Color-coded ligands: Tracking the catalyst using highly pigmented porphyrazine ligands in biphasic reactions

Article abstract of DOI:10.1021/om501094p

We demonstrate the concept that highly colored imidazolium porphyrazine ligands and complexes may be useful for visualizing the location of a catalyst and/or ligand. This was analytically demonstrated by the application of UV-vis spectrophotometry to detect the ligand and ICP-OES spectroscopy for quantification of the Pd in the biphasic systems. In the first instance, a toluene/water system was used, in which the complex and ligand preferred the organic phase. Water-soluble substrates were made to react under these conditions by employing the complexes as catalysts in Heck and Suzuki C-C bond forming reactions. In the second instance, an ionic liquid containing biphasic mixture was used, in which the ligand/catalyst was retained in the ionic liquid. Catalyst recycling experiments were met with limited success.

Full text of DOI:10.1021/om501094p

Communication
pubs.acs.org/Organometallics
Color-Coded Ligands: Tracking the Catalyst using Highly Pigmented
Porphyrazine Ligands in Biphasic Reactions
†
,†,‡
Mmboneni Tshivhase and D. Bradley G. Williams*
†Research Centre for Synthesis and Catalysis, Department of Chemistry, University of Johannesburg, P.O. Box 524, Auckland Park,
2
006 Johannesburg, South Africa
‡
Ferrier Research Institute, Victoria University of Wellington, 69 Gracefield Road, Lower Hutt 5010, New Zealand
S Supporting Information
*
ABSTRACT: We demonstrate the concept that highly colored imidazo-
lium porphyrazine ligands and complexes may be useful for visualizing the
location of a catalyst and/or ligand. This was analytically demonstrated by
the application of UV−vis spectrophotometry to detect the ligand and ICP-
OES spectroscopy for quantification of the Pd in the biphasic systems. In
the first instance, a toluene/water system was used, in which the complex
and ligand preferred the organic phase. Water-soluble substrates were made
to react under these conditions by employing the complexes as catalysts in
Heck and Suzuki C−C bond forming reactions. In the second instance, an ionic liquid containing biphasic mixture was used, in
which the ligand/catalyst was retained in the ionic liquid. Catalyst recycling experiments were met with limited success.
1
0
iphasic catalysis is an important approach to homogeneous
catalysis and allows the reagents or products to be readily
complexes are stable and excellent catalysts. While most
palladium biscarbene complexes contain two identical ligands or
are symmetrical when the carbenes are linked together, some
B
separated from the catalyst. At least two large-scale processes rely
1
11
upon a biphasic setup. Quite a variety of biphasic systems have
mixed carbene−phosphine ligands are known. Typically,
2
3
been used, including organic/water, organic/ionic liquid, ionic
precursor imidazolium halides readily react with a metal source
such as palladium acetate to yield the palladium biscarbene
complex, and a wide range of such complexes has been previously
4
5
liquid/water, and fluorous biphasic systems. While the catalyst
is typically held in one or the other phase, we recently developed
phase-switchable catalysts (organic/water) where cooperative
10
prepared. Herein we detail the synthesis of imidazolium-
containing porphyrazines, their palladium complexes, and the
use thereof in biphasic catalyst systems. We cover organic/water
and ionic liquid/water scenarios in the catalysis reactions with a
focus on Suzuki C−C bond-forming chemistry. This reaction
previously has been the subject of attention when using N-
heterocyclic carbene complexes. We also detail our initial
efforts at recycling these catalysts.
To synthesize our imidazolium ligands, 4,5-diaminophthaloni-
6
ligand effects have been identified. Each potential solution to the
problem of recovery of homogeneous catalysts has its benefits,
and each suffers its drawbacks, but one universal problem is the
ability to track the movement of the ligand and/or catalyst with
ease. To address this issue, we proposed to use a ligand that
would be visible at low concentrations, which would serve as a
proxy by which to track the movement/retention of a catalyst in
biphasic systems. To be successful, the ligand would have to be
vividly colored and possess a high extinction coefficient.
Porphyrin-type compounds represent a class of structures with
very high extinction coefficients (log ε values of 4 or higher are
12
trile (1) was allowed to react with formic acid to produce the
13
benzimidazole 2 in 65% yield (Scheme 1). Monoalkylation of
the product was effected by treatment thereof with NaH and 1-
iodopentane or 1-bromoisobutane in DMF, giving products 3 in
acceptable yields. Alkylation was accompanied by a significant
upfield shift in the resonance associated with the imidazole H
atom, from 8.71 ppm in the parent compound 2 to around 7.85
ppm for the monoalkylated derivatives 3. Reaction of 3 with neat
-chloropentane under reflux afforded alkylpentylimidazolium
chlorides 4 in high yield. Linstead macrocyclization of 3 with 2,3-
dipropylmaleonitrile in n-BuOH in the presence of Mg-
OBu)2 followed by treatment with acetic acid to remove the
7
8
common), and they are readily manipulated to bind metals.
This implies that it is possible to detect such materials at low
levels by making use of simple visual techniques or by applying a
spectroscopic technique such as UV−vis spectroscopy. The data
generated also may be used to quantify the amount of pigment
present in a sample, if the extinction coefficient is known. If the
spectroscopic data are coupled with a method such as ICP-MS by
which the metal can be tracked, quantitative data can be
generated as to the amount of ligand in each phase and the
amount of metal in each phase. In this way, movement of the
ligand (as a proxy for the catalyst) can be qualitatively assessed
visually and quantitatively measured spectroscopically.
1
7a
7,8
(
core-encapsulated Mg ion (Scheme 1), gave porphyrazine
derivatives 5. The yields of the porphyrazines are consistent
7
,8
Palladium complexes embody an important class of catalysts
for many useful organic reactions. Palladium biscarbene
Received: October 30, 2014
Published: December 1, 2014
9
©
2014 American Chemical Society
7023
dx.doi.org/10.1021/om501094p | Organometallics 2014, 33, 7023−7026

Organometallics
Communication
Scheme 1. Synthesis of Imidazolium Ligands
Furthermore, there was little difference in terms of intensity and
absorption maxima when comparing the imidazole- and
imidazolium-derivatized porphyrazines and their Pd complexes
(
Figure 1 and Figure S1 in Supporting Information), and the
Figure 1. UV−vis spectra of 5a and 6a.
7
,8
with the literature. Quaternization of the imidazole moieties
was achieved by reaction of 5 with neat excess 1-chloropentane
under reflux, affording imidazolium derivatives 6 in acceptable
yields.
In addition, bisimidazole 8 was prepared by reaction of 4,5-
dichlorophthalonitrile (7) with imidazole in DMF in the
presence of potassium carbonate (Scheme 2). Very simple
recorded extinction coefficients were consistently above log ε = 4
(
see the experimental details in the Supporting Information).
This was useful and nicely served the purposes of the study.
Organic homogeneous reactions were carried out to measure
the performance of the catalyst. The phthalonitrile precursors of
the porphyrazines were also used for comparison. This set of
ligands would allow comparison of the influence of the alkyl
substitution on the outcome of the reaction. Also, each
phthalonitrile could be compared with its derivative porphyr-
azine, which could give pointers as to the influence of the more
remote functionality (nitriles or macrocyclic rings) on the
outcomes of the reactions. Pd catalysts of ligands 4 and 6 were
Scheme 2. Synthesis of Bisimidazolium Ligands
14
prepared in situ, in accordance with the literature, while
bisimidazolium species 9 and 12 were converted into the Pd
complexes 10 and 13 in advance. The Suzuki C−C bond-forming
reaction transformation was investigated, making use of aryl
bromide substrates with phenylboronic acid (Table 1), under
conditions that were selected to match those previously
employed for catalysts generated from simple imidazolium
14
ligands. Accordingly, the reagents were heated in 19/1 DMF/
H O in the presence of the catalyst and potassium carbonate as
2
base. For the bisimidazolium-derived systems, we also employed
15
conditions previously reported for similar catalysts. Here the
reactions were performed in toluene using potassium carbonate
as base at 120 °C for 24 h. For Tables 1 and 2, “ligandless” Pd
salts were used to provide qualitative benchmark reactions only.
The results obtained for the Suzuki reaction (Table 1) were
NMR spectra (see the experimental details in the Supporting
Information) confirm the anticipated symmetry of the
compound. The porphyrazine analogue of 8 was prepared in a
fashion similar to that described above, i.e. by Mg-templated
macrocyclization of 8 with 2,3-dipropylmaleonitrile, to produce
the free base porphyrazine 11 after removal of the core-
encapsulated Mg ion with acetic acid. Successful conversion of
bisimidazoles 8 and 11 into their bisimidazolium derivatives 9
and 12, respectively, was achieved by reaction of each precursor
molecule with an excess of 1-bromobutane in the presence of n-
1
4,15
consistent with those for similar catalysts,
and little
difference was noted between the simple ligands and their
porphyrazine derivatives. The substrates bearing electron-
withdrawing groups, producing 15−17, showed higher activity
14
than the electron-rich substrate, producing 14, as expected.
Ortho substitution on the aryl bromide was generally tolerated
across the series of catalysts, and all performed well at 0.1 and
0.01 mol % catalyst loadings. These reactions confirmed the
overall activity of the Pd complexes as catalysts for the Suzuki
reactions. In particular, they demonstrated good activity even
under low catalyst loadings. That there is little change in outcome
when moving to higher or lower catalyst loading has been
Bu NI in toluene under reflux. The Pd complexes were readily
4
prepared from these imidazolium products by reaction thereof
with Pd(OAc) in dimethyl sulfoxide at 120 °C for 2 h, followed
2
by removing the solvent under vacuum and washing the residue
with ether to provide complexes 10 and 13, respectively.
When they were investigated by UV−vis spectrophotometry,
the porphyrazines and their quaternized derivatives displayed
spectra consistent with unsymmetrical porphyrin-type mole-
cules. These spectra were characterized by a strong Soret (B
16
previously noted and may be related to the aggregation of
palladium and/or to the formation of catalytically inactive Pd
17
black or colloids at higher concentrations.
In the biphasic reactions, the solvent would be replaced with
the hydrophobic ionic liquid N-octylpyridinium bistriflimide
([OPy][BTI]) possessing a noncoordinating counterion for the
7
,8
band) absorption and by split Q bands in the visible region.
7
024
dx.doi.org/10.1021/om501094p | Organometallics 2014, 33, 7023−7026

Organometallics
Communication
Table 1. Suzuki Reactions using Imidazolium Salts 4a,b and
Table 2. Suzuki Reactions with and without Ionic Liquid using
4-Iodobenzoic Acid as Substrate
a
6
a,b
[
OPy][BTI]/
2
yield (%)
H O yield
toluene/H O
2
entry amt of cat. (mol %) ligand or catalyst
14
15
16
17
(%)
yield (%)
a
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
0.1
PdCl2
41
61
55
53
41
43
59
49
42
33
27
58
32
31
62
52
66
92
82
92
94
62
88
70
80
83
52
68
64
52
54
63
56
78
96
69
98
55
70
80
56
78
62
79
87
64
75
88
49
74
88
51
82
51
63
56
43
70
62
79
87
52
66
70
toluene yield (%),
cat.
cycle 80 °C reflux 80 °C reflux
reflux
0.1
4a
b
0.1
4b
Pd(OAc)2
1
2
3
1
2
3
1
2
3
37
21
c
48
c
32
22
c
54
18
c
52
68
64
0.1
6a
0.1
6b
c
a
0.01
0.01
0.01
0.01
0.01
0.1
PdCl2
10
13
55
32
16
52
27
c
63
c
68
36
18
65
39
c
75
39
c
4a
4b
c
6a
61
32
c
74
48
21
0
1
2
3
4
5
6
6b
b
Pd(OAc)2
a
0.1
10
Conditions: catalyst (Pd(OAc) or 10 or 13, 0.1 mol %), PhB(OH)
(11 mmol), 4-iodobenzoic acid (10 mmol), total volume 11 mL made
up of 1/10 ionic liquid/H O or toluene/H O or neat toluene.
2
2
0.1
13
0.01
0.01
0.01
Pd(OAc)2
2
2
b
c
10
13
Pd(OAc) , no added ligand. Pd black observed, little or no
2
conversion of substrate to product.
a
Conditions for entries 1−10: PdCl and ligand (2.0 equiv) as shown
2
(
entries 1 and 6, no added ligand), aryl halide (1.0 mmol),
phenylboronic acid (1.2 mmol), K CO (2.0 equiv), DMF (10 mL),
2
3
b
H O (0.5 mL), 110 °C. Conditions for entries 11−16: Pd(OAc) or
catalyst as shown (entries 11 and 14, no added ligand), aryl halide
2
2
(
10.0 mmol), phenylboronic acid (11.0 mmol), K CO (20 equiv),
2 3
toluene (10 mL), 120 °C.
Pd on the one hand, and toluene on the other. The second part of
the biphasic system would be water in both cases. The reactions
were performed at 80 °C and under reflux, for comparison
purposes (Table 2). The two layers were vigorously stirred to
improve contact between catalyst and substrates (4-iodobenzoic
acid and phenylboronic acid). At the end of the reaction the
mixture separated into two layers; the aqueous layer contained
the product and the other (organic or ionic liquid) the catalyst,
Figure 2. Toluene/water mixture (left) and ionic liquid/water mixture
(right).
recycling; however, the stability of the complexes needs to be
addressed to maintain activity, which is lost with each cycle.
To assess the leaching/retention of the ligand and Pd catalyst
in each phase, analyses were performed on the various phases of
the biphasic mixtures. For these analyses, UV−vis spectropho-
tometry as well as inductively coupled plasma optical emission
spectroscopy (ICP-OES) were performed. These two techniques
are complementary, since the former assesses the amount of
ligand present while the latter tests for Pd. Both are sensitive
methods and should allow good correlation of the results.
When performing reactions in the ionic liquid/water and
toluene/water mixtures with catalyst 13, the ligand (catalyst)
could be visually located as a blue solution in one of the phases
allowing easy isolation of the product from the catalyst. The
separation of the ligand (catalyst, present at 9 × 10−
4
M
concentration; approximately 97 ppm of Pd for catalysts 10 and
3) into one of the layers (ionic liquid or toluene) is evident from
1
Figure 2. The ionic liquid was found to be stable under these
nonforcing conditions. To recover the reaction products, the
aqueous layer was lyophilized and reconstituted in anhydrous
DMF and the product methylated under the influence of methyl
iodide in the presence of a base (Cs CO ). The methyl ester
2
3
product was isolated from this reaction mixture.
(
Figure 2). Reactions were set up as follows:
The results show that the catalysts performed at approximately
the same levels in the biphasic system as they did in the
homogeneous toluene solvent (Table 2). This was a good result,
since biphasic systems often suffer from sluggish rates in
1
2
3
. reaction in toluene/water
. as in 1 with no ligand or catalyst (control)
. reaction in ionic liquid/water
18
4. as in 3 with no ligand or catalyst (control)
comparison to homogeneous counterparts. There was some
improvement in the yields obtained when moving to the higher
of the two temperatures, but it was not significant (Table 2). It is
also clear that the ligands impart an advantage over the ligandless
These reactions were performed under standard conditions
(120 °C, 24 h) in duplicate. For one set of reaction mixtures,
UV−vis measurements were performed on the water layers of
both reaction mixtures and on the toluene layer. These
measurements showed the complete absence of ligand in the
water layers. In contrast, the spectra of the toluene layer
catalyst (Table 2, entry for Pd(OAc) ), as can be deduced on the
2
basis of the yields of the primary reactions and of the recycling
experiments. The principle is shown that the ligands enable
7
025
dx.doi.org/10.1021/om501094p | Organometallics 2014, 33, 7023−7026

Organometallics
Communication
contained peaks corresponding to the catalyst used, with the
absorption maxima mimicking the freshly prepared complex. It is
possible that the carbene-bound Pd could become core
encapsulated by the porphyrin structure. However, the structure
of the UV−vis spectra generated here after reaction indicated
that this likely did not occur. Free base porphyrazines are
characterized by split Q bands, such as those reproduced in
ACKNOWLEDGMENTS
We thank the University of Johannesburg, Sasol Ltd., and THRIP
for financial support.
■
REFERENCES
■
(
1) (a) Kuntz, E. G. (Rho
News of the Week” in Chem. Eng. News 2003, 81, 9 (March 31, 2003).
̂
ne-Poulenc) Fr. Patent 2314910, 1975. (b)
“
Figures 1 and 2, consistent with D symmetry. Simpler spectra
2h
(2) (a) Aqueous Phase Organometallic Catalysis-Concepts and
accompany a more symmetrical porphyrin structure, such as that
Applications; Cornils, B., Herrmann, W. A., Eds.; Wiley-VCH:
Weinheim, Germany, 2005. (b) Dyson, P. J.; Ellis, D. J.; Welton, T.
Platinum Met Rev. 1998, 42, 135−140.
created when a metal is accommodated in the core, leading to C
2
v
7
c
symmetry. While shoulders on the Q bands may be noted for
7
,8
(
2
3) (a) Williams, D. B. G.; Ajam, M.; Ranwell, A. Organometallics 2006,
metalated porphyrins, significant Q band splitting such as that
observed in free base systems is atypical.
5, 3088−3090. (b) Williams, D. B. G.; Ajam, M.; Ranwell, A.
Organometallics 2007, 26, 469−4695.
4) Cao, Y.; Xing, H.; Yang, Q.; Li, Z.; Chen, T.; Bao, Z.; Ren, Q. Ind.
Eng. Chem. Res. 2014, 53, 10784−10790.
5) Perperi, E.; Huang, Y.; Angeli, P.; Manos, G.; Mathison, C. R.;
Cole-Hamilton, D. J.; Adams, D. A.; Hope, E. G. Chem. Eng. Sci. 2004,
In the second set of reaction mixtures, each water layer
obtained after completion of the reactions was immediately
acidified (nitric acid, 0.5 mL, 69%) to stabilize the Pd. An aliquot
of each acidified aqueous layer (1.0 mL) was diluted to 10 mL
using deionized water, after which ICP-OES measurements were
performed. Pd was detected at the level of 0.12 ppm in the
aqueous layer from reaction 1 (i.e., toluene/water mixture),
signifying very little leaching of Pd from the toluene into the
water. Analysis of the aqueous layer from reaction 3 (ionic liquid/
water) showed the absence of palladium, indicating that the Pd is
retained essentially completely in the ionic liquid layer. These are
exciting proof of concept observations which bode well for the
further development of these ligand/catalyst systems where the
integrity of the complex and its catalytic activity are retained
through many reaction cycles. If catalyst longevity can be
achieved, monitoring of the system over time would be
dramatically simplified, since the ligand has a high extinction
coefficient and is therefore detectable at very low concentrations,
even by simple visual inspection.
(
(
5
9, 4983−4989.
(6) Mokhadinyana, M.; Desset, S. L.; Williams, D. B. G.; Cole-
Hamilton, D. J. Angew. Chem., Int. Ed. 2012, 51, 1648−1652.
(7) (a) Williams, D. B. G.; Mbatha, G. B. Dyes Pigm. 2011, 88, 65−74.
(
b) Hachiya, S.-I.; Cook, A. S.; Williams, D. B. G.; Garrido Montalban,
A.; Barrett, A. G. M.; Hoffmann, B. M. Tetrahedron 2000, 56, 6565−
569. (c) Bellec, N.; Garrido Montalban, A.; Williams, D. B. G.; Cook, A.
S.; Anderson, M. E.; Feng, X.; Barrett, A. G. M.; Hoffman, B. M. J. Org.
Chem. 2000, 65, 1774−1779.
8) Andersen, K.; Anderson, O. P.; Miller, T.; Mani, N. S.; Baumann, T.
6
(
F.; Anderson, M.; Broderick, W. E.; Eichhorn, D. M.; Goldberg, D.;
Jarrell, W.; Lange, S. J.; Lee, S.; Nie, H.; Sabat, M.; Sibert, J. W.; Stern, C.;
Hoffman, B. M.; Baum, S.; Beall, L. S.; Cook, A. S.; Mccubbin, Q. J.;
Garrido Montalban, A.; Rodriguez-Morgade, M. S.; White, A. J. P.;
Williams, D. B. G.; Williams, D. J.; Barrett, A. G. M.; Hope, H.;
Olmstead, M. M. J. Heterocycl. Chem. 1998, 35, 1013−1042.
(
9) (a) Makume, B. F.; Bredenkamp, T.; Williams, D. B. G.
In conclusion, imidazolium-containing phthalonitriles and
their porphyrazine analogues generated active Pd catalysts for the
Suzuki reaction. These were tolerant of various reaction
conditions in homogeneous systems, and the activity was
maintained under biphasic conditions (toluene/water and ionic
liquid/water), where the reagents and catalyst were present in
different phases. The intensely colored porphyrazine catalysts
could be visually tracked, providing a simple means by which to
assess the presence or absence of the catalyst in a given phase.
ICP-OES analysis for Pd showed that only very little Pd leached
into the aqueous layer, while no ligand was detected there by
UV−vis spectroscopy; the catalyst was retained in the non-
aqueous layer (toluene or ionic liquid). Recycling experiments
allowed the catalysts to be reused, albeit with limited success.
This work demonstrates the concept that highly colored ligands
may be useful for the visual location of a catalyst and/or ligand.
Future work in this area will focus on improving the activity of the
catalyst through various recycling experiments.
ChemCatChem 2014, 6, 2801−2804. (b) Williams, D. B. G.; Shaw,
M.; Green, M. J.; Holzapfel, C. W. Angew. Chem., Int. Ed. 2008, 47, 560−
5
2
63. (c) Williams, D. B. G.; Bredenkamp, T. ChemCatChem 2012, 4,
06−208. (d) Williams, D. B. G.; Shaw, M. L.; Hughes, T.
Organometallics 2011, 30, 4968−4973. (e) Williams, D. B. G.; Evans,
S. J.; de Bod, H.; Mokhadinyana, M. S.; Hughes, T. Synthesis 2009,
3106−3112.
(10) Hahn, F. E.; Jahnke, M. C. Angew. Chem., Int. Ed. Engl. 2008, 47,
3122−3172.
(
11) (a) Schmid, T. E.; Jones, D. C.; Songis, O.; Diebolt, O.; Furst, M.
R. L.; Slawin, A. M. Z.; Cazin, C. S. J. Dalton Trans. 2013, 42, 7345−
253. (b) Bernhammer, J. C.; Chong, N. X.; Jothibasu, R.; Zhou, B.;
Huynh, H. V. Organometallics 2014, 33, 3607−3617.
7
(
12) (a) Rajabi, F.; Thiel, W. R. Adv. Synth. Catal. 2014, 356, 1873−
1
2
2
877. (b) Yang, Y.; Feng, M. T.; Lu, J. M. Org. Biomol. Chem. 2013, 11,
266−2272. (c) Wang, Z.-Y.; Chen, G. Q.; Shao, L.-X. J. Org. Chem.
012, 77, 6608−6614.
(13) Kudrik, E. V.; Shaposhnikov, G. P.; Balakirev, A. E. Russ. J. Gen.
Chem. 1999, 69, 1321−1324.
(14) Huang, W.; Guo, J.; Xiao, Y.; Zhu, M.; Zou, G.; Tang, J.
Tetrahedron 2005, 61, 9783−9790.
ASSOCIATED CONTENT
■
(
15) Herrmann, W. A.; Reisinger, C.; Spiegler, M. J. Organomet. Chem.
*
S
Supporting Information
Text giving experimental and analytical details. This material is
available free of charge via the Internet at http://pubs.acs.org.
1
998, 557, 93−96.
(16) (a) Williams, D. B. G.; Shaw, M. L. Tetrahedron 2007, 63, 1624−
1629. (b) Williams, D. B. G.; Pretorius, M. J. Mol. Catal. A 2008, 284,
7
(
(
7−84.
17) Farina, V. Adv. Synth. Catal. 2004, 346, 1553−1582.
AUTHOR INFORMATION
■
*
18) Van Leeuwen, P. W. N. M. Homogeneous Catalysis: Understanding
Corresponding Author
E-mail for D.B.G.W.: bradley.williams@vuw.ac.nz; bwilliams@
uj.ac.za.
the Art; Kluwer: Dordrecht, The Netherlands, 2004.
Notes
The authors declare no competing financial interest.
7
026
dx.doi.org/10.1021/om501094p | Organometallics 2014, 33, 7023−7026