Controlling reactivity through liquid assisted grinding: The curious case of mechanochemical fluorination

Article abstract of DOI:10.1039/c6gc03139k

We have identified an example of a mechanochemically milled organic reaction where liquid-assisted grinding controls the selectivity, such a phenomenon has not been reported/observed before. It was found that upon milling dibenzoylmethane with Selectfluor in the absence of any solvent, a 3:1 ratio of monofluorinated:difluorinated product was observed. Whereas, addition of 0.125 mL of acetonitrile (～10% of the total volume of materials present) to the ground reaction mixture afforded 50:1 selectivity. Furthermore, this phenomenon is applicable to a small range of diketone substrates thus far explored. Additionally, we have demonstrated that difluorination can be achieved by simply switching from adding acetonitrile to addition of sodium carbonate. Most notable, in the latter case, is the reduced reaction time compared to a conventional solvent approach, 2 hours in the mill and 24 hours in the flask.

Full text of DOI:10.1039/c6gc03139k

Green Chemistry
View Article Online
View Journal
PAPER
Controlling reactivity through liquid assisted
grinding: the curious case of mechanochemical
Cite this: DOI: 10.1039/c6gc03139k
fluorination†
Joseph L. Howard, Yerbol Sagatov, Laura Repusseau, Christiane Schotten and
Duncan L. Browne*
We have identified an example of a mechanochemically milled organic reaction where liquid-assisted
grinding controls the selectivity, such a phenomenon has not been reported/observed before. It was
found that upon milling dibenzoylmethane with Selectfluor in the absence of any solvent, a 3 : 1 ratio of
monofluorinated : difluorinated product was observed. Whereas, addition of 0.125 mL of acetonitrile
(
∼10% of the total volume of materials present) to the ground reaction mixture afforded 50 : 1 selectivity.
Received 14th November 2016,
Accepted 8th December 2016
Furthermore, this phenomenon is applicable to a small range of diketone substrates thus far explored.
Additionally, we have demonstrated that difluorination can be achieved by simply switching from adding
acetonitrile to addition of sodium carbonate. Most notable, in the latter case, is the reduced reaction time
compared to a conventional solvent approach, 2 hours in the mill and 24 hours in the flask.
DOI: 10.1039/c6gc03139k
www.rsc.org/greenchem
Mechanochemical milling methods represent an attractive
process for the preparation of chemical products from a sus-
1
tainability perspective. The concept of running reactions in
the neat phase without solvent waste is irrefutably an impor-
tant pursuit for a more sustainable future. Indeed, recent
Table 1 Optimization of conditions for the selective mono or difluori-
nation of dibenzoylmethane under mechanochemical conditions
2
a
advances in the area of metal organic frameworks have demon-
strated that such processes can be scaled to the manufacture
3
level using a twin screw extrusion apparatus. Such equipment
is already present in many industrial manufacturing plants for
formulation, where reliable processing of powdered materials
is necessary to meet regulatory demands. Most notably, this
chemical processing tool has been found to vastly increase
space time yields for MOF preparation. Still, solid-state
milling, as applied to synthetic organic chemistry, is a rela-
tively underexplored area given the potential gains that could
be made against the economy and environment. In large part,
we believe that the slow uptake of this technology is attribu-
table to both a lack of understanding of the potential enabling
attributes for organic synthesis and a poor understanding of
how to optimize reactions (with respect to yield) through
b
b
Selectfluor
(equiv.)
Time
(h)
2
3
Entry
Additive
[%]
[%]
1
2
3
4
5
6
7
8
9
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
1
0.5
1
2
2
2
2
2
2
2
2
2
2
2
—
—
—
53%
4%
11%
4%
0%
7%
38%
0%
3%
2%
0%
87%
53%
79%
91%
61%
0%
MeCN (0.25 mL)
MeCN (0.25 mL)
—
H O (0.25 mL)
2
i-PrOH (0.25 mL)
PhMe (0.25 mL)
9%
1
c,4
30%
20%
100%
6%
2%
2%
control of the operating parameters.
Liquid assisted grind-
10
2
CH Cl
2
(0.25 mL)
ing (LAG) represents one such phenomenon whereby the
addition of small amounts of liquid can have a profound effect 12
on the outcome of a milled reaction. Recently it was reported
11
MeCN (0.125 mL)
Na CO (1 equiv.)
0%
2
3
94%
87%
68%
19%
13
14
15
K
Cs
CaCO
2
CO
3
(1 equiv.)
(1 equiv.)
(1 equiv.)
2
CO
3
3
53%
School of Chemistry, Main Building, Park Place, Cardiff University, Cardiff,
CF10 3AT, UK. E-mail: dlbrowne@cardiff.ac.uk
a
Dibenzoylmethane (1 mmol), Retsch MM400, 10 mL stainless steel
b
milling jars with one 10 mm (4 gram) stainless steel ball. Determined
by F NMR with trifluorotoluene as internal standard, remaining
mass balance is recovered starting material.
1
9
†
Electronic supplementary information (ESI) available: Detailed experimental
procedures and characterization data. See DOI: 10.1039/c6gc03139k
This journal is © The Royal Society of Chemistry 2016
Green Chem.

View Article Online
Paper
Green Chemistry
that both the quantity and nature of added liquid can result in entries 8–10). Interestingly, added water seemingly resulted in
5
the switching between polymorphs. Herein we describe a a complete inhibition of the reaction, although notably
related observation concerning LAG for a synthetic fluorina- Selectfluor is known to not be degraded by water (Table 1,
9
tion reaction. Fluorinated molecules are in ever increasing entry 7). Reducing the amount of added acetonitrile from
demand due to their ability to dramatically enhance the pro- 0.250 mL to 0.125 mL (∼10% of the total volume of all
6
,7
perties of materials.
materials in the milling jar) resulted in further improvements
Given the prevalence of the carbon–fluorine bond in a to 100 : 0 (i.e. no difluorination). To put these results into per-
range of important chemicals we were keen to investigate spective the solvent based reaction was also performed (in a
whether such a bond can be formed using solid-state milling similar fashion to Banks and co-workers), but with two equiva-
1
0
techniques. Initial investigations commenced by treating solid lents of Selectfluor to make for a more accurate comparison.
dibenzoylmethane (1) with one equivalent of Selectfluor under It was found that this method requires 3.5 hours to reach
mechanochemical mixer mill conditions. The reactions were completion and afforded excellent selectivity (Fig. 1). Also
performed on a 1 mmol scale in 10 mL stainless steel jars with explored was the opportunity of proceeding directly to the
one stainless steel ball (10 mm, 4.0 g) and with a frequency of difluorinated compound by simply adding base to the milled
3
0 Hz applied. Temperature is a difficult variable to control reaction. Addition of one equivalent of sodium carbonate pro-
under milling conditions, but in all cases described here the vided an isolated yield of 90% of a 9 : 1 mixture of di- to mono-
jars were cool enough to handle immediately following the fluorinated products, again within 2 hours (Table 1, entry 12).
end of the grinding period, thus suggesting that the jar itself Other carbonate bases proved less effective to mediate this
did not exceed 50 °C.
transformation.
Pleasingly, after milling for 1 hour with equal reagent stoi-
When compared to the analogous solvent based reaction,
chiometries, a 53% yield of the monofluorinated product (2) the most notable observation was the reduced reaction time
was obtained (Table 1, entry 1) with 4% of the difluorinated afforded by the mechanochemical technique, 24 hours against
material (3) also present. The remaining mass balance was 2 hours (Fig. 1).
confirmed to be starting material (1). Initially increasing the
In order to further probe these observations and see if they
equivalents of Selectfluor resulted in 87% of mono- and 11% are more generally applicable, the monofluorination con-
difluorinated product (Table 1, entry 2). Addition of a liquid, ditions with and without LAG for a small range of other
so called liquid assisted grinding or LAG, appeared to slow 1,3-diketones were assessed (Fig. 2). It was found, in all cases
down the reaction, but most importantly, increased the selecti- examined, that addition of 0.125 mL of acetonitrile to the
8
vity towards the monofluorinated product (Table 1, entry 4).
solid reagents provided superior selectivity ratios to those
Pushing the LAG reaction conditions further by increasing the without added acetonitrile, with all other variables remaining
time to two hours resulted in complete consumption of start- constant. In a similar manner, the generality of the obser-
ing material with a 91 : 7 ratio of mono- to difluorinated pro- vation that mechanochemical milling results in a reduced reac-
ducts (Table 1, entry 5).
tion time for the difluorination was also explored across the
Neat milling under the same conditions resulted in inferior same range of substrates (Fig. 3). Indeed, this was found to be
selectivity of 3 : 2, indicating that LAG is enabling improved true: milling this reaction results in significantly faster conver-
selectivity. The nature of the liquid used for liquid assisted sion to the desired product for all cases explored. Having
grinding was also explored, with isopropanol, toluene and di- assessed the reactivity of 1,3-diketones with Selectfluor under
chloromethane all providing vastly inferior results (Table 1, mechanochemical conditions we turned to β-ketoesters which
Fig. 1 Summary of key mechanochemical observations and comparison to solvent based method. Isolated yields reported.
Green Chem.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Green Chemistry
Paper
Fig. 2 Scope of the enhanced selectivity effect afforded by liquid
assisted grinding of mechanochemical monofluorination reactions of
1
,3-diketones.
Fig. 4 Summary of key mechanochemical observations and comparison
to solvent based method for liquid substrates. Isolated yields reported.
a
Fig. 3 Scope of the reduced reaction time effect afforded by solid-state
grinding of mechanochemical difluorination reactions of 1,3-diketones.
NaCl used as a grinding agent/auxiliary/adsorbent for liquid reactants.
The amount used is equal to twice that of the total of all other reactants.
1
0
are both less reactive and liquid substrates. Milling of liquid/ of LAG and compared to the traditional solvent based round
solid mixtures can result in the gumming of the reactor vessel bottom flask reaction (Fig. 4). Most striking is the reduction in
and therefore poor mass transfer. Additionally, low boiling reaction time that is afforded by the solid-state milling
liquids can vaporize and release from the milling jars if appro- approach (Fig. 4). In this example the solvent based reaction
priate care is not taken. To account for this, it is common to occurs over 120 hours (5 days) whereas the milled reaction,
use an auxiliary material such as silica, alumina, talc or in- with added acetonitrile, requires only two hours to run to com-
1
1
organic salts as a milling agent/auxiliary or adsorbent. In pletion! In the absence of (added) LAG the reaction mixture
synthetic organic reactions some such materials could be con- returned 67% recovered starting material and 32% yield of the
sidered far from innocent (such as the base used in Fig. 3!). In monofluorination product. Difluorination of the liquid ethyl
pharmaceutical formulation science these materials are benzoylacetate was also possible within the same timeframe
termed ‘glidants’ or ‘lubricants’ and assist in the uniform by adding sodium carbonate to the milled reaction mixture
1
2
passage of powdered materials through screw extruders. On (in the absence of LAG), contrasting against the 5 day solvent-
exploring the mechanochemical electrophilic fluorination of based reaction (Fig. 4). With this newly found capability we
ethyl benzoylacetate, it was found that sodium chloride was a finally explored the difluorination of 2,2,6,6-tetramethyl-
suitable material to permit adequate mass transfer and satis- 3,5-heptanedione,
a liquid substrate whose intermediate
factory results. Milling of the liquid ethyl benzoylacetate with monofluorinated compound features a highly stereocongested
Selectfluor and NaCl was explored in the presence and absence enolate.
This journal is © The Royal Society of Chemistry 2016
Green Chem.

View Article Online
Paper
Green Chemistry
In this instance the solvent based reaction conditions were
found to require 9 days to proceed to 69% conversion (as
measured by F NMR spectroscopy with an internal standard),
whereas the mechanochemically milled reaction provided 59%
conversion in just two hours (Fig. 4).
In conclusion, carbon–fluorine bond formation is possible
under solid-state mechanochemical milling conditions.
However, more significant is the observation that liquid
assisted grinding can be used to favor the formation of one
reaction product over another and can give rise to improved
selectivity. The precise effects of LAG on organic reactions are
(b) B. Rodríguez, A. Bruckmann, T. Rantanen and C. Bolm,
Adv. Synth. Catal., 2007, 349, 2213; (c) A. Stolle, T. Szuppa,
S. E. S. Leonhardt and B. Ondruschka, Chem. Soc. Rev.,
2011, 40, 2317; (d) G.-W. Wang, Chem. Soc. Rev., 2013, 42,
7668; (e) D. Braga, L. Maini and F. Grepioni, Chem. Soc.
Rev., 2013, 42, 7638; (f) E. Boldyreva, Chem. Soc. Rev., 2013,
42, 7719; (g) G. N. Hermann, P. Becker and C. Bolm, Angew.
Chem., Int. Ed., 2016, 55, 3781; (h) S. Lou, Y. Mao, D.-Q. Xu,
J. He, Q. Chen and Z. Xu, ACS Catal., 2016, 6, 3890;
(i) Y. Zhou, F. Guo, C. E. Hughes, D. L. Browne,
T. R. Peskett and K. D. M. Harris, Cryst. Growth Des., 2015,
15, 2901; ( j) N. R. Rightmire, D. L. Bruns, T. P. Hanusa and
W. W. Brennessel, Organometallics, 2016, 35, 1698.
2 (a) Ball Milling Towards Green Synthesis, ed. B. Ranu and A.
Stolle, Royal Society Of Chemistry, Cambridge, U.K., 2014;
(b) A. Sarkar, S. Santra, S. K. Kundu, A. Hajra,
G. V. Zyryanov, O. N. Chupakhin, V. N. Charushin and
A. Majee, Green Chem., 2016, 18, 4475; (c) R. A. Haley,
A. R. Zellner, J. A. Krause, H. Guan and J. Mack, ACS
Sustainable Chem. Eng., 2016, 4, 2464; (d) T. X. Métro,
J. Bonnamour, T. Reidon, J. Sarpoulet, J. Martinez and
F. Lamaty, Chem. Commun., 2012, 48, 11781; (e) T. X. Métro,
J. Bonnamour, T. Reidon, A. Duprez, J. Sarpoulet,
J. Martinez and F. Lamaty, Chem. – Eur. J., 2015, 21, 12787;
(f) E. Colacino, P. Nun, F. M. Colacino, J. Martinez and
F. Lamaty, Tetrahedron, 2008, 64, 5569.
1
9
8
,13
poorly characterized
and the exact rationale for the
1
4
observed selectivity in this reaction remains unclear. Our
current hypothesis is that it derives from changes in the crys-
talline form of the mono-fluorinated product, with such forms
only accessible in the presence of added acetonitrile. These
forms may be meta-stable nano-crystals as described by
Belenguer, Hunter, Sanders and co-workers. Indeed, Jones
and co-workers have described how different quantities of
added LAG can lead to different polymorphs to that of a neat
reaction. The latter point is a significant one as it implies
that to understand such a phenomenon will require expertise
in solid-state chemistry, organic reaction mechanism and
mechano- or tribo-chemical methods. We believe the reaction
manifold described herein is an ideal tool for such a study as
5
a
5
b
1
9
the reaction is clean, can be monitored by F NMR and ceases
to continue as soon as the product mixture is triturated away
from the insoluble Selectfluor material. In addition, we have
also demonstrated that mechanochemical milling can vastly
reduce reaction times with little effect on yield and selectivity.
This has been achieved with comparisons run in our labora-
tories with very closely related reaction conditions. This effect
is applicable across both solid and liquid reagents, as long as
the appropriate grinding agent is used.
3 D. Crawford, J. Casaban, R. Haydon, N. Giri, T. McNally
and S. L. James, Chem. Sci., 2015, 6, 1645.
4 (a) R. Trotzki, M. M. Hoffmann and B. Ondruschka, Green
Chem., 2008, 10, 767; (b) G. Kaupp, CrystEngComm, 2011,
13, 3108; (c) T. Friščić, I. Halasz, P. J. Beldon,
A. M. Belenguer, F. Adams, S. J. Kimber, V. Honkimäki and
R. E. Dinnebier, Nat. Chem., 2013, 5, 66; (d) R. Schmidt,
C. F. Burmeister, M. Balaz, A. Kwade and A. Stolle, Org.
Process Res. Dev., 2015, 19, 427.
5
(a) A. M. Belenguer, G. I. Lampronti, A. J. Cruz-Cabeza,
C. A. Hunter and J. K. M. Sanders, Chem. Sci., 2016, 7,
Acknowledgements
6617; (b) D. Hasa, E. Miniussi and W. Jones, Cryst. Growth
D. L. B. thanks Cambridge Reactor Design for a Ph.D. award to
J. L. H., the Bolashak International Scholarship of the
President of the Republic of Kazakhstan for a Scholarship
award to Y. S., the Erasmus programme for support of
L. R. and the School of Chemistry at Cardiff University for
generous support. We thank the EPSRC UK National Mass
Spectrometry Facility at Swansea University for mass spectro-
metry measurements.
Des., 2016, 16, 4582.
6 (a) B. E. Smart, J. Fluorine Chem., 2001, 109, 3; (b) R. Filler
and R. Saha, Future Med. Chem., 2009, 1, 777;
(c) M. Morgenthaler, E. Schweizer, A. Hoffmann-Röder,
F. Benini, R. E. Martin, G. Jaeschke, B. Wagner, H. Fischer,
S. Bendels, D. Zimmerli, J. Schneider, F. Diederich,
M. Kansy and K. Muller, ChemMedChem, 2007, 2, 1100;
(d) D. B. Berkowitz and M. Bose, J. Fluorine Chem., 2001,
112, 13; (e) D. J. O’Hagan, J. Fluorine Chem., 2010, 131,
1071.
7
For reviews about organo-fluorine methods see:
a) J. Wang, M. Sánchez-Roselló, J. L. Aceña, C. del Pozo,
Notes and references
(
1
For reviews about mechanochemistry and selected recent
examples see: (a) S. L. James, C. J. Adams, C. Bolm,
D. Braga, P. Collier, T. Friščić, F. Grepioni, K. D. M. Harris,
G. Hyett, W. Jones, A. Krebs, J. Mack, L. Maini, A. G. Orpen,
I. P. Parkin, W. C. Shearouse, J. W. Steed and
D. C. Waddell, Chem. Soc. Rev., 2012, 41, 413;
A. E. Sorochinsky, S. Fustero, V. A. Soloshonok and H. Liu,
Chem. Rev., 2014, 114, 2432; (b) T. Furuya, C. A. Kuttruff
and T. Ritter, Curr. Opin. Drug Discovery Dev., 2008, 11, 803;
(c) I. Hyohdoh, N. Furuichi, T. Aoki, Y. Itezono, H. Shirai,
S. Ozawa, F. Watanabe, M. Matsushita, M. Sakaitani,
P. S. Ho, K. Takanashi, N. Harada, Y. Tomii, K. Yoshinari,
Green Chem.
This journal is © The Royal Society of Chemistry 2016

View Article Online
Green Chemistry
Paper
K. Ori, M. Tabo, Y. Aoki, N. Shimma and H. Iikura, ACS
Med. Chem. Lett., 2013, 4, 1059; (d) C. N. Neumann and
9 G. Stavber, M. Zupan, M. Jereb and S. Stavber, Org. Lett.,
2004, 6, 4973.
T. Ritter, Angew. Chem., Int. Ed., 2015, 54, 3216; 10 R. E. Banks, N. J. Lawrence and A. L. Popplewell, J. Chem.
e) J.-P. Bégué and D. Bonnet-Delpon, J. Fluorine Chem., Soc., Chem. Commun., 1994, 3, 343.
006, 127, 992; (f) K. L. Kirk, Org. Process Res. Dev., 2008, 11 (a) R. Schmidt, R. Thorwirth, T. Szuppa, A. Stolle,
(
2
1
2, 305; (g) T. Liang, C. N. Neumann and T. Ritter, Angew.
B. Ondruschka and H. Hopf, Chem. – Eur. J., 2011, 17,
8129; (b) J. L. Do, C. Mottillo, D. Tan, V. Štrukil and
T. Friščić, J. Am. Chem. Soc., 2015, 137, 2476;
(c) R. Thorwirth, A. Stolle and B. Ondruschka, Green Chem.,
2010, 12, 985.
Chem., Int. Ed., 2013, 52, 8214; (h) J. A. Ma and D. Cahard,
Chem. Rev., 2004, 104, 6119; (i) D. L. Browne and
P. Richardson, Fluorination Approaches, in Synthetic
Methods in Drug Discovery, Royal Society of Chemistry,
Cambridge, U.K., 2016, vol. 2, p. 263.
12 J. Li and Y. Wu, Lubricants, 2014, 2, 21.
8
(a) T. Friščić, Chem. Soc. Rev., 2012, 41, 3493; 13 D. Tan, C. Mottillo, A. D. Katsenis, V. Štrukil and T. Friščić,
b) A. V. Trask, N. Shan, W. D. Motherwell, W. Jones, Angew. Chem., Int. Ed., 2014, 53, 9321.
S. Feng, R. B. Tan and K. J. Carpenter, Chem. Commun., 14 However, we have further probed this effect across different
(
2
005, 7, 880; (c) T. Friščić, S. L. Childs, S. A. A. Rizvi and
time points and have demonstrated that in this case LAG is
W. Jones, CrystEngComm, 2009, 11, 418.
causing improved selectivity, see ESI† for details.
This journal is © The Royal Society of Chemistry 2016
Green Chem.