons, the reaction
was subsequently run with a catalyst concentration of 5
mol%. The reaction conditions allowed a scale between
0.5 and 5 mmol without loss of yield and quality of the
product, e.g. at 5 mmol scale the yield was 78%.
At the initial phase of this project, a few experiments
were also performed in a preparative microwave oven
but the yields thus obtained were, on average, lower
than those of the corresponding ball-milling experi-
ments. Finally, multi-coupling reactions with 1,3-diiod-
obenzene and 1,3,5-triiodobenzene were attempted but
these aryl halides failed to give any significant amount
of target product and invariably resulted in a complex
mixture of products.
From the model studies, it is apparent that it is not
sufficient to simply heat the reaction mixture in a stirred
or unstirred vessel or to apply static 200 kg/cm2 pressure
to it. It is most probably the combination of pressure,
heat, grinding and stirring that provides the conditions
necessary for the successful outcome of the reaction.
Furthermore, the same pattern of reactivity as observed
in the study of the Suzuki coupling is repeated in the
Heck coupling [6]. Thus, coupling of electron-deficient
aryl or heteroaryl halides, such as 4-iodonitrobenzene
or 2-bromo-6-methylpyridine, which were previously
successfully coupled in DMF solution, was not success-
ful in the ball-milling but electron-rich substrates like
4-iodoanisole and 4-iodoaniline were efficient coupling
partners. Prior protection of the amine functionality
was not necessary and to our knowledge there are no
examples of coupling between protected amino acrylates
and halo aniline derivatives. 4-Iodoanisole has been
used together with 2-acetamidoacrylic acid under
Scheme 1.
Table 1
Optimum yields in the ball-milling Heck–Jeffrey reactions
Arylhalide
Olefin
Yield of product (%)a
Iodobenzene
Iodobenzene
A
B
A
B
B
B
B
A
A
B
B
B
A
A
B
A
A
76
77
Bromobenzene
Bromobenzene
4-Iodoaniline
57
57
88
74
4-Iodoanisole
3-Iodobenzonitrile
3-Bromofurane
2-Iodothiophene
2-Iodothiophene
4-Iodonitrobenzene
2-Bromo-6-methylpyridine
4-Iodophenol
28
13
25
25
No reaction
No reaction
No reaction
No reaction
No reaction
Complex mixture
Complex mixture
Chlorobenzene
Chlorobenzene
1,3-Diiodobenzene
1,3,5-Triidobenzene
a
The reactions were performed and the products were isolated
according to the general procedure.
In our ball-milling experiments we adopted the com-
position of the reaction mixture essentially as given by
Calo et al. [11] in their paper on Heck couplings in mol-
ten tetra butyl ammonium bromide using conventional
laboratory glassware without grinding. Tetra butyl
ammonium halides are solids at room temperature but
in the molten state they act as ionic liquids that the
authors claim to have a stabilizing effect on the catalytic
species i.e. Pd(0). Thus, the following ingredients consti-
tuted the standard reaction mixture: aryl halide, pro-
tected aminoacrylate, Pd(OAc)2, NaCl, NaHCO3,
HCO2Na, and nBu4NCl. Our preliminary studies con-
firmed the importance of the tetra butyl ammonium ha-
lide, which in our case was nBu4NCl, for the success of
the reaction. Failure to add this ingredient resulted in no
reaction whatsoever. Furthermore, the addition of so-
dium formate as a reductant for Pd(II) improved the
yields. In previous ball-milling experiments NaCl was
used as an additive presumably as a grinding aid [5,6].
In our experiments we found that the reaction per-
formed best when no NaCl or a relatively small amount
of NaCl was added; 5 mg/mg aryl halide seemed to be a
good proportion.
Apparently, the ball-milling created conditions that
are difficult to obtain under usual conditions. The rota-

3780
E. Tullberg et al. / Journal of Organometallic Chemistry 689 (2004) 3778–3781
ordinary Heck conditions [13], but this methodology
seems hampered by low reproducibility and a problem-
atic work-up [10,12].
ket. Two similarly loaded vessels were then fastened
securely into place and the cover on the ball mill
was closed. The ball mill was then kept at full speed
and left thus for 60 min. After the vessels had cooled
down, (typically two reactions were run at the same
time to balance the rotor of the ball mill), they were
taken out of the ball mill and carefully opened. The
contents were first scraped out into a beaker and the
vessel, including the balls, were washed with 200 ml
of acetone. The contents of the beaker were poured
into a round-bottomed flask together with 600 mg
of silica and the solvent was removed in vacuo. The
solid residue absorbed onto silica, was placed on a sil-
ica column and eluted with heptane:ethylacetate in
proportions described above approximately 250 ml.
The results are shown in Table 1. Compounds previ-
ously known in the literature were checked by running
Ball-milling conditions are undoubtedly very different
from conventional conditions, therefore we could not
assume that the E, Z-isomeric ratio would automatically
be the same as experiments run under conventional con-
ditions. The products were therefore subjected to NOE
analysis. In all cases we found no effect on the vinylic
proton when running spectra in amide proton saturation
mode, but a NOE effect was observed between the amide
proton and the aromatic ortho protons. This observa-
tion is consistent with the product being in the Z-form
as demonstrated by our group in earlier studies [10].
The planetary ball-mill methodology is cheap, scala-
ble and may be an efficient alternative to its high-tech-
nology counterparts.
1
melting points and H NMR and found to be in good
accordance with the literature data [10]. The following
new compounds were prepared using this description.
3. Experimental
General information: all reagents were used as deliv-
ered without further purification. Inorganic salts were
dried in an oven at 110 °C overnight. Chromatographic
workup was performed on a 24 mm column filled with
3.2. (Z)-3-(4-Amino-phenyl)-2-tert-butoxycarbonylamino-
acrylic acid methyl ester
According to the procedure above: amorphous brown
solid; m.p. 148.3–150.1 °C,1H NMR(CDCl3): d 7.43, 6.64
(AB q, 4H, JAB = 8.4 Hz, Ar–H), 7.27 (s, 1H, HC‚), 6.04
(br, 1H, NH), 3.83 (s, 3H, OCH3), 1.43(s, 9H, C(CH3)3;
13C NMR(CDCl3): d 166.6, 153.3, 147.9, 132.0, 130.9,
123.9, 121.0, 114.5, 80.6, 60.4, 28.1; HRMS (FAB+) m/e
Calc. for C15H20N2O4 292.1423. Found: 292.1419;
IR(KBr) cmÀ1: 3300, 2900, 1699, 1599, 1517. Anal. Calc.
for C15H20N2O4 C, 61.63; H, 6.90; N, 9.58. Found: C,
61.51; H, 7.03; N, 9.49.
˚
25 cm of Matrex 60 A 35–70 mesh silica. The eluents
were heptane:ethylacetate 5:1 for reactions using olefin
A and heptane:ethylacetate 3:2 when using olefin B.
TLC analyses were performed on Merck silica gel 60
coated aluminium plates with heptane:ethyl acetate 1:1
as eluent. Olefins A and B were synthesized according
to the literature procedure [10]. The reactions were per-
formed using a Fritsch Planetary Micro Mill model
‘‘Pulverisette 7’’ housing two stainless steel cups con-
taining eight stainless steel balls each and sealed by a
stainless steel lid fitted with a Teflon gasket. NMR spec-
tra were recorded on a Bruker ARX 300 instrument,
NOE experiments were recorded in d6-DMSO at 80 °C
on a Bruker DRX 500c, 500 MHz instrument. The delay
time was 1 ms, the irradiation time 10 s, the acquisition
time 3 s, the repetition time 13 s and the transmitter ef-
fect 2 Hz. The samples were degassed in an ultrasound
bath for 15 min prior to recording. Melting points were
measured on a Gallenkamp melting point microscope
and are uncorrected, IR spectra were recorded on a
Shimadzu FTIR-8300 instrument and mass spectra were
recorded on a JEOL SX-102 mass spectrometer.
3.3. (Z)-2-tert-Butoxycarbonylamino-3-(4-methoxy-phe-
nyl)-acrylic acid methyl ester
According to the procedure above: amorphous
white solid; m.p. 107.5–108.7 °C,1H NMR(CDCl3): d
7.53, 6.88 (AB q, 4H, JAB = 8.7 Hz, Ar–H), 7.51 (s,
1H, HC‚), 6.17 (br, 1H, NH), 3.83 (s, 3H,
CO2OCH3), 3.82 (s, 3H, ArOCH3) 1.42(s, 9H,
C(CH3)3; 13C NMR(CDCl3): d 166.3, 160.3, 153.0,
131.6, 130.5, 126.5, 122.2, 113.9, 80.8, 55.2, 52.4,
28.1; HRMS (FAB+) m/e Calc. for C16H21NO5
307.1419. Found: 307.1427; IR(KBr) cmÀ1: 3209,
2910, 1705, 1637, 1604, 1514. Anal. Calc. for
C16H21NO5 C, 62.53; H, 6.89; N, 4.56. Found: C,
62.60; H, 7.04; N, 4.41.
3.1. General reaction conditions
To the reaction vessel was added: 1.00 eq. of the
aryl halide, 1.05 eq. of the aminoacrylate, 2.50 eq.
NaHCO3, 0.20 eq. HCO2Na, 1.20 eq. nBu4NCl, 0.05
eq. Pd(OAc)2 and NaCl 5 mg/mg arylhalide. Then 8
steel balls were added to the vessel, which was then
purged with argon and closed with lid and Teflon gas-
3.4. (Z)-2-tert-Butoxycarbonylamino-3-(3-cyano-phe-
nyl)-acrylic acid methyl ester
According to the procedure above: amorphous
white solid; m.p. 102.1–104.0 °C,1H NMR(CDCl3): d

E. Tullberg et al. / Journal of Organometallic Chemistry 689 (2004) 3778–3781
3781
7.79 (s, 1H, Ar–H), 7.69 (d, 1H, JAB = 7.5 Hz, Ar–H),
7.56 (d, 1H, J = 7.7 Hz, Ar–H), 7.46 (t, 1H, J = 7.7
Hz, Ar–H), 7.21 (s, 1H, HC‚), 6.55 (br, 1H, NH),
3.88 (s, 3H, OCH3), 1.40 (s, 9H, C(CH3)3; 13C
NMR(CDCl3): d 165.4, 151.7, 135.9, 133.6, 132.5,
131.8, 129.1, 125.8, 125.5, 118.4, 112.6, 81.6, 52.9,
28.0; HRMS (FAB+) m/e Calc. for C16H18N2O4
302.1266. Found: 302.1272; IR(KBr) cmÀ1: 3306,
2981, 2229, 1693, 1649, 1493. Anal. Calc. for
C16H18N2O4: C, 63.56; H, 6.00; N, 9.27. Found: C,
63.39; H, 5.90; N, 9.16.
3.7.
rylic acid benzyl ester
(Z)-2-tert-Butoxycarbonylamino-3-furan-2-yl-ac-
According to the procedure above: white semi solid,
1H NMR(CDCl3): d 7.72 (s, 1H, furanyl), 7.35–7.43
(m, 5H, J = 6.3 Hz, 1H, thienyl), 6.67 (s, 1H, HC‚),
6.17 (br, 1H, N–H), 5.27 (s, 2H, O–CH2Ph), 1.46 (s,
9H, C(CH3)3; 13C NMR(CDCl3): d 165.2, 153.3, 145.2,
143.6, 135.6, 128.5, 128.3, 128.0, 124.4, 123.3, 120.4,
109.9, 80.8, 67.1, 28.1; HRMS (FAB + Na) m/e Calc.
for C19H21NO5Na 366.1317. Found: 366.1344; IR(-
NaCl) film cmÀ1: 3332, 2977, 2252, 1699, 1504, 1261.
Anal. Calc. for C19H21NO5: C, 66.46; H, 6.16; N, 4.08.
Found: C, 66.50; H, 6.18; N, 4.01.
3.5. (Z)-2-tert-Butoxycarbonylamino-3-thiophen-2-yl-ac-
rylic acid benzyl ester
According to the procedure above: amorphous white
solid; m.p. 121.0–122.0 °C,1H NMR(CDCl3): d 7.77 (br,
1H, HC‚), 7.52 (d, 1H, J = 5.07 Hz, thienyl), 7.34–7.45
(m, 5H, J = 6.5 Hz, Ar–H ), 7.34 (d, 1H, J = 3.25 Hz,
thienyl), 7.09 (t, 1H, H–Ar, J = 4.96 Hz), 5.89 (br, 1H,
NH), 5.29 (s, 2H, OCH2Ph), 1.48 (s, 9H, C(CH3)3; 13C
NMR(CDCl3): d 165.1, 153.6, 136.6, 135.7, 133.0,
130.8, 129.0, 128.5, 128.2, 127.0, 121.6, 81.0, 67.1,
28.2; HRMS (FAB+) m/e Calc. for C19H21NO4S
359.1191. Found: 359.1203; IR(KBr) cm À1: 3324,
2968, 2374, 1713, 1622, 1480. Anal. Calc. for
C19H21NO4S: C, 63.49; H, 5.89; N, 3.90. Found: C,
63.42; H, 6.20; N, 3.57.
Acknowledgements
We gratefully acknowledge the financial support of
the Swedish Research Council, the Knut and Alice Wal-
lenberg Foundation and the Royal Physiographic Soci-
ety in Lund. We also direct our sincere gratitude to
Dr. Karl-Erik Bergqvist who designed and ran all the
NOE-experiments and helped us to interpret them.
References
[1] B.A. Lorsbach, M.J. Kurth, Chem. Rev. 99 (1999) 1549.
[2] M.-S. Schiedel, C.A. Briehn, P. Ba¨uerle, J. Organomet. Chem.
653 (2002) 200.
[3] T. Jeffery, Tetrahedron Lett. 26 (1985) 2667.
[4] A.-S. Carlstro¨m, T. Frejd, Acta Chem. Scand. 46 (1992) 163.
[5] M.O. Rasmussen, O. Axelsson, D. Tanner, Synth. Commun. 27
(1997) 4027.
3.6. (Z)-2-tert-Butoxycarbonylamino-3-thiophen-2-yl-ac-
rylic acid methyl ester
According to the procedure above: yellow paste,1H
NMR(CDCl3): d 7.71 (br, 1H, HC‚), 7.50 (d, 1H,
J = 4.97 Hz, thienyl), 7.32 (d, 1H, J = 3.52 Hz, thie-
nyl), 7.08 (t, 1H, J = 3.87, thienyl), 5.86 (br, 1H, N–
H), 3.84 (s, 3H, OCH3), 1.48 (s, 9H, C(CH3)3; 13C
NMR(CDCl3): d 165.7, 153.5, 136.6, 132.8, 130.6,
129.0, 127.0, 121.6, 81.0, 52.4, 28.2; HRMS (FAB +
Na) m/e Calc. for C13H17NO4SNa 306.0776. Found:
306.0768; IR(KBr) cmÀ1: 3230, 2977, 1701, 1627,
1458, 1365, 1265. Anal. Calc. for C13H17NO4S: C,
55.11; H, 6.05; N, 4.94. Found: C, 55.09; H, 5.94;
N, 4.94.
[6] S.F. Nielsen, D. Peters, O. Axelsson, Synth. Commun. 30 (2000)
3501.
[7] G.-W. Wang, T.-H. Zhang, P. Li, H. Zhan, Y.-C. Liu, Y. Murata,
K. Komatsu, Tetrahedron Lett. 44 (2003) 4407.
[8] V.P. Balema, J.W. Wiench, M. Pruski, V.K. Pecharsky, J. Am.
Chem. Soc. 124 (2002) 6244.
[9] U. Schoeberl, T.F. Magnera, R.M. Harrison, J. Am. Chem. Soc.
119 (1997) 3907.
[10] A.-S. Carlstro¨m, T. Frejd, Synthesis (1989) 414.
[11] V. Calo, A. Nacci, A. Monopoli, L. Lopez, A. di Cosmo,
Tetrahedron 57 (2001) 6071.
[12] P.J. Harrington, L.S. Hegedus, J. Org. Chem. 49 (1984) 2657.
[13] M. Cutolo, V. Fiandanese, F. Naso, O. Sciacovelli, Tetrahedron
Lett. 24 (1983) 4603.