Frederic
Ballaschk
and
Stefan F.
Kirsch
*
Organic Chemistry, Bergische Universität Wuppertal, Gaußstr. 20, 42119 Wuppertal, Germany. E-mail: sfkirsch@uni-wuppertal.de
First published on 3rd October 2019
It is shown how secondary alcohols are oxidized to provide the corresponding ketones by use of Oxone® and solid-supported hypervalent iodine catalysts. Under experimentally simple conditions with acetonitrile at elevated temperatures, excellent conversions were achieved with low catalyst loadings (0.2–5 mol%) when employing the conjugates 5 and 6 derived from IBX and IBS. The catalysts are broadly applicable to a range of alcohol substrates. Of primary importance with respect to sustainability issues, the metal-free catalysts are easily removed from the reaction mixture through filtration, and they can be re-used in oxidation processes for multiple times, without loss of catalytic activity.
![]() | ||
Scheme 1 A) Structures of IBX and IBS. (B) DMAP–peptide conjugates through Fmoc-SPPS on Boc-Gly Merrifield resin. (C) Solid-supported oxidation catalysts derived from IBX and IBS. |
Despite the many advantages IBX offers, several issues are unattractive to practitioners: (i) similar to many other beloved oxidation agents, stoichiometric amounts of the hypervalent iodine compound are required for alcohol oxidation. This is particularly problematic since IBX is reported to be shock- and heat-sensitive,25,26 at least under certain conditions, thus weakening its routine for large-scale applications. (ii) The regeneration of the IBX species from the 2-iodosobenzoic acid (IBA) formed during the oxidative process is typically avoided, resulting in significant amounts of “organic waste”.27 (iii) DMSO is the preferred solvent for all variants of IBX oxidations,28,29 requiring sophisticated work-up protocols to remove the remainder of the hypervalent iodine reagent, for example by addition of chlorinated solvents. It is therefore reasonable to summarize that IBX and its derivatives are not the ideal reagents with respect to the principles of Green Chemistry30,31 since they are mostly used as stoichiometric and non-recyclable reagents.
Several strategies were developed to tackle those issues. For example, modified IBX reagents32–37 for the use in common solvents other than DMSO and safer formulations38,39 of IBX with reduced explosive properties were described. Experimental procedures that aimed at the effective recovery of the reduced form of the hypervalent iodine compounds using biphasic protocols or fluorous techniques led to further improvements.40–44 Polymer-supported IBX reagents were also reported, with the goal to ease work-up protocols and to improve recyclability,45 including polymer-supported variants of IBX,46–50 IBX-amides,51–56 IBX-esters,57 and 2-iodylphenol ethers.58 In addition, the oxidation with catalytic amounts of IBX59–61 and particularly 2-iodoxybenzenesulfonic acid (IBS, 2)62–65 was shown to be highly efficient.66 Those catalytic methods typically rely on Oxone® as the stoichiometrically used co-oxidant,67–74 a reagent that is considered environmentally safe, non-toxic and non-explosive.75 To facilitate the recovery of the hypervalent iodine organocatalyst, fluorous IBX was recently introduced.76 However, the catalytic use of hypervalent iodine oxidants bound to a polymer support was not reported, until now. This is despite the fact that there would be obvious advantages: the iodine-based catalysts are an environmentally sustainable alternative to transition metals, relatively inexpensive and easily recoverable.77
In this paper, we present two new solid-supported hypervalent iodine reagents for the oxidation of secondary alcohols, derived from readily available starting materials: one is based on IBX, and the other is an IBS derivative. In the presence of Oxone®, the two reagents are shown to be highly effective catalysts when using a water–acetonitrile solvent mixture, with catalysts loadings as low as 1 mol%. We also demonstrate that the two metal-free organocatalysts are easily reusable, and work-up protocols consist of simple filtrations. Therefore, our hypervalent iodine compounds may become attractive ‘green catalysts’ in the ever-evolving fields of alcohol oxidations and other challenging oxidative processes.30,31,78–81
With the Fmoc-protected amino acid 10 in hand, the immobilization on glycin-modified Merrifield resin was accomplished using standard solid-phase peptide synthesis (see ESI† for details). Of note, alanine was chosen as the linker amino acid because of the ease of its introduction with SPPS techniques, its lack of functional groups, and its stability toward oxidative conditions (particularly in comparison to glycine). We decided to start with the conjugate 4 having four alanines as linker unit since, at the outset of our studies, we expected that the use of an elongated and flexible linker that provides a spatial separation of the catalytically active part from the polymer body will be advantageous.51 The loading of the resin was determined to be around 0.46 mmol g−1via UV-VIS measurements (see ESI†). The amino acid 10 was also linked to the aminoethyl polystyrene resin giving conjugates 5. We constructed those conjugates with several linker lengths for further tests, with 5a having no alanine unit (n = 0), 5b having two alanines (n = 2) and 5c having four alanines (n = 4). In all cases, a loading between 0.3 and 0.4 mmol g−1 was achieved.
We next focused on the synthesis of the solid-supported precatalyst 6 (Scheme 2B), relying on our previous synthetic approach to the oxidant IBX-SO3K.37,84,85 We began with 4-aminobenzoic acid, which was selectively sulfonylated with fuming sulfuric acid to give the acid 11. Sandmeyer-type conditions furnished iodide 12. All the steps of this sequence are high-yielding without the need for a chromatographic purification, and multiple runs were successfully and reproducibly carried out in multigram scale. The carboxylic acid 12 was then coupled to the polymer resin bearing a single alanine linker to give the conjugate 6′ with a loading of 0.50 mmol g−1. We note that the direct coupling of 12 to the resin was slow and low-yielding when no alanine was involved. The polymer-supported iodoxy reagent 6 was easily obtained upon oxidation of 6′ with the Oxone® tetrabutylammonium salt (TBA-Oxone®).
Entry | Catalyst (mol%) | Cycle no.a | 13a/14ab |
---|---|---|---|
a Recycling and reuse of catalyst by filtration. b Ratio 13a/14a was determined via GC-FID (calibrated with stock solutions of 13a and 14a) after 18 h. c With 40 mol% of nBu4NHSO4, 2 h, MeCN. | |||
1 | — | — | 92/8 |
2 | 4 (4) | 1 (initial) | 0/100 |
3 | 4 (4) | 2 (1st reuse) | 18/82 |
4 | 4 (4) | 3 (2nd reuse) | 64/36 |
5 | 4 (4) | 4 (3rd reuse) | 82/18 |
6 | 4 (4) | 5 (4th reuse) | 85/15 |
7 | 5c (5) | 1 (initial) | 0/100 |
8 | 5c (5) | 2 (1st reuse) | 0/100 |
9 | 5c (5) | 3 (2nd reuse) | 0/100 |
10 | 5c (5) | 4 (3rd reuse) | 0/100 |
11 | 5c (5) | 5 (4th reuse) | 19/81 |
12c | 6 (1) | 1 (initial) | 21/79 |
13c | 6 (1) | 2 (1st reuse) | 14/86 |
14c | 6 (1) | 3 (2nd reuse) | 12/88 |
15c | 6 (1) | 4 (3rd reuse) | 14/86 |
16c | 6 (1) | 5 (4th reuse) | 19/81 |
17c | 6 (1) | 6 (5th reuse) | 22/78 |
The failure of 4 was attributed to the weakness of the ester bond between the Merrifield resin and the glycine linker toward hydrolysis under the reaction conditions. Indeed, control experiments showed that esters were partly hydrolysed in aqueous acetonitrile at 70 °C in the presence of Oxone® while amide bonds remained fully stable. As a result, column chromatography was required to obtain analytically pure oxidation products when employing precatalyst 4. Simple filtration, on the other hand, gave the ketones with traces of impurities that were not unequivocally identified, but may certainly stem from the partial hydrolysis of 4. The design error of 4 then led us to carefully study the systems 5 and 6 where the hypervalent iodine precursor units were connected to the aminoethyl polystyrene resin via more stable amide bonds.
Gratifyingly, it was easily possible to reuse the catalysts 5 and 6 for multiple times, after recycling of the resins through simple filtration. As shown in Table 1 (entries 7–11), the IBX-derived system 5c kept its activity over five cycles, and complete oxidation of borneol was achieved using 5 mol% of the catalyst with 1.6 equivalents of Oxone® at 70 °C in aqueous acetonitrile. Lower catalyst loadings (<2.5 mol%) were shown to provide incomplete conversions, even after elongated reaction times. The IBS variant 6 was effectively reused for five times (Table 1, entries 12–17), accomplishing a constant conversion between 78% and 88%, albeit with lowered catalyst loading (1 mol%) and reduced reaction times (2 h). As a result of the recycling tests, we decided that both systems 5 and 6 based on the aminoethyl polystyrene resin qualify well for catalyzing the oxidation of secondary hydroxy groups under the premise that they deliver easy work-up and multiple reuses.
Our optimization attempts with catalyst 5c then showed that a broad range of reaction conditions is feasible, and the oxidation proceeds equally well in water-free MeNO2, aqueous MeNO2 (30% water) and aqueous MeCN (5–30% water). However, the use of dry acetonitrile as solvent led to significantly diminished yields. Best conversions were achieved between 70 °C and 90 °C, while temperatures below 60 °C did not provide the product of oxidation. To our surprise, the linker length had almost no influence on the catalyst efficiency. As summarized in Table 2, similar conversions were achieved when using 2.5 mol% of the polymers 5a, 5b for the oxidation of nonanol 13b.
The IBS-derived catalyst system 6 turned out to be markedly more active, compared to 5: Table 3 demonstrates that 3 mol% of 6 or of its unoxidized precursor 6′ are sufficient to achieve an excellent conversion at 70 °C in aqueous acetonitrile. The required reaction times were notedly shorter than with IBX derivative 5. Under non-aqueous conditions in acetonitrile or nitromethane, the oxidation proceeded slowly, and the weak conversions were attributed to the low solubility of Oxone® (entries 3 and 4). The addition of substoichiometric amounts of nBu4NHSO4 as phase transfer catalyst resulted in a tremendous rate enhancement.62 In the presence of 0.4 equivalents of nBu4NHSO4, it was easily possible to use 6 with extremely low catalysts loadings of 1 mol% and 0.2 mol% to accomplish useful conversions (entries 8 and 9). Under the conditions, a TON up to 455 was reached with regard to each catalytic center (with a TOF of around 5.3 × 10−3 s−1).
Entry | Catalyst (mol%) | Additive | Solvent | 13b/14ba |
---|---|---|---|---|
a Ratio 13b/14b was determined via GC-FID (calibrated with stock solutions of 13b and 14b). b 6 h. c 2 h. d 4 h. e 24 h. | ||||
1b | 6′ (3) | — | MeCN–H2O (7![]() ![]() |
5/95 |
2b | 6 (3) | — | MeCN–H2O (7![]() ![]() |
3/97 |
3c | 6 (2.5) | — | MeCN | 94/6 |
4c | 6 (2.5) | — | MeNO2 | 91/9 |
5c | 6 (2.5) | n Bu4NHSO4 (10 mol%) | MeCN | 43/57 |
6c | 6 (2.5) | n Bu4NHSO4 (10 mol%) | MeNO2 | 63/37 |
7c | 6 (2.5) | n Bu4NHSO4 (20 mol%) | MeCN | 3/97 |
8d | 6 (1) | n Bu4NHSO4 (40 mol%) | MeCN | 2/98 |
9e | 6 (0.2) | n Bu4NHSO4 (40 mol%) | MeCN | 9/91 |
10d | — | n Bu4NHSO4 (40 mol%) | MeCN | 94/6 |
The scope of the catalyzed oxidation of secondary alcohols with Oxone® was then explored, using the solid-supported compounds 5c and 6 under standardized conditions. As summarized in Table 4, various structurally diverse secondary alcohols 13 were examined as substrates using method A [5c (5 mol%), Oxone® (1.6 equiv.), 18 h, 70 °C, MeCN–H2O (7:
3)] and method B [6 (5 mol%), Oxone® (1.6 equiv.), nBu4NHSO4 (0.4 equiv.), 18 h, 70 °C, MeCN]. We point out that the addition of the phase transfer agent was only required when using non-aqueous conditions (method B); aqueous solvent mixtures (method A) do not require the additive since Oxone® is easily dissolved under the conditions. A general trend was that the performance of the IBS-derived catalyst 6 was superior. In particular, sterically demanding secondary alcohols (e.g., 13a, 13d, 13f and 13k) were easier oxidized to the corresponding ketones using method B. The reaction times until full conversion of the starting substrate were also markedly shorter with 6, compared to 5c (method A). Another advantage of catalyst 6 was that it was employed in a combination with nBu4NHSO4 under non-aqueous conditions, allowing to use other solvents than acetonitrile if required due to solubility issues of the substrate. For example, 5α-cholestan-3β-ol 13s was successfully oxidized with Oxone® and 6 in toluene under otherwise identical conditions (entry 19). The main advantage of method A (with 5c) was the simple work-up protocol consisting only of filtration to remove 5c and subsequent extraction of the aqueous layer. If the starting material was completely converted, further purification by column chromatography was typically not necessary. In the case of method B (with 6), an additional filtration over silica for the removal of the tetrabutylammonium salts was required to obtain analytically pure compounds; the phase transfer agent was not recovered.
Entry | Alcohol 13 | # | Ketone 14 | Yield [%]a,b | Entry | Alcohol 13 | # | Ketone 14 | Yield [%]a,b | ||
---|---|---|---|---|---|---|---|---|---|---|---|
A | B | A | B | ||||||||
a Isolated yield. b n.r. = experiment was not run. c 3 h. d 24 h. e 72 h in acetone. f 90% from benzoin. g 18 h in toluene. h Determined via GC-FID (calibrated with stock solutions of 13t/14t). | |||||||||||
1 |
![]() |
a |
![]() |
79 | 91c | 11 |
![]() |
k |
![]() |
76 | 98 |
2 |
![]() |
b |
![]() |
70 | 79c | 12 |
![]() |
l |
![]() |
82 | 84 |
3 |
![]() |
c |
![]() |
80 | 85c | 13 |
![]() |
m |
![]() |
71 | 97f |
4 |
![]() |
d |
![]() |
32 | 65d | 14 |
![]() |
n |
![]() |
n.r. | 89 |
5 |
![]() |
e |
![]() |
71 | 83 | 15 |
![]() |
o |
![]() |
0 | 66c |
6 |
![]() |
f |
![]() |
0 | 32e | 16 |
![]() |
p |
![]() |
0 | 66c |
7 |
![]() |
g |
![]() |
82 | 85 | 17 |
![]() |
q |
![]() |
95 | 96 |
8 |
![]() |
h |
![]() |
84 | 84 | 18 |
![]() |
r |
![]() |
0 | 61 |
9 |
![]() |
i |
![]() |
n.r. | 98c | 19 |
![]() |
s |
![]() |
0 | 78g |
10 |
![]() |
j |
![]() |
75/58/69/82/63 | 86/79/85/86/82 | 20 |
![]() |
t |
![]() |
n.r. | 75h |
We point out that primary alcohols also provide oxidation products under the conditions of method B. However, aldehydes were produced in markedly reduced yields and mixed with various products of over-oxidation, thus limiting the use of the catalytic system tremendously. In the case of primary allylic alcohols 15, it was possible to isolate the corresponding aldehydes in pure form and moderate yields, as outlined for 16a and 16b in Scheme 3. Two other limitations of our catalyst system for the oxidation of secondary alcohols were identified in the course of our studies: (1) phenols are not stable under the conditions,14,15 and (2) amines are not tolerated due to various competing condensation and oxidation reactions that were not further analyzed.
Method B using IBS-derived polymer 6 is also easily applicable to larger scales, as demonstrated in Scheme 4 for the conversion of secondary alcohol 13b into ketone 14b: 50 mmol of 13b led to the isolation of the desired ketone with 80% yield after 5 h, using only 1 mol% of catalyst 6. Of note, 95% of the supported catalyst were recovered for re-uses.
Our future studies will make use of the catalyst systems for chemoselective oxidations, by varying the linker units. We will also show in due course how the immobilized catalysts are of use in flow chemical oxidation processes.
Footnote |
† Electronic supplementary information (ESI) available: Experimental procedures, analytical data and copies of 1H, 13C NMR-spectra. See DOI: 10.1039/c9gc02605c |
This journal is © The Royal Society of Chemistry 2019 |