Download PDF
Synlett 2024; 35(09): 1001-1006
DOI: 10.1055/a-2201-7326
cluster
Chemical Synthesis and Catalysis in Germany

Alkene versus Aryl Chlorination in Asymmetric Hypervalent Iodine Catalysis: A Case Study

Andreas M. Arnold
a   Institute of Organic Chemistry, Leipzig University, Johannisallee 29, 04103 Leipzig, Germany
b   Department of Chemistry, Technical University of Munich, Lichtenbergstraße 4, 85748 Garching, Germany
,
Julia Binder
a   Institute of Organic Chemistry, Leipzig University, Johannisallee 29, 04103 Leipzig, Germany
b   Department of Chemistry, Technical University of Munich, Lichtenbergstraße 4, 85748 Garching, Germany
,
Martin Kretzschmar
a   Institute of Organic Chemistry, Leipzig University, Johannisallee 29, 04103 Leipzig, Germany
,
Tanja Gulder  ∗
a   Institute of Organic Chemistry, Leipzig University, Johannisallee 29, 04103 Leipzig, Germany
b   Department of Chemistry, Technical University of Munich, Lichtenbergstraße 4, 85748 Garching, Germany
› Author Affiliations

This work was funded by the Deutsche Forschungsgemeinschaft (DFG, GU 1134/5). A.M.A. thanks the Fonds der Chemischen Industrie for a PhD Fellowship.
› Further Information
 
 PDF Download Permissions and Reprints All articles of this category


Abstract

Hypervalent λ3-iodanes have become a prominent tool for halofunctionalizations of alkenes. Despite many examples of asymmetric fluorinations reported lately, the corresponding enantioselective chlorination reactions using iodoresorcinol-based catalysts are significantly less developed, with only one example known to date. Here, we show how competing aromatic chlorination of the iodoarene catalyst is a significant obstacle in these transformations, hinting towards a conceptual issue with this well-established catalyst class for enantioselective chlorinations. Consequently, the reaction conditions and the catalyst design must be adapted to facilitate an effective chirality transfer. Hence, attention should be paid when selecting the oxidizing agent, the stoichiometry, and careful reaction analysis must be conducted to identify the factual catalytically active species.


#

Key words

chlorination - hypervalent iodane catalysis - asymmetric reactions - iodoresorcinols - rearrangement

The arena of synthetic chemistry continues to evolve with a relentless pursuit of new methodologies that not only facilitate precise bond formations but also enable the construction of increasingly complex molecular architectures. Within this dynamic landscape of organic method development, hypervalent λ3-iodanes have developed from a chemical curiosity to valuable and reliable tools within recent years.[1] They are generally applicable, mild, nontoxic, and environmentally friendly substitutes for toxic and precious metal reagents, orchestrating remarkable transformations not accessible otherwise.[2] In addition to their use in ‘simple’ one-electron transfer reactions, such as the oxidation of alcohols[3] or the dearomatization of phenols,[4] iodanes have also conquered other areas of organic synthesis.[5] Today, many different hypervalent iodanes are available as reagents[6] and catalysts.[7] They have been involved, i.e., in diverse atom transfer reactions,[8] cyclizations,[9] oxidative rearrangements[5c] [10] and fragmentations,[11] α-functionalizations of carbonyl compounds,[12] and even in asymmetric synthesis.[13]

The halogenation of π-systems is one of the oldest, most useful, and most reliable functionalizations of substrates.[14] Although a massive number of different methods have already been described in the literature since the discovery of the dibromination of alkenes over a century ago,[15] there is still demand for novel and mild methods uncovering unusual reactivities and selectivities, thus granting access to new structural scaffolds. Hypervalent λ3-iodanes are an excellent playground for discovering such unprecedented chemistry.[13f] [16] Their role in halogenation reactions has lately garnered profound attention.[6e]

Zoom Image
Scheme 1 a) Hypervalent iodane catalyzed rearrangements of imides 1 to α-hydroxy amide derivatives 4 as key transformation in the cascade reactions to β-lactams 5 and β-amino acids 6, b) first attempt on an enantioselective dichlorination using hypervalent iodine catalysis, and c) problems associated with chlorinations using resorcinol-based hypervalent iodines

Within our research program on halogenations and halofunctionalizations, our group disclosed 2-iodohippuric acid (2) as a simple yet effective catalyst in different types of hypervalent λ3-iodane-triggered and -catalyzed transformations, such as aromatic chlorinations,[17] bromocyclizations,[18] biomimetic dihalogenations, and rearrangements.[16`] [b] [c] , [19] The power of the established protocols was demonstrated in the efficient preparation of bioactive natural products, such as marinopyrrol[17] (not shown), and the incorporation of these transformations in triple and even quadruple reaction cascades.[20] The latter provided access to crucial bioactive substance classes, like, e.g., β-lactams 5 and β-amino acids 6, from simple starting materials.[20] The core of these straightforward multistep approaches constitutes the selective, hypervalent iodane mediated rearrangement of imide 1, which proceeds via the cyclic oxazolidinonium intermediate 3. Upon workup, 3 readily hydrolyzes to furnishing α-hydroxy carboxylamide 4 with high diastereoselectivities (Scheme [1a]).[19] Given the versatility of 4 as a synthetic building block, its synthesis in an optically pure form sparked our interest. A privileged structural motif triggering enantioselective transformations harnessing the iodine(I)/iodine(III) manifold is the C 2-symmetric iodoresorcinol lactate 9. Initially developed by Ishihara et al.,[13g] [h] both the ester and amide analogues of 9 have conquered other transformations,[21] such as amino, fluoro, and oxy functionalizations of alkenes[7a] [13a] [b] [d] [f] [16e] [22] besides rearrangements.[10b] Key to their success is the distinct arrangement of hydrogen-bonding interactions inside the catalyst, resulting in a helically chiral confinement. This, together with a vast structural variability, allows for easy fine-tuning and, thus, adaption of the chiral scaffold to the needs of a specific reaction.[13j] We assumed that these chiral iodoresorcinols 9 are ideally suited for rearranging 1 under suitable conditions. However, the literature does not present reports of utilizing 9 as reagents or catalysts in enantioselective bromo functionalizations.[23] However, recent studies by Wilson and Dutton showed that PhIBr2 is an elusive compound,[24] making the application of linear bromo iodanes not feasible here. Therefore, we wondered if we could use the corresponding, more stable chloro iodanes instead to synthesize 4 in optically pure form. To our knowledge, the Gilmour group reported the only example using catalysts 9 in enantioselective chlorinations within their study on catalytic hypervalent iodine 1,2-dichlorinations of styrenes (Scheme [1b]).[23] There, they showed in one example that this reaction can, in principle, also be carried out enantioselectively. However, the conversion of 8 using the adamantyl derivative 9d as catalyst proved to be very slow (3 days; 23% yield) and gave product 11 with an e.r. of 64:36.

In this paper, we initially intended to develop an enantioselective chlorination-mediated rearrangement of imide 1a employing the iodine(I)/iodine(III) manifold (Scheme [1c]). During our studies we observed predominant aromatic chlorination of the catalyst scaffold. As this unwanted reaction pathway has a significant influence on reaction and catalyst design, we had a closer look on the processes going on during iodoresorcinol-catalyzed chlorinations with particular emphasis on the actual catalytic species.

First, we explored if a hypervalent chloro iodane induced rearrangement is possible by treating imide 1a with stoichiometric amounts of Willgerodt’s reagent (13). Product 4a was isolated in 95% yield (Table [1], entry 1). A catalytic reaction was likewise feasible. Without further optimization, amide 4a was readily accessible in a good 78% yield employing our standard reaction conditions[19] with 10 mol% 2 by simply replacing NBS with NCS (Table [1], entry 2). Here, the cyclic imide 14 resulting from a Friedel–Crafts-type opening of the haliranium-ion intermediate was isolated in 21% as side product. Encouraged by these results, we focused on the asymmetric variant using catalytic amounts of lactate ester 9a instead of 2. Unfortunately, no conversion of starting material 1a was observed even by raising the amounts of 9a to 1.5 equivalents. In all reactions, the chlorine source NCS was consumed to the extent of the added iodobenzene.

Puzzled by this outcome, we looked at the general formation and stability of different dichloro iodoarenes 12 by treating lactate derivatives 9 with chlorine gas (Scheme [2]). In neither case, we could isolate the corresponding λ3-iodane species 12 (Scheme [2a]). Instead, we observed dichlorination of the arene moiety in quantitative yield for ester 15a (R1 = H). Only in the case of the less electron-rich benzoic acid ester derivative 9b a mixture of mono- and dichlorinated iodoarene 11b and 15b (40:60) was produced, likewise in quantitative yield. At the same time, the analogue 9c bearing an amide moiety in the side chain decomposed completely. To exclude that steric congestion at the iodine atom originating from the lactate chains prevents iodo dichlorination, dimethoxy-2-iodoresorcinol (not shown) was subjected to the reaction conditions, but only the dichlorobenzene derivative was afforded here as well (not shown, see the Supporting Information for further information).[17]

Table 1 Hypervalent Iodine Mediated Chlorination-Induced Rearrangements of Imide 1a

Entry

Conditions

Yield of 4a (%)

1

13 (1.5 equiv), DCM

95

2

 2 (10 mol%), NCS, HFIP

78

3

 9a (10 mol%), NCS, DCM

–

4

 9a (10 mol%), NCS, HFIP

–

5

 9a (150 mol%), NCS, DCM

–

6

 9a (150 mol%), NCS, HFIP

–
Zoom Image
Scheme 2 Attempts to generate the elusive dichloroidoanes 9: a) by direct chlorination, b) via ligand exchange, and c) rearrangement of 1a triggered by in situ generated 12a

In an alternative pathway, we pursued the formation of chloro iodane 12 via ligand exchange (Scheme [2b]) using the bisacetoxyiodoarene 16a as the starting material. In contrast to the reactions employing electrophilic chlorine reagents as shown before (cf. Table [1] and Scheme [2a]), no direct electrophilic aromatic chlorination should be possible in this redox neutral pathway as chloride anions are intrinsically not able to react with the aryl ring.

Surprisingly, only chloroiodobenzene 11a was produced in 76% (Scheme [2b]). Even the in situ generation of the iodine(III) species, followed by its immediate trapping by chloride anions, which constitutes a precondition for rendering this reaction catalytic, was successful when using Selectfluor as an external oxidant and CsCl as a chlorine source.[25] Here, the mono 11a and dichloro product 15a were furnished depending on the ratio of reagents and substrate (Scheme [2c]). These studies led to the assumption that the desired dichloroiodane 12 is indeed formed in situ but immediately reacts with itself or another substrate molecule 16, giving 11 and 15. Encouraged by these results, we next blocked the two positions meta to the iodine at the aryl ring (→ 16b) with chlorine substituents to avoid aromatic chlorination. Ligand exchange in 16b occurred smoothly at –20 °C. The formed dichloro iodane reacted directly with 1a upon warming the reaction mixture to room temperature, giving 4a in quantitative yields, albeit with no enantioselectivity for now (Scheme [2d]).

This result, however, clearly showed that the formation of hydroxy amides 4 can be accomplished using resorcinol derivatives 9, but in contrast to fluorinations, the corresponding chlorinations need to be carefully analyzed regarding catalyst integrity. Therefore, we looked closer to see if the aromatic substitution of 9 also occurs during other hypervalent iodine catalyzed chlorinations, such as the 1,2-dichlorination of styrenes 8 (Table [2]). In styrenes, the alkene moiety is much more electrophilic than in imides 1; thus, its chlorination should be much more feasible.

Table 2 Catalytic Chlorination of Styrene 8 and Chlorofunctionalization of Imide 1a a

Entry

Starting material

Catalysts (mol%)

Conversion (%)

Product

Yield (%)b

e.r.c

 1

8

 9a (20)

quant.

10

63

46:54

 2

8

15a (20)

87

10

43

48:52

 3

8

 9b (20)

quant.

10

56

47:53

 4[23]

8

 9d (20)

quant.

10

39

64:36

 5

8

  –

70

10

<5

–

 6

1a

 9a (20)

83

 4a

66

50:50

 7

1a

 9a (50)

60

 4a

45

49:51

 8

1a

 9b (20)

52

 4a

42

48:52

 9

1a

 9d (20)

41

 4a

36

50:50

10

1a

  –

–

 4a

<5

–

a Starting material (200 μmol, 1.0 equiv), 101 mg CsCl (600 μmol, 3.0 equiv), and 78.0 mg Selectfluor (220 μmol, 1.1 equiv) were dissolved in 900 μL abs. DCM together with 190 μL HFIP. The solution was cooled to –20 °C, and the catalyst was added.

b Determined by 1H NMR spectroscopy from the crude reaction mixture using an internal standard.

c Determined by HPLC on a chiral phase

We converted 8 using the mild conditions described by Gilmour et al.[23] The results of the Gilmour group were fully reproducible (entry 4), giving 10 in 39% yield and 64:36 e.r. Using the ethyl ester 9a, the 1,2-dichloro product 10 was delivered in a higher 63% yield (entry 1) together with a significant amount of monochlorinated catalyst 11a (68%) and untouched catalyst 9a (32%). Changing to catalysts 9b and 15a bearing a less electron-rich benzene ring resulted in a drop in overall product yield 10 (43% and 56%, respectively, entries 2 and 3) and a hampered aromatic chlorination activity.

Switching to substrate 1a, no native catalyst 9 was visible in the reaction mixture in all reactions shown below (entries 6–9). By increasing catalyst loading to 50 mol% (entry 7), only 60% conversion was observed, and the yield of amide 4a dropped while the ratio of 11:15 remained the same. These results corroborate our assumption that aromatic substitution of 9 is preferred over electrophilic alkene addition in 1a. When electron-deficient catalyst 9b was used, the imide 1a was only converted with 52%, and product yield was decreased to 42%. In this case, the catalyst was quantitatively monochlorinated with no traces of dichlorinated 15b detectable. The lowered conversion may be attributed to the iodoarene 9b/11b being too electron poor to promote efficient cyclization of 1a. The importance of the iodoarene catalyst for transforming 1a and 8 was established by unproductive control experiments (entries 5 and 10), emphasizing that both substrates must be converted via a hypervalent iodane intermediate such as 12. Unfortunately, there is no asymmetric induction with the employed catalysts except for the 1,2-dichlorination using the sterically demanding adamantyl ester 9d as the chiral catalyst.

The monochloro catalyst 11a was determined to be the actual catalytic species in the chlorination-triggered rearrangement of 1a by kinetic studies (Figure [1]). After only 30 min at –20 °C, catalyst 9a was entirely consumed, and only monochlorinated iodoarene 11a was present in the reaction mixture. Almost no imide 1a (>98%) was converted during this time. After warming the mixture to ambient temperature after 90 min, product formation rapidly occurred, leading to 68% amide 4a after 120 min. Installing a second chlorine atom at the aryl ring in 11a appeared to be very slowly; only 4% dichlorinated catalyst 15a was produced after 3 h. For the 1,2-dichlorination of styrene 8, the aromatic substitution and the electrophilic addition happened simultaneously, so the actual catalyst responsible for the targeted reaction is inconclusive (see the Supporting Information). Interestingly, only when styrene 8 was consumed completely did the formation of 15a start.

Zoom Image
Figure 1 Exploring the time course of the aromatic substitution versus electrophilic alkene addition in the catalytic conversion of imide 1a with resorcinol catalyst 9a. Reagents and conditions: imide 1a, CsCl (3.0 equiv), and Selectfluor (1.1 equiv) were mixed in DCM/HFIP (900:190 μL). At –20 °C, 9a (0.2 equiv) was added to the reaction mixture. At t = 90 min, the mixture was warmed to room temperature. As only 1.1 equiv Selectfluor was added together with 0.2 equiv 9a, the conversion of 1a was not complete at the end of the reaction, as at least 0.24 equiv oxidant were consumed in the aromatic chlorination.

The C 2-symmetric iodoresorcinol lactates 9 are powerful reagents and catalysts in asymmetric hypervalent iodane chemistry. Despite great success in iodine-catalyzed fluorinations, their use in other halogenation reactions is limited to a single example in the literature. We tried to extend their application scope to enantioselective chlorination-mediated rearrangement of imides 1. Although the investigated iodoarenes are, in principle, suitable to catalyze these transformations, aromatic chlorination of the catalysts 9 was identified as the dominant reaction even at low temperatures. This phenomenon was likewise observed, even though to a smaller extent, for the conversion of more reactive compounds, such as styrenes 8 and catalysts 9b, equipped with electron-withdrawing substituents. The chlorinated iodoarene 11 was still able to promote the chloronium transfer in our case, however, this reaction must be considered when adjusting the electronic nature of the iodoarene and the stoichiometry of the oxidizing agent. Moreover, mono- or dichlorination of the iodoarenes 9 alters their electronic nature, which has a significant impact on the oxidizing potential of the iodoarenes, the stability of the resulting linear dichloroiodoarenes,[26] and the transfer of chirality. Based on the studies presented here, stereoselective chlorination-mediated reactions are currently developed in our group.


#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

We thank F. Göricke for assistance in HPLC analysis.

Supporting Information

    Supporting information for this article is available online at https://doi-org.accesdistant.sorbonne-universite.fr/10.1055/a-2201-7326.
  • Supporting Information

Corresponding Author

Tanja Gulder
Institute of Organic Chemistry, Leipzig University
Johannisallee 29, 04103 Leipzig
Germany   

Publication History

Received: 05 October 2023

Accepted after revision: 30 October 2023

Accepted Manuscript online:
30 October 2023

Article published online:
06 December 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 a) Hypervalent iodane catalyzed rearrangements of imides 1 to α-hydroxy amide derivatives 4 as key transformation in the cascade reactions to β-lactams 5 and β-amino acids 6, b) first attempt on an enantioselective dichlorination using hypervalent iodine catalysis, and c) problems associated with chlorinations using resorcinol-based hypervalent iodines
Zoom Image
Scheme 2 Attempts to generate the elusive dichloroidoanes 9: a) by direct chlorination, b) via ligand exchange, and c) rearrangement of 1a triggered by in situ generated 12a
Zoom Image
Figure 1 Exploring the time course of the aromatic substitution versus electrophilic alkene addition in the catalytic conversion of imide 1a with resorcinol catalyst 9a. Reagents and conditions: imide 1a, CsCl (3.0 equiv), and Selectfluor (1.1 equiv) were mixed in DCM/HFIP (900:190 μL). At –20 °C, 9a (0.2 equiv) was added to the reaction mixture. At t = 90 min, the mixture was warmed to room temperature. As only 1.1 equiv Selectfluor was added together with 0.2 equiv 9a, the conversion of 1a was not complete at the end of the reaction, as at least 0.24 equiv oxidant were consumed in the aromatic chlorination.