Download PDF
Synlett 2023; 34(05): 405-413
DOI: 10.1055/a-1990-5495
account

Discovering the Site-Selective Umpolung of Ketones Triggered by Hypervalent Fluoro-Iodanes – Why Investigating Side Reactions Matters!

Martin Kretzschmar
a   Institute of Organic Chemistry, Leipzig University, 04103 Leipzig, Germany
,
Tanja Gulder  ∗
a   Institute of Organic Chemistry, Leipzig University, 04103 Leipzig, Germany
b   Department of Chemistry, Technical University Munich, 85748 Garching, Germany
› Author Affiliations

This work was funded by the Emmy-Noether Program (GU 1134/3) and the Heisenberg Program (GU 1134/4) of the Deutsche Forschungsgemeinschaft (DFG).
› Further Information
 
 PDF Download Permissions and Reprints All articles of this category


Abstract

In this account, we describe our journey leading to the discovery of a generally applicable umpolung method for the α-functionalization of ketones. Central to this reaction is the cyclic hypervalent fluoro-iodane, which is mostly known for various alkene functionalizations enabling, for example, the synthesis of fluoro-benzoxazepines, indoles, and ketones. During this work, we encountered α-functionalized ketones as minor side products. This observation prompted us to further investigate this reactivity, thus revealing a directed umpolung of pyridyl ketones by the fluoro-iodane. The key to the success was the unexpected non-covalent interaction between the nucleophile, substrate, and iodane.

1 Introduction

2 Cyclizations Triggered by the Fluorination of Styrenes

3 Umpolung Reactions Facilitated by Hypervalent Iodanes

4 Discovering and Evolving a Fluoro-Iodane-Triggered Regioselective α-Functionalization of Carbonyl Compounds

5 First Investigations on the Nitrogen-Directed Umpolung

6 Conclusion


#

Key words

hypervalent iodane - fluorination - umpolung of ketones - non-covalent interactions - α-functionalization - halogen bonding

Biographical Sketches

Zoom Image

Martin Kretzschmar graduated from Leipzig University with an M.S. in chemistry in 2014. Afterwards he joined the group of Prof. Christoph Schneider to work on his Ph.D. thesis in the field of enantioselectively catalyzed transformations of o-quinone methide imines, which was funded by the DBU (German Federal Environmental Foundation) with a scholarship. After finishing his Ph.D., he joined the group of Prof. Tanja Gulder in 2020 as a permanent research associate to work in the field of synthetic halogenations.

Zoom Image

Tanja Gulder studied chemistry at the University of Wuerzburg where she received her diploma in 2004. After earning her Ph.D. with distinction in 2008 under the supervision of Prof. G. Bringmann, she pursued postdoctoral studies with Prof. P. S. Baran at The Scripps Research Institute (La Jolla, CA, USA). In 2011, she returned to Germany and began her independent career supported by a Liebig Fellowship (Fonds der Chemischen Industrie) at RWTH Aachen. In 2014, she moved to TU Muenchen as an Emmy ­Noether research group leader and was appointed to the Heisenberg Professorship of Biomimetic Catalysis in 2018. Since 2020, she has been the Chair of Organic Chemistry at Leipzig University. Her laboratory is dedicated to biomimetic catalysis (enzyme mimicking), with special focus on halogenations and their application in the synthesis of natural products and therapeutics

1

Introduction

Right from the start of our research group, we have been fascinated by the unique properties and reactivities of halogenated compounds. This prompted us to establish a research program on developing versatile and broadly applicable methods for the selective introduction of halogen atoms into molecular scaffolds. As such, finding new ways for installing fluorine, as the smallest but most significant and impactful member of the halogen family, seemed highly appealing. Since the first explicit report on organofluorine compounds in 1835,[1] the potential of fluorine-containing molecules has been rapidly recognized leading to their current essential status in materials science[2] as well as in agro[3] and pharmaceutical[4] chemistry. This popularity is easily explained when comparing a C–F bond with its corresponding C–H analog. Due to its very high electronegativity, substituting a hydrogen with a fluorine atom significantly alters the electronic properties of an organic molecule while leaving its steric properties almost unchanged. Therefore, fluorine atoms are widely used to tune the physical, chemical, and pharmaceutical properties of molecules.[4c] [e]

Zoom Image
Scheme 1 Structures of known fluorinating agents and selected examples of methods employing fluoro-iodane 1 in electrophilic fluorinations

In contrast, accessing new methods for installing fluorine atoms into organic structures has been associated with several obstacles, such as the challenges in handling F2 gas or XeF2. Therefore, fluorination reactions were, for a long time, far less straightforward than other electrophilic halogenations facilitated by molecular chlorine, bromine, and iodine. The introduction of mild and robust fluorinating reagents (F-reagents), such as N-fluoropyridinium salts 2 [5] and Selectfluor (3) (Scheme [1]), helped to overcome these problems,[6] and as a consequence, their use has enabled a plethora of synthetic fluorination pathways.[7]

In 2015, we started exploring the reaction scope of the back then newly discovered cyclic hypervalent λ3-fluoro-benziodoxole 1 (Scheme [1]).[8] Various linear λ3-iodane reagents, such as, p-(difluoroiodo)toluene (4),[9] have been known for a long time and have received particular interest among the synthetic community.[10] These reagents offer new reactivities compared to other common electrophilic F-reagents and thus open ways to novel fluorine-containing scaffolds.[10b] Their broad application, however, was hampered by their inherent chemical instability. The cyclic iodoxole 1 has the advantage of being bench-stable and thus storable while retaining the reactivity specific for F-iodanes. When we started our studies, gem-difluorinations[11] and fluoro lactonizations[5] of styrenes 5 and 6, respectively, as well as α-fluorinations of β-ketoesters 7 [8b] were among the known reaction types addressable by 1. These initial transformations have been complemented over the years and today encompass ring openings,[12] oxygen,[13] amino[14] and carbon[14] fluoro functionalizations, syntheses of heterocycles,[15] Balz–Schiemann reactions, and Hofmann degradations,[16] besides many other synthetic applications.[13] [17] This manifold of reactions triggered by iodane 1 shows the versatility of this intriguing reagent, not only for already well-established fluorinations and fluoro-functionalizations, but also in our pursuit to expand the reactivity space of cyclic hypervalent iodanes.

Zoom Image
Scheme 2 (a) Chemoselective fluorinations of styryl-benzamides leading to two distinct products depending on the employed fluorinating agent. (b) Putative mechanism of the benzoxazepine formation.

# 2

Cyclizations Triggered by the Fluorination of Styrenes

At the beginning, we chose styryl-benzamide 11a as the model substrate and treated it with the F-iodane 1 expecting a 6-exo-trig cyclization to give the benzoxazine 12a (Scheme [2, a]). The reaction proceeded rapidly in the presence of molecular sieves, which were added to avoid putative hydrolysis of the iodane 1. To our surprise, instead of 12a, a 7-membered ring product was formed, i.e., the fluoro-benzoxazepine 13a.[18] Besides the, at the first glance unusual 7-endo ring closure, the methyl group in 13a (R3) was formally shifted from the benzylic position in 11a to the homobenzylic carbon atom in 13a. Overall, product 13a was obtained in 77% yield already after 5 minutes of stirring at room temperature in acetonitrile (MeCN). In terms of substrate scope, the reaction also proved to be quite advantageous. We were able to employ substrates 11 with different aromatic, heteroaromatic and alkyl residues at the amide (R1) functionality (Scheme [3, a]). Electron-withdrawing and electron-donating substituents on the aromatic core (R2) and different alkyl substituents (R3) at the alkene moiety were also well tolerated (61–85% yield). In addition, trisubstituted derivatives (R4 ≠ H) likewise furnished the corresponding products 13 in good yields (76% for R4 = Me) and with decent diastereomeric ratios (d.r.) of up to 80:20.[18]

As a cross-check, we treated styryl-benzamide 11a with other electrophilic fluorinating reagents. Only Selectfluor (3) succeeded in converting our starting material 11a (Scheme [2, a]), indeed giving the chemoselectivity expected in the first place. The benzoxazine 12a bearing an exocyclic fluoromethylene unit was delivered after 24 h at room temperature in 61% yield. This result corroborated our assumption that the formation of the fluoro-benzoxazepines 13 is unique to F-iodanes such as 1.

Mechanistic studies involving an isotopically labeled external alkene moiety in 11a (Scheme [2, b]), together with DFT calculations, confirmed our suggestion that the reaction sequence involves a 1,2-rearrangement of the aryl portion rather than a 1,2-methyl shift. Such aryl migrations proceeding via phenonium ions 16 have been known in hypervalent iodane chemistry for a long time.[5] [11] [19] However, they usually lead to regioisomeric products as a result of unselective attack of the inter- or intramolecular nucleophile at C1 or C2 of the cyclopropyl ring in 16 (Scheme [2, b]). In our case, the opening of the three-membered spirocycle occurred solely at the fluorine-containing C1 center.

This unexpected regioselectivity prompted us to undertake further investigations, leading to the following putative mechanistic sequence (Scheme [2, b]).[18] Initially, iodane 1 is activated and pre-organized in proximity to substrate 11a by forming a hydrogen bond between the amide NH and the F atom in 1.[20] σ-Metathesis involving the cyclic, 4-membered, transition state 14 affords the 1,2-iodofluorine species 15. Our mechanistic scenario for the addition of 1 to 11a was corroborated by DFT calculations for similar transformations conducted by Szabó and Himo.[21] Intermediate 15 further reacts by nucleophilic attack of the aryl ring, furnishing the cyclopropyl species 16 upon reductive cleavage and release of aryl iodide (ArI). The aryl iodide(I) compound can be recovered from the reaction mixture during workup and then be recycled into iodane 1, which contributes to a more sustainable reaction sequence.

The cyclopropyl group in 16 then undergoes spontaneous ring opening thermodynamically driven by re-aromatization. The obtained carbocation 17 is stabilized by the 2p non-bonded electron-pair back bonding of the F atom (α-F effect) and the lone pair of the amide carbonyl group.[22] The carbocation 17 is thus the decisive intermediate determining the regioselectivity. Next, trapping of 17 occurs via cyclization followed by deprotonation to yield exclusively the fluoro-benzoxazepine 13a. The entire mechanism can therefore be summed up as a fluorination/1,2-aryl migration/cyclization cascade.[18]

Zoom Image
Scheme 3 (a) + (b) λ3-Fluoro-iodane 1 triggered conversion of styrenes 11 and 18 into benzoxazepines 13 and 19 via a fluorination/1,2-aryl shift/cyclization cascade. (c) Further transformations of benzoxazepines 19 into ketones 20 and indoles 21.

Having achieved our initial goal of finding new applications based on the unique reactivity of fluoro-iodane 1, we further explored the boundaries of our concept. The next step was to broaden the substrate scope. As pyridine-containing compounds play a superior role as pharmacophores,[4b] [23] we first replaced the phenyl moiety with a pyridine ring in starting material 18 (Scheme [3, b]).[24] Notably, when trying to convert 18 employing our reaction conditions, we did not detect any conversion. We attributed this result to the highly deactivated alkene moiety in 18 due to the electron-withdrawing nature of the pyridine ring. On adding Lewis acidic additives such as AgBF4 we re-installed the desired reactivity by further activating iodane 1. The fluoro-aza-benzoxazepines 19 were thus accessible, however, the reaction time significantly increased to almost 3 hours. The substrate scope for this transformation was similar to that observed for the benzoxazepines 13.[24]

With these fluoro-aza-benzoxazepines 19 in hand, we triggered further downstream processes. Hydrolysis of 19 under basic conditions gave the benzyl ketone 20 in quantitative yields (Scheme [3, c]). However, when analyzing the resulting reaction mixture more closely, we also found traces of azaindole 21. As such structural scaffolds are of relevance in medicinal chemistry[23] [25] the reaction was pushed further in that direction. By directly treating the crude reaction mixture obtained after basic hydrolysis with a strong acid, we achieved full conversion into the azaindole 21. The thereby developed one-pot synthesis of 21 gave rise to the desired unprotected azaindole 21b in an excellent 97% yield.[24]

Zoom Image
Scheme 4 (a) Extension of the previously established fluorination–cyclization methods of styryl-benzamides 11 enabling the synthesis of ketones 20. (b) + (c) Fluorination–cyclizations to give indoles 21 and 22 with reversed substitution patterns.

Our success in obtaining ketone 20 and azaindole 21 prompted us to directly target the synthesis of the corresponding aryl derivatives as well. Looking closer at the mechanism outlined in Scheme [2], this goal can also be reached by trapping the decisive α-fluoro carbocation 17 with water. Indeed, traces of water introduced to the reaction mixture by simply omitting the molecular sieves delivered the benzylic ketones 20 from styrenes 11 in up to quantitative yields and in a single step (Scheme [4, a]).[22] DFT calculations confirmed the open chain cation 17 (see Scheme [2, b]) to be sufficiently stable to be trapped by an external nucleophile such as H2O, further strengthening our mechanistic assumptions.

Subsequently, we showed that o-nitrogen-substituted ketones 20 were also easily cyclized to give the corresponding indoles 21. To facilitate the cyclocondensation step, we added 50 mol% of camphorsulfonic acid (CSA) to our reaction mixture. Under these conditions, we directly isolated indoles 21 from styrenes 11 (Scheme [4, b]). As the residues R3 and R4 formally switched their positions in products 21, the indole assembly must have occurred via the postulated quadruple cascade reaction involving a fluorination/1,2-aryl migration/water addition/cyclocondensation sequence. Because of the mild reaction conditions, diverse indoles 21 were generated from styrenes 11 in excellent yields (51–96%) and all examples exclusively delivered the rearranged products 21. Even structurally complex molecules, such as isotryptophanes (not shown), were accessible using our method.[22] Interestingly, by replacing the N-protecting group (R1) with an electron-donating group, such as a methyl or tosyl, the spirocyclopropane intermediate 23 was stabilized, thus enabling the intramolecular nitrogen atom to directly open the three-membered ring. As this reaction follows a SN2 mechanism, the cyclopropyl group is attacked exclusively at the sterically less hindered carbon atom to afford indoles 22 with a complementary substitution pattern (Scheme [4, c]).[22] This variant was applicable to a wide range of structurally diverse substrates with variations tolerated at all relevant positions.[22]

Zoom Image
Scheme 5 An observed side reaction when employing pyridyl-styrene 18a in the established synthesis of indoles 21

During these investigations on establishing fluorine as a traceless directing group in indole syntheses (Scheme [4, b]), we detected the formation of the α-functionalized ketone 24a as a minor component when employing pyridyl-styrene 18a (Scheme [5]). Although we were initially puzzled by the generation of product 24a bearing a camphorsulfonic acid moiety α to the ketone, we quickly became aware that 24a must be built via α-umpolung of the in situ generated ketone 20a. This possible mechanistic scenario initially triggered our interest to look closer into the field of iodane-triggered umpolung reactions.


# 3

Umpolung Reactions Facilitated by Hypervalent Iodanes

Carbonyl compounds, especially ketones, are among the most commonly used synthetic building blocks in organic synthesis, as they can act as both nucleophiles and electrophiles.[26] To further broaden the reaction portfolio of ketones, the umpolung concept was introduced by Corey and Seebach, changing the electrophilic reactivity of the carbonyl C atom by converting it into a 2-lithio-1,3-dithiane.[27] Later, methods were developed to achieve umpolung of the inherent nucleophilic polarity of the α-position of carbonyl compounds (Scheme [6, a]). Halogen atoms,[28] Lewis acids,[29] and transition-metal catalysts[30] were mainly employed to achieve this task.

Zoom Image
Scheme 6 Concepts for using λ3-iodanes to facilitate the umpolung and subsequent addition of nucleophiles α to ketones

Hypervalent iodanes serve as versatile reagents in such α-umpolung reactions. The first example dates back to work by Mizukami in 1978.[31] Mechanistically these transformations proceed via SN2′ attack of a nucleophile on an in situ formed enolonium 32.[32] Since the initial discovery of this reactivity, iodane-mediated umpolung methods have been significantly advanced over time, now covering a broad spectrum of transformations.[33] Today, they can be classified into two distinct categories. On the one hand, there are strategies using iodanes 28 that already contain the nucleophile that is to be transferred to the ketone 27. PIDA (30)[31] and Togni Reagent II (31)[33a] are commonly used to achieve such reactions (Scheme [6, b]).[17a] , [33`] [c] [d] [e] [f] In these cases, the choice of the nucleophile is limited by its ability to form a stable λ3-hypervalent iodane reagent 28. Another issue concerns the regioselectivity if more than one enolizable position is present in 27. Both obstacles lead to significant restrictions in substrates and nucleophiles suitable for this kind of transformation.[33g]

A solution to the latter problem is to preform the enol species 33 either as an iridium enolate (X = IrLn)[33h] or as a silyl enol ether (X = OSiR3) (Scheme [6, c]),[33`] [j] [k] [l] [m] [n] [o] which constitutes the precondition for umpolung reactions of the second category: here, the iodane and the nucleophile are separately applied to the reaction mixture. This enables the successful introduction of a plethora of external nucleophiles. For example, amine,[33i] nitrile,[33j] azide,[33k] aryl[32] , [33`] [m] [n] and alkyl[33o] nucleophiles. However, an additional synthetic step is required to obtain the enol 33, which presents a shortcoming of this method. Overall, there were no direct, regioselective hypervalent iodane-mediated α-functionalizations of ketones bearing two enolizable α-positions using external nucleophiles when we started our investigations in this field.


# 4

Discovering and Evolving a Fluoro-Iodane-Triggered Regioselective α-Functionalization of Carbonyl Compounds

Zoom Image
Scheme 7 A successful attempt at the direct oxidative α-umpolung of pyridyl ketones 36 employing fluoro-iodane 1

For a closer look into the reactivity observed for 18a and to verify our hypothesis of ketone 20a being the decisive intermediate of the assumed α-functionalization (cf. Scheme [5]), we directly employed ketone 36a in the reaction and again obtained the corresponding camphorsulfonic acid substituted product in a low, albeit similar yield as before (not shown). Switching, however, to benzotriazole (BTA-H) (37) as the nucleophile and dichloromethane (DCM) as the solvent significantly increased the formation of the analogous α-functionalized product 38a to 47% (Scheme [7]). To our delight, the reaction did not exhibit the previously described limitations of the literature known, iodane-mediated α-functionalizations.[31] [32] [33] Although starting material 36a contained acidic protons on both carbon atoms adjacent to the carbonyl functionality, we solely functionalized the benzylic CH2 position without any pretreatment of substrate 36a, while still using an external nucleophile.

The reaction conditions were further evaluated using the unsubstituted pyridyl ketone 36b as a model substrate, together with BTA-H (37) as the nucleophile (Scheme [8]).[34] To avoid geminal disubstituted side products the amount of fluoro-iodane 1 was reduced to 0.9 equivalents. Other hypervalent iodane reagents, especially those known to efficiently facilitate α-umpolung reactions such as PIFA (39), Koser’s reagent (40), and PIDA (30) were tested. These studies showed that only the linear, chemically sensitive F-iodane 41 delivered product 38b in a comparable yield (53%). The weakly oxidizing PIDA (30) was likewise capable of converting 36b, but here only the acetoxylated product (not shown) was isolated in 20% yield, hinting at transfer of the internal nucleophile as reported in the literature[31] (cf. Scheme [6]).

Zoom Image
Scheme 8 Exploring the reaction conditions

Because of its chemical stability, the cyclic fluoro-iodane 1 was selected as the optimal umpolung agent for this transformation, affording product 38b in 72% yield under the optimized conditions, while no relevant background reactions were observed.[34]

The method showed a broad scope with respect to the applicable external nucleophiles (Scheme [9]). Besides different N-nucleophiles, such as heterocycles (38c, 38d) and a sulfonamide (38e), a wide range of structurally different O- and S-nucleophiles were selectively installed to give products 38f–n in good yields (41–75%).[34] A necessary precondition for a successful conversion constituted the presence of a slightly acidic proton in the nucleophile. Therefore, enamines, azides and allyl silanes, together with chloride (in the form of HCl), were not suitable. Structural variations in the ketone 36 were also well tolerated as long as the pyridyl N atom remained at the ortho-position. Substrates bearing different heterocycles, e.g., with additional N atoms (38o–q), and various substituents on the pyridine core (38r–v) were accepted in this α-functionalization (Scheme [10]).[34] The range of substituents tolerated at the other site of the carbonyl (R2) spanned from alkyl, aryl, and even benzyl portions to alkynyl as well as electron-withdrawing fragments, such as CF3 or esters (38w–ac). An enamide was likewise transformed, leading to product 38ad in 51% yield. This shows that the method is not limited to carbonyl compounds and thus an extension to a broader array of substrate classes is in principle possible.[34]

Zoom Image
Scheme 9 Substrate scope and limitations of the developed Umpolung reaction in terms of tolerated nucleophiles

# 5

First Investigations on the Nitrogen Directed-Umpolung

The regioselectivity observed here for the α-functionalization of ketones 27 might be explained by precoordination of the hypervalent iodane 1 to the Lewis-basic nitrogen of the pyridine ring. The subsequent nucleophilic addition of the enol to the electrophilic iodane 1, which now proceeds intra- and not intermolecularly, can then only lead to intermediate 45, which, compared to 27, exhibits a reversed polarity at the carbon atom adjacent to the carbonyl functionality (Scheme [11]). Displacement of the hypervalent iodane hypernucleophuge by an external nucleophile then gives rise to the target compound 38 in a proximity-controlled reaction. We proposed fluorine-containing iodanes to be capable of offering significant non-covalent interactions based on halogen bonding[35] between the electrophilic iodine atom and intra- and intermolecular nucleophiles. Hints toward the possibility of such interactions were given by Togni et al. (Scheme [12, a], left).[36] The existence of the Weiss reagent (48) further proves that iodine(III) species can be stabilized, in particular with pyridine moieties, which further corroborates our working hypothesis.[37]

Zoom Image
Scheme 10 Substrate scope of the developed umpolung reaction in terms of tolerated substitutions on the ketone 36
Zoom Image
Scheme 11 Mechanistic proposition containing all determined and relevant interactions

To shed light on the underlying principles leading to the substrate-directed umpolung, we had a closer look at the role of the Lewis basic pyridyl N atom. Shifting the position of the interacting nucleophilic ortho-nitrogen to the meta 36ae or para 36af position of the pyridine ring or using the benzyl analog 36ag (Scheme [12, b]) resulted in no conversion of 36. This emphasizes its decisive role. Only the activated benzylsilyl enol ether 49 showed some conversion, although only the N-2-linked product 50 was isolated in 24% yield, and no product 38 was detectable. In contrast, when the o-pyridyl enol ether 51 was used, no difference in the chemical yield of the obtained addition product 38b (66%) was observed compared to the reactions employing the corresponding ketone 36b. A reaction pathway proceeding via an in situ formed pyridinium species, such as 52, could be excluded as only starting material was re-isolated when ketone 36b was treated with 1. Overall, these preliminary mechanistic studies point at the importance of both the presence and the position of the Lewis basic ortho nitrogen atom close to the acidic α-C-H functionality, confirming its role as a directing (44, see Scheme [11]) and intermediate stabilizing group (45, Scheme [11]).

Zoom Image
Scheme 12 Mechanistic considerations and investigations mainly concerning the role of the acidic nucleophile and the pyridyl N atom

Next, we drew our attention to the necessity of a finely balanced Brønsted acidic nucleophile (Scheme [12, c]).[34] The in situ NMR studies showed a strong H-bonding interaction of the fluorine atom in iodane 1 and the nucleophile, here BTA-H (37), but no ligand exchange of the fluoride by BTA. This non-covalent interaction should further activate both nucleophile 37 and iodane 1 at the same time (see complex 44, Scheme [11]). In iodane 1 the electron density is pulled out of the F–I bond leading to further strengthening of the electrophilic σ hole located at the iodine atom and thus the halogen bonding between the pyridine moiety and the hypervalent iodane. Further in-depth studies of this phenomenon are underway in our laboratory.


# 6

Conclusion

In summary, we have described our journey in exploring the addressable reaction scope of the cyclic fluoro-iodane 1 in this account. Starting from alkene functionalizations we moved to the umpolung of the carbon atoms adjacent to carbonyl functionalities. Our efforts were driven by our curiosity that triggered a thorough analysis of unsuccessful transformations and/or formed side products. In doing so, we established a generally applicable protocol for the addition of nucleophiles, such as BTA (37), to the α-position of ketones by taking advantage of a polarity reversal induced by the iodane 1. With our method, we have contributed to overcome the previously described shortcomings of this reaction type, such as the lack of regioselectivity together with preforming enol ethers or specific nucleo­phile–iodane complexes.

Even more important, within our investigations we showed the general feasibility of exploiting secondary interactions in hypervalent λ3-fluoro-iodanes to enable and control reactivity. It is noteworthy that the hypervalent F-iodane is most likely not the optimal binding partner for pyridyl Lewis bases. Further investigations on the halogen bonding of λ3-F-iodanes and harnessing these non-covalent interactions to steer the course of reactions are currently underway in our laboratory. The presented study provides a first step to a much broader understanding of the reactivity of F-iodanes, contributing to the development of a generalizable concept that allows for the dynamic prediction of product formation and the potential to expand the addressable transformations in the future.


#
#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

We thank Dr. G. M. Kiefl for the fruitful discussion during the preparation of this Account.


Corresponding Author

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

Publication History

Received: 14 November 2022

Accepted after revision: 30 November 2022

Accepted Manuscript online:
30 November 2022

Article published online:
02 January 2023

© 2022. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


   Permissions and Reprints All articles of this category
Zoom Image
Zoom Image
Zoom Image
Scheme 1 Structures of known fluorinating agents and selected examples of methods employing fluoro-iodane 1 in electrophilic fluorinations
Zoom Image
Scheme 2 (a) Chemoselective fluorinations of styryl-benzamides leading to two distinct products depending on the employed fluorinating agent. (b) Putative mechanism of the benzoxazepine formation.
Zoom Image
Scheme 3 (a) + (b) λ3-Fluoro-iodane 1 triggered conversion of styrenes 11 and 18 into benzoxazepines 13 and 19 via a fluorination/1,2-aryl shift/cyclization cascade. (c) Further transformations of benzoxazepines 19 into ketones 20 and indoles 21.
Zoom Image
Scheme 4 (a) Extension of the previously established fluorination–cyclization methods of styryl-benzamides 11 enabling the synthesis of ketones 20. (b) + (c) Fluorination–cyclizations to give indoles 21 and 22 with reversed substitution patterns.
Zoom Image
Scheme 5 An observed side reaction when employing pyridyl-styrene 18a in the established synthesis of indoles 21
Zoom Image
Scheme 6 Concepts for using λ3-iodanes to facilitate the umpolung and subsequent addition of nucleophiles α to ketones
Zoom Image
Scheme 7 A successful attempt at the direct oxidative α-umpolung of pyridyl ketones 36 employing fluoro-iodane 1
Zoom Image
Scheme 8 Exploring the reaction conditions
Zoom Image
Scheme 9 Substrate scope and limitations of the developed Umpolung reaction in terms of tolerated nucleophiles
Zoom Image
Scheme 10 Substrate scope of the developed umpolung reaction in terms of tolerated substitutions on the ketone 36
Zoom Image
Scheme 11 Mechanistic proposition containing all determined and relevant interactions
Zoom Image
Scheme 12 Mechanistic considerations and investigations mainly concerning the role of the acidic nucleophile and the pyridyl N atom