Bridge Functionalisation of Bicyclo[1.1.1]pentane Derivatives

Abstract “Escaping from flatland”, by increasing the saturation level and three‐dimensionality of drug‐like compounds, can enhance their potency, selectivity and pharmacokinetic profile. One approach that has attracted considerable recent attention is the bioisosteric replacement of aromatic rings, internal alkynes and tert‐butyl groups with bicyclo[1.1.1]pentane (BCP) units. While functionalisation of the tertiary bridgehead positions of BCP derivatives is well‐documented, functionalisation of the three concyclic secondary bridge positions remains an emerging field. The unique properties of the BCP core present considerable synthetic challenges to the development of such transformations. However, the bridge positions provide novel vectors for drug discovery and applications in materials science, providing entry to novel chemical and intellectual property space. This Minireview aims to consolidate the major advances in the field, serving as a useful reference to guide further work that is expected in the coming years.


Introduction
Over the past three decades,t he application of bicyclo-[1.1.1]pentane (BCP) derivatives in materials science as molecular rods, [1] molecular rotors, [2] supramolecular linker units, [3] liquid crystals, [4] FRET sensors [5] and metal-organic frameworks [6] has been extensively investigated and is the subject of ar ecent review. [7] Since the seminal reports of Barbachyn (1993) [8] and Pellicciari (1996), [9] theB CP motif has also emerged within drug discovery as av aluable bioisostere for internal alkynes, tert-butyl groups and monosubstituted/1,4-disubstituteda renes. [10] Thea ppeal of the abiotic BCP fragment, among related systems such as cubanes and higher bicycloalkanes,o riginates from its ability to add three-dimensional character and saturation to compounds. Increasing the fraction of sp 3 -hybridised carbon atoms in ad rug molecule,F sp 3 ,h as been found to make al ead oral drug compound "more developable" [11] and correlates positively with clinical success. [12] To this end, multiple research groups have documented the increased or equal solubility [13] / potency [8-9, 13a,d, 14] /metabolic stability [13a,c,e, 15] and decreased non-specific binding [13c,f] of lead compounds that can be achieved through such bioisosteric replacements.I ncreasing the solubility and potencyo fam edicine can reduce the therapeutic dose required, potentially avoiding drug-drug interactions and drug-induced liver injury through metabolic activation. Bioisosteric replacement has also been identified as as trategy to circumvent Markush structure patent claims on drug candidates. [16] Multiple reviews have addressed the 12 + unique synthetic approaches that have been reported for construction of the BCP framework. [7,10,17] In particular,c arbene insertion into the central bond of bicyclo[1.1.0]butanes,a nd nucleophilic/radical addition across the central bond of [1.1.1]propellanes,have emerged as the two most practical and scalable methods (Scheme 1). Methodology now exists to install most fragments of interest in drug discovery at the bridgehead (1,3) positions.Incontrast, methodology for substitution of the bridge (2,4,5) positions remains underdeveloped. [18] Most reported methods forge the key substituent bond prior to construction of the BCP core itself.F or reasons outlined in Section 2.2, examples of the direct functionalisation of the bridge positions in ac ontrolled manner are rare.
Bridge-substituted BCP derivatives represent novel chemical space and provide novel vectors for ligandtarget interactions in drug discovery. As successful drug discovery becomes more challenging year on year, [19] accessing new chemical space becomes increasingly important. Introduction of the abiotic BCP motif may also provide an ew strategy for overcoming genetic resistance to existing therapeutic compounds.B rown has recently commented on this issue, [20] highlighting that abiotic scaffolds may prove important in the discovery of anticancer,a ntibiotic, antiviral and antiparasitic drugs,w here the causative agents can be highly proficient at developing resistance to established treatments.
Aspects of the bridge functionalisation of BCP derivatives have been previously addressed by Michl and co-workers. [17a] Herein, we seek to update this work and provide asummary, from as ynthetic perspective,o ft he known methods for the preparation of bridge-substituted BCP derivatives reported up to mid-2021. Thed iscussion is organized by the nature of the substituent installed. Further distinction is drawn between strategies that install the substituent before synthesis of the BCP core,and those proceeding via direct functionalisation of the hydrocarbon skeleton.
"Escaping from flatland", by increasing the saturation level and three-dimensionality of drug-like compounds,c an enhance their potency,selectivity and pharmacokinetic profile.One approacht hat has attracted considerable recent attention is the bioisosteric replacement of aromatic rings,i nternal alkynes and tert-butyl groups with bicyclo[1.1.1]pentane (BCP) units.While functionalisation of the tertiary bridgehead positions of BCP derivatives is well-documented, functionalisation of the three concyclic secondary bridge positions remains an emerging field. The unique properties of the BCP core present considerable synthetic challenges to the development of such transformations.However,t he bridge positions provide novel vectors for drug discovery and applications in materials science,p roviding entry to novel chemical and intellectual property space.This Minireview aims to consolidate the major advances in the field, serving as auseful reference to guide further work that is expected in the coming years.  [21] kcal mol À1 for the parent hydrocarbon (cf.2 7.5 kcal mol À1 for cyclopropane [21] )a nd thermal stability up to ca. 300 8 8C. [17a,22] This strain primarily arises from significant destabilizing overlap of the rear lobes of bridgehead orbitals directed at substituent positions. [23] In many cases,t his transannular communication attenuates and directs the reactivity of the cage.

Orbital Hybridisation and CÀHB ond Strengths in the BCP Core
Theh ybridisations of bridge and bridgehead carbon orbitals used for substituent bonding on the parent hydrocarbon have been established as sp 2.5 and sp 2.0 respectively through 1 J C-H coupling constant analysis, [24] in good agreement with values obtained from ab initio [25] and semiempirical [26] calculations.T he high s character of these orbitals has two effects:1 )bicyclopent-1-yl and bicyclopent-2-yl fragments are net electron-withdrawing; [17a] and 2) BCP C À Hb onds have al arge homolytic bond dissociation energy (BDE), greater than that of typical saturated hydrocarbons.I ndeed, Maillard and Walton have demonstrated that hydrogen atom abstraction by tert-butoxyl radicals proceeds more rapidly from cyclopropane than from the parent BCP. [27] Joseph

Angewandte Chemie
Furthermore,a lthough simple consideration of hybridisation suggests that the bridge C À Hbonds should be weaker than bridgehead C À Hb onds, only the bridgehead radical is observed by EPR on exposure of BCP to tert-butoxyl radicals. This regioselectivity is remarkable considering that the bridge and bridgehead radicals have approximately equal calculated enthalpies of formation, [28] and that the bridge hydrogens have at hreefold statistical advantage over the bridgehead hydrogens.A side from the bridgehead C À Hb onds being more sterically accessible, [29] the selectivity arises from the aforementioned transannular interactions within the BCP core.The transition state for hydrogen atom abstraction by an electronegative radical is polarised, with development of partial positive charge on the adjacent carbon. During hydrogen abstraction from ab ridgehead site,t his positive charge can be extensively stabilised by the transannular bridgehead orbital interaction. [17a] In contrast, the bridge carbon does not have access to such stabilisation and so hydrogen atom transfer (HAT) from this position is kinetically inhibited. Moreover,i fo ne bridgehead position is functionalised with an electron-withdrawing group,this intracage stabilisation is compromised and so the expected selectivity of C-H abstraction from ab ridge position is returned (see Section 4.1). [30] In summary,the high BDE of BCP CÀHbonds means that ad irect HATa pproach for decoration of bridge positions would likely be confined to avery limited range of substrates; they could possess no other alkyl C À Hbonds,and blocking of unfunctionalised bridgehead positions may be necessary. More recently,attempts at directed C-H activation chemistry on the bridge positions have also proven unsuccessful. [31] Consequently,p refunctionalisation at the bridge positions is expected to be necessary for the divergent introduction of substituents.

Reactive Intermediates on BCP Bridge Positions
Theg eneration and properties of reactive intermediates (carbocations,carbanions and radicals) on the bridgehead and bridge positions of BCP derivatives have been extensively discussed by Michl. [17a] Key points relevant to the synthesis of bridge-functionalised derivatives are: 1) Bridge-centered carbenium ions are unstable,undergoing rapid and irreversible skeletal rearrangement to ringopened products (see Sections 3.1, 4.1 and 5.2). As such, they are not viable intermediates for the preparation of bridge-substituted BCP derivatives. 2) Bridge-centered carbanions are kinetically stable against ring opening, and are putative intermediates in the Haller-Bauer cleavage and Birch reduction of 2-substituted BCP derivatives (see Section 5.2). 3) Bridge carbon-centered radicals are kinetically stable against b-scission to afford cyclobutyl products,e ven though this process is calculated to be strongly exothermic. [32] 3. Oxygen Substituents

Bicyclo[1.1.1]pentan-2-ol and Derivatives
Methods for the preparation of BCP bridge alcohols are rare.A ttempts at direct C-H oxidation using persulfate salts have been unsuccessful. [30] Theparent bicyclo[1.1.1]pentan-2ol (14 b)h as been most efficiently accessed through Baeyer-Villiger oxidation of mixed ketones 12 (see Section 5.1), hydrolysis of the resulting esters and separation of the isomeric alcohols (Scheme 2A). [29] Compound 14 b has also been accessed in low yield through reduction of the corresponding ketone (see Section 3.2). The2 -substituted phenyl analogue (16 a)i sm ore accessible, [33] representing the major product of the photolysis of cyclobutyl phenyl ketone via (thermally reversible) [33a,b] Norrish-Yang (NY) [34] cyclisation (Scheme 2B). TheNYcyclisation strategy is compatible with electron-poor and electron-neutral substrates (16 b-f), but is less tolerant of electron-rich systems such as vinyl and 2-furyl cyclobutyl ketones. [35] While the lithium salt of the parent bicyclo[1.1.1]pentan-2-ol is stable in aprotic solvents (Scheme 2A), [29] increased strain at the bridge position means that 2-substituted bicyclo-[1.1.1]pentan-2-ol alkoxides are unstable.E xposure of alcohols 16 a and 17 to catalytic NaOMe in methanol results in rapid cycloreversion to the corresponding cyclobutyl phenyl ketones, [33b,36] analogously to the ring-opening of cyclopropyl alcohols. [37] Deuterium-labelling studies have identified two distinct pathways for ring opening (Scheme 3);however,both are driven by release of ring strain in the transition state. [36a] Ring opening occurs rapidly at cryogenic temperatures; attempts at O-alkylation of the alkoxide of 16 a with methyl iodide or dimethyl sulfate have been unsuccessful. [36a] Alkylation is presumably retarded by the decreased nucleophilicity of the alkoxide from the electron-withdrawing effect and steric demand of the BCP moiety.
Moreover,t he p-bromophenylcarbamate derivative of alcohol 16 a has been prepared through direct reaction with the isocyanate without prior deprotonation. [33c] Similarly, nitrobenzoyl esters 20 [38] and 21 [36a,39] have been prepared by treatment of alcohols 14 b and 16 a,r espectively,w ith the

Angewandte
Chemie appropriate benzoyl chloride.Solvolysis of these esters leads only to ring-opened products,i llustrating the instability of BCP-centered carbocations.S imilarly,t reatment of free alcohol 16 a with acetic acid [36a] produces 22 in 40 %y ield and the corresponding acetate in 60 %yield. Themechanism of rearrangement has been studied kinetically and spectroscopically. [36a, [38][39][40] It is proposed that departure of the leaving group is assisted by participation of an adjacent CÀC s-bond (Scheme 4). Rearrangement, presumably via housane cation II,provides cyclopentenyl cation III which may be trapped by nucleophiles.W hile cation I has been observed by NMR spectroscopy under cryogenic,s uperacidic conditions,c ation II was not detected. In the absence of nucleophiles, III is observed as allylic cation IV arising from formal Wagner-Meerwein rearrangement. [40] In summary,access to the parent bicyclo[1.1.1]pentan-2-ol (14 b)r emains extremely inefficient (Schemes 2a nd 5). Additionally,t he instability of 14 b and derivatives makes them challenging synthetic intermediates to handle,u nless they can be protected as solvolytically stable functionalities under strictly neutral conditions.

Bicyclo[1.1.1]pentan-2-one and Derivatives
Ketone 24 has been prepared by Dougherty through ozonolysis of alcohol 16 a. [42] NaBH 4 reduction then provided alcohol 14 b in 23 %y ield (Scheme 5). Thep oor yield of the ozonolysis reaction precludes 24 as au seful synthetic intermediate for the preparation of 2-substituted BCP derivatives.C ompound 24 is thermally unstable,u ndergoing cycloreversion at elevated temperatures to afford allylketene and derived oligomers/nucleophilic trapping products.Cycloreversion is much more facile than for typical cyclobutanones. Ketalisation of 24 has only been achieved through irradiation in CD 3 OD in the absence of base. [42] Ther elated 4,5-dimethylenebicyclo[1.1.1]pentan-2-one (30)w as later prepared by Dowd from tricyclopentane 25, [41] where generation of the highly strained BCP scaffold is compensated by extrusion of dinitrogen from unstable intermediate 29 (Scheme 6). Thei nstability of 30 similarly precludes its use as asynthetic intermediate to access bridgesubstituted BCP derivatives.

Halogen Substituents
Only bridge-fluorinated, -chlorinated and -brominated BCP derivatives are known. While fluorine and chlorine substituents have been introduced through both dihalocar-

Angewandte
Chemie bene insertion into the central bond of bicyclo[1.1.0]butane precursors and through direct reaction of BCP derivatives with the elemental halogens,b romine has only been introduced through dibromocarbene insertion. [30,43]

Bridge-Chlorinated derivatives
Thed ichlorocarbene insertion method of Cherkofsky [43a] (Scheme 1) is arguably the most convenient approach to bridge chlorine-substituted BCP derivatives.B oth chloroform [43a] and sodium trichloroacetate [13d] are viable carbene precursors.M onodechlorination is readily achieved using tin [30] or silicon [13d] hydride radical chemistry,f acilitated by stabilisation of the transition state for chlorine abstraction by the remaining halogen atom. Removal of the second halogen atom, to afford ab ridge-unfunctionalised BCP derivative,i s possible but requires forcing reaction conditions and extended reaction times.
Thed irect chlorination of BCP derivatives has been extensively investigated and represents the first reported direct functionalisation of the BCP core.P hotochemical chlorination has been demonstrated using both t BuOCl [44] and elemental Cl 2 [29] by Wiberg,with 22.5:1 and 7:1selectivity for bridgehead:bridge substitution, respectively (Scheme 7). Notably,the preference for bridgehead substitution contrasts with the direct chlorination of other bicyclic hydrocarbons (e.g. norbornane) which typically give zero or little bridgehead substitution. [29] Repeating the chlorination with Cl 2 at higher chlorine concentration increased the total yield to 50 %, with abroadly similar product distribution except with zero production of 39. [29] This result was initially taken to affirm the kinetic stability of the BCP bridgehead radical:only when the supply of Cl 2 is limited is the radical sufficiently long-lived to rearrange to give the ring-opened cyclobutane product. However,this was later disputed by Della and co-workers, [45] who argued that 39 instead arises from heterolysis of the bridgehead CÀCl bond of 1-chlorobicyclo[1.1.1]pentane and rearrangement of the resulting cation.
In 1989, Michl demonstrated the bridge 2,2-dichlorination of 1,3-disubstituted BCP derivatives with gaseous Cl 2 (Scheme 8A). [30] Bicyclo[1.1.1]pentane-1-carboxylic acid, bearing no substituent at one bridgehead site,also underwent selective bridge dichlorination. As discussed in Section 2.2, this arises from the electron-withdrawing carboxyl group decreasing the stabilisation conferred to the bridgehead radical through the 1,3-transannular interaction. No significant quantities of monochlorinated products were observed in any case:t his was rationalised through stabilisation of the Scheme 7. Direct chlorinationo fthe parent BCP hydrocarbon.

Angewandte
Chemie transition state for the second HATe vent by the installed chlorine atom. While these studies increase the synthetic utility of the previous work of Wiberg,they are still limited by the requirement for Cl 2 gas and the use of CCl 4 ,C FCl 3 and TFAsolvents.
Finally,i ts hould be noted that bridge-chlorinated BCP derivatives likely cannot be accessed through functionalisation of 2,2-dichloro[1.1.1]propellane (47). Compound 47, prepared in 1992 by Michl and co-workers, [46] was found to be highly susceptible to oligomerisation and loss of ab ridge chlorine atom, producing unstable cation 49 (Scheme 9). This loss of chlorine is consistent with that reported previously for ah igher gem-dichloropropellane. [47] By analogy with computational studies on gem-difluorinated BCP derivatives, [17a, 48] the instability of 47 could be rationalised through angular distortion of the propellane cage caused by geminal bridge substitution, which would favour an increased C-C-C valence angle at the bridge.This strain can be partially relieved either through oligomerisation to 48,orthrough loss of chloride and subsequent cage rearrangement. It is therefore postulated that other propellanes with bridge substituents that are competent leaving groups will also be unstable. [31a]

Bridge-Fluorinated Derivatives
Bridge-fluorinated BCP derivatives have garnered significant interest in the past two years.U nsurprisingly,d irect fluorination of 1,3-disubstituted BCPs with limited gaseous F 2 is extremely unselective,p roducing 14 out of the 15 possible bridge-fluorinated derivatives. [48] Exhaustive fluorination gives access to penta-and hexafluorinated derivatives in synthetically useful yields. [25b] Recent attempts [31a] at con-trolled fluorination on the BCP core through the electrochemical method of Baran, [49] and the amine-directed, Pdcatalysed approaches of Xu [50] and Hrdina, [51] have all proven unsuccessful. Only via difluorocarbene insertion have bridgefluorinated BCP derivatives been prepared in ac ontrolled fashion.
Thec oncurrent works of Mykhailiuk [31a] and Ma [52] disclosed the first examples of this chemistry in 2019, using the Ruppert-Prakash reagent (TMS-CF 3 )a nd the Dolbier reagent (TFDA, FSO 2 CF 2 CO 2 TMS) as difluorocarbene precursors,respectively (Scheme 10). Anderson has also recently demonstrated difluorocarbene insertion using az witterionic phosphonium carboxylate precursor. [53] Preliminary results towards the generation of 2-chloro-2-fluorobicyclo-[1.1.1]pentane derivatives,using CHFCl 2 as achlorofluorocarbene source,were also presented by Mykhailiuk. Across both reports,t he reaction scope was limited to aryl-or vinylsubstituted bicyclo[1.1.0]butanes,alimitation apparently not shared by the related dichlorocarbene insertion reaction. [54] However,b etween the two reports,f luorinated BCP derivatives bearing both electron-neutral and electron-poor aromatic substituents are accessible (Scheme 10). Bicyclo-[1.1.0]butanes bearing electron-rich aromatic substituents are not viable substrates on account of their instability with respect to ring opening, even at room temperature.
Thes tability of the difluorinated BCP motif to acidic, basic, reducing and forcing palladium cross-coupling conditions was also demonstrated between the two reports. Although vinyl or aryl substituents were essential for the difluorocyclopropanation, both can subsequently be oxidised to carboxylic acids [31a, 54] and further derivatised. This was later exploited in af urther publication by Ma in 2020, [55] demonstrating various transformations of gem-difluorinated BCPs at the bridgehead positions.
Af urther general limitation of dihalocarbene insertion reactions arises from the nature of the synthesis of the starting materials.A ne lectron-withdrawing group (or ah alogen) is required in intermediates 3,s oa sto facilitate deprotonation or metal-halogen exchange to enable the required intramolecular cyclisation reaction. Access to the cyclobutanol intermediates (2)i sa lso nontrivial, as the addition of sterically hindered, electron-poor and heteroaryl Grignard reagents to cyclobutanones is challenging. [52] Ma and co-workers have,h owever,i llustrated the construction of fivemembered heteroaryl rings appended to the difluoro-BCP core at al ate stage using classical carbonyl-centred ring syntheses. [55]

Carbon Substituents
BCP derivatives bearing bridge carbon substituents have been accessed in four fundamental ways:t hrough chlorocarbonylation of aB CP precursor;t hrough NY cyclisation of cyclobutyl aryl ketones;f rom substituted [1.1.1]propellanes; and from intramolecular nucleophilic displacement on asubstituted cyclobutane.T he synthesis of tricyclo-[2.1.0.0 2,5 ]pentanes (where two bridge positions on the BCP core are directly bonded), and miscellaneous synthetic routes that afford bridge carbon-substituted BCP derivatives as minor products,a re beyond the scope of this review.B oth topics are addressed by Michl. [17a] Thei nstallation of carbon substituents on the BCP core through intermolecular carbene insertion into ab icyclo-[1.1.0]butane (Scheme 1) has not yet been demonstrated, although there are as mall number of reports of ar elated intramolecular reaction providing [1.1.1]propellanes (see Section 5.3). Attempts at intermolecular delivery of aC H 2 fragment via photolysis with diazomethane [56] and under Simmons-Smith conditions have been unsuccessful. [13d, 57] Attempted reaction of bicyclo[1.1.0]butanes with bis(carbomethoxy)carbene was also unsuccessful. [58] Experimental and computational studies of the mechanism of carbene insertion into bicyclo[1.1.0]butanes have been previously discussed. [17a, 58]

Carbon Substitution through Direct Chlorocarbonylation
Acyl chloride 11 b represents the first example of aB CP derivative bearing ab ridge carbon substituent. Thep roduct distribution of chlorocarbonylation of 31 with oxalyl chloride corresponds to as electivity of 17:1 for bridgehead:bridge substitution. As described previously,manipulation of mixed acyl chlorides 11 also gave first access to bicyclo[1.1.1]pentan-2-ol (14 b)( Scheme 11).

Carbon Substitution through Norrish-Yang (NY) Cyclisation
Thes ynthetic utility of the NY cyclisation of cyclobutyl phenyl ketone (Section 3.1) was explored by Alexander in the 1970s [59] and later in 1993 [35b] (57). Presumably the generation of an intermediate benzylic anion promotes formation of 57 rather than the bicycloalkyl carboxamide,t he typical product of the Haller-Bauer reaction. Compound 57 was more conveniently accessed by Wiberg [35b] through acylation of 16 a and subsequent reduction with metallic sodium.
Direct deoxygenation of 16 a with Li/NH 3 was unsuccessful due to rapid cycloreversion to cyclobutyl phenyl ketone (15 a)u nder basic conditions (see Section 3.1). Reductive deoxygenation of 16 a using Et 3 SiH/BF 3 ·Et 2 Oa nd halogenative deoxygenation under Appel conditions were also unsuccessful. [35b] Both resulted in fragmentation of the BCP core,f urther corroborating the instability of the putative bridge carbocation. Both Alexander and Wiberg demonstrated oxidation of the aromatic ring of 57 to the carboxylic acid, which provided access to 2-benzoylbicyclo[1. Across two studies in 2005 [60] and 2012, [61] Xia, Yang et al. reported findings from as tudy of enantioselective photochemical solid-state NY cyclisations of ketones 63.Asymmetry originates from the homochiral environment of the crystalline lattice of the solid salts.Ketones 63 were prepared from alcohols 62,presumably originating from carboxylic acid 59 or ad erived ester.U ltraviolet irradiation of the solid ketones gave the corresponding tricyclic alcohols 64 with modest enantioselectivity (maximum 90 % [60] and 60 % [61] ee). Extended reaction times and/or higher reaction temperatures were necessary to achieve satisfactory conversion in the solid state;h owever,i ncreased conversion was generally to the detriment of asymmetric induction.  [62] allows opening of the central bond with ab road variety of carbon-and heteroatom-centered radicals and nucleophiles (treatment with electrophiles [62] or transition metal catalysts [63] affords exclusively ring-opened products). As such, [1.1.1]propellanes could represent key divergent intermediates for the library synthesis of BCP derivatives in drug discovery.Avariety of bridge-substituted [1.1.1]propellanes have been prepared, including 2,2'-bonded dimers and those featuring spirocyclic attachment of another ring, either through apolar cyclisation strategy [64] or through intramolec-Scheme 11. Preparationo fa cyl chlorides 11 a,b through photochemical chlorocarbonylation of the parent BCP hydrocarbon.

Angewandte Chemie
Minireviews ular carbene insertion. [65] However,m any of these substituents are simple hydrocarbons or contain few functional handles for further derivatisation.
Furthermore,r elatively few substituted propellanes have then been opened to afford monomeric bicyclo[1.1.1]pentane derivatives. [64a,d-h, 65a, 66] Some have been polymerised to afford [n]staffane chains;t his has been recently reviewed by Bräse. [67] Co-polymerisation of substituted propellanes with acrylates has also been studied, as well as the preparation of bridge-substituted-BCP-containing polyamide polymers. [67] Simple examples of the functionalisation of bridgesubstituted propellanes were provided by Szeimies, [ resulted in exclusive halogenation of the two-carbon bridge over the BCP core, consistent with the BDE trends described previously (Section 2.2). Tr eatment of the obtained mixed chlorides with sodium amide furnished olefin 70,w hich underwent ozonolysis in high yield to afford cis-endo-tetrasubstituted BCP derivative 71.C ompound 71 then served as ac ommon precursor for the synthesis of four other tetrasubstituted analogues (72-75)t hrough conventional carbonyl manipulations.A cid 73 spontaneously formed anhydride 76 during isolation, which went to completion on heating above 80 8 8C for several hours.
In 1993 and 1995 reports,S chlüter and co-workers reported the synthesis of five propellanes containing ap rotected and free hydroxymethyl functionality (Scheme 14). [64d,e] These were accessed through amodified Szeimies procedure employing functionalised olefin 86.A side from the synthesis of [n]staffane polymers,t he synthetic chemistry of these propellanes would surprisingly not be explored until late 2020 in the works of Ma [68] and Baran. [31b] Thew ork of Ma delineated the synthesis of analogues of propellane 88 through as imilar method. Theg enerated propellanes (95)w ere then subjected to the strain-release amination conditions of Baran, [69] furnishing 1-amino-2-functionalised BCP products (96)( Scheme 15). MOM, TBS and Scheme 12. NY cyclisation of cyclobutyl phenyl ketone and manipulation of resulting alcohol 16 a.The synthesis of compounds 14 b/24 and 60 (described in Sections 3.1 and 6, respectively)a re also included for clarity." quant" = quantitative;R*= enantiopure chiral fragment.

Angewandte
Chemie benzyl O-protecting groups were all compatible with the synthetic sequence.Adioxolanylmethyl substituent at C2 was also tolerated. As mall range of secondary amines were proficient substrates,i ncluding allylic/benzylic amines and piperidine derivatives;h owever, sterically hindered amines, cyclopropylamines and pyrrolidine derivatives were unsuccessful. After methoxymethyl cleavage,the alcohol functionality of bicyclopentane 97 was derivatised to the corresponding fluoride,a zide,t hioacetate and N-(4-fluorophenyl)carbamate under standard conditions. These studies represent the first peer-reviewed investigation into aconvenient and general strategy for the synthesis of bridge-substituted BCP derivatives.H owever,t he chemistry is currently limited by its poor performance with more complex amines and the difficulty of achieving simultaneous 1,3-disubstitution across the propellane:preliminary attempts at aminoalkylation under the conditions of Gleason, [70] and aminoalkenylation under the Zweifel conditions of Aggarwal, [71] were unsuccessful and extremely low-yielding, respectively.F or the bioisosteric replacement of an ortho-or metasubstituted benzenoid fragment this may not be necessary,but having equal access to both 1,2-disubstituted and 1,2,3trisubstituted BCP derivatives would greatly increase the synthetic utility of the chemistry and broaden the chemical space it can access.Restricting studies to the synthesis of 1,2disubstituted BCP derivatives can arguably never fully capitalise on the three-dimensionality of the BCP core.
Compounds 103 and 104 were further used to prepare analogues of six drugs containing ortho-o rmeta-substituted benzene rings,including MTP inhibitor lomitapide (107), Hh signalling inhibitor sonidegib (109)a nd histamine H 1 antagonist meclizine (111) (Figure 1). Forthe latter and three other Fort he direct appendage of aromatic rings to the BCP core,c onversion of BCP carboxylic acids to N-hydroxyphthalimide (NHP) esters permitted application of the groupst hermal decarboxylative Negishi arylation chemistry (generalised in Scheme 17).
In summary,t he work of Baran broadens the range of synthetic transformations known for decoration of the BCP core,a nd provides the first examples of bridge-substituted BCP analogues of drug compounds.F uture work will help elucidate the ortho vs. meta character of 1,2-disubstituted BCPs in this context. Thew ork confirms the viability of thermal decarboxylation of redox-active esters on the bridge positions of BCPs (previously demonstrated only on the bridgehead positions), but the full synthetic potential of the generated radical is yet to be explored. As discussed, future work focusing on 1,2,3-trisubstitution of the BCP system will further increase the value of the developed methods within drug discovery.

Carbon Substitution through Intramolecular Nucleophilic Displacement
In early 2021, Qin presented an elegant contemporary adaptation (Scheme 18 B) [73] of the original intramolecular displacement methodology employed in the first synthesis of bicyclo[1.1.1]pentane (Scheme 18 A). [44,74] Cyclobutanones 118 were accessed through the groupsb oron-preserving cross-coupling of aldehydes 117 followed by ketal hydrolysis. Conversion of 118 to mesitylsulfonyl hydrazones,followed by addition of Cs 2 CO 3 ,furnished intermediate diazo compounds 119.T rapping of the pendant boron atom delivered highenergy zwitterionic intermediates 120,w hich rearranged to afford 2-substituted bicyclo[1.1.1]pentanes 121 via 1,2-metalate rearrangement with extrusion of dinitrogen. Akey benefit of the developed process lies in its robustness:i nc ontrast to reactions employing propellanes,t his reaction can be performed in air with only asmall decrease in yield. Additionally, as stoichiometric water is generated during sulfonylhydrazone formation (Scheme 18 B, step ii)), the reaction is entirely tolerant of trace water in reagents and solvents.
Ar ange of secondary BPin esters were viable substrates, providing bridge alkyl-(122)a nd cycloalkyl-(123-125) substituted BCP derivatives.Aprimary BPin ester could also be cyclised to afford the bridge-unsubstituted product. Te rtiary BPin esters also successfully underwent cyclisation, providing geminal bridge-disubstituted products (e.g. 126). Heteroaryl (125), isopropyl ester (127 a), vinyl (127 b), ethynyl (127 c), methyl, amide and carbamate groups were tolerated as substituents on the cyclobutane ring (R). A positive correlation was identified between the steric demand of the "R" substituent and the reaction yield:t his is aconsequence of increased preference for the BPin-containing side chain to adopt the pseudo-axial position required for cyclisation. This requirement means that cyclobutanones with small "R" substituents are not viable substrates:c ompounds 128 a,b,with R = BPin (A-value = 0.42 kcal mol À1 ) [75] and R = H, respectively,w ere not accessible using this methodology. Thei ntroduction of ab ridge methyl substituent was also demonstrated through a-substitution on the cyclobutanone precursor,r ather than substitution adjacent to the boron centre.Finally,asymmetric borylation under the conditions of Aggarwal provided an enantioenriched secondary BPin ester cyclisation precursor,w hich was cyclised with only 3% erosion of enantiopurity to afford BCP derivative 129.T o our knowledge,t his represents the first example of asym- TheB Pin group of compound 123 was derivatised to hydroxyl, vinyl and heteroaryl functionalities under standard conditions.M atteson reaction [76] provided the homologated primary BPin ester,w hile conversion to the trifluoroborate salt further enabled proto-deboronation and Suzuki crosscoupling reactions.T he trifluoroborate salt could also be employed in another round of boron-preserving cross-coupling chemistry with ak etone,p roviding another means of formal homologation of the CÀBb ond.
In summary,t he work of Qin provides ac onceptually distinct method for the synthesis of bridge-substituted BCP derivatives.T he developed methodology is operationally simple and can deliver nonsymmetrical 1,3-di-, 1,2,3-tri-, and 1,2,2,3-tetrasubstituted BCP products containing versatile functional handles for downstream derivatisation. However,alimitation is arguably its reduced divergencycompared to the methods of Baran and Ma, wherein a[1.1.1]propellane provides an early common intermediate to potentially hundreds of unique derivatives.W ith the present method, bespoke cyclisation precursors must be synthesised (asymmetrically,i fr equired) for each fundamental series of compounds,l imiting its applicability to late-stage functionalisation. Only the BCP products post-cyclisation represent points of divergency.

Nitrogen Substituents
Surprisingly,n one of the works of Baran, Ma or Qin considers the synthesis of BCP derivatives containing bridge nitrogen substituents.T his is in contrast to the considerable progress made towards installation of nitrogen atoms at the bridgehead positions, [17b,69, 70,77] for example,f or bioisosteric replacement of anilines,c ommon structural alerts and toxicophores in drug design. To date,n itrogen has only been installed on the bridge positions of BCP derivatives through Schmidt rearrangement of carboxylic acid 59 (Schemes 12 and 19) by Wiberg. [35b] Alexander previously reported the same transformation in 1978; [59b] however, this claim was not substantiated by characterisation data for the amine product. As expected, the electron-withdrawing BCP core decreases the basicity of 130 compared to other cycloalkylamines.
Peroxyacid oxidation of 130 then furnished the corresponding nitro compound (60)i n2 4% yield. [35b] Like nitrocyclopropane and in contrast to nitrocyclopentane and nitrocyclohexane, 60 was found to be insoluble in aqueous NaOH. This is attributed to its increased pK a compared to other nitrocycloalkanes,aconsequence of the reluctance of the bridge carbon atom to engage in p-bonding with associated increase in strain.

Boron Substituents
As part of their synthetic campaign, Baran and co-workers aimed to establish ar oute to ab oron-containing BCP as ah igh-value building block. [31b] After propellane 88 was opened with benzyl iodoacetate,b orylated BCP 134 was accessible via the groupsc opper-catalysed decarboxylative Themethodology of Qin [73] also provided BCP derivative 133 containing ab ridge BPin substituent (Scheme 21). Likewise,notransformations of the bridge functionality were then explored.

Summary and Outlook
Thew orks of Baran, Ma and Qin highlight two general synthetic strategies with potential to address the long-standing question of the preparation of bridge-functionalised BCP derivatives.T he use of as ubstituted [1.1.1]propellane intermediate could facilitate divergent synthesis through the exploitation of its broad and ambiphilic reactivity profile. Thea pplication of decarboxylative radical transformations, dozens of which have now been reported, [78] adds af urther dimension of divergency.T his is in stark contrast to the prior art in this field, which has been sporadic and largely reliant on inconvenient and low-yielding synthetic methods.
With decarboxylative radical transformations on the BCP core successfully demonstrated, in the near future we are likely to witness the pursuit of ag reater range of related transformations.T he application of enabling technologies such as photoredox catalysis and synthetic electrochemistry will likely be central to this development. As such transformations often proceed under very mild conditions, [79] the late-stage bridge functionalisation of BCP derivatives may become commonplace.Judicious selection of different radical precursor groups,orthe identification of synthetic sequences that allow the installation of asingle radical precursor group at different stages,may also prove to be an effective strategy for the functionalisation of multiple bridge positions.I n addition, the deployment of dual catalysis platforms such as metallaphotoredox catalysis,organophotoredox catalysis and mediated electrochemistry,c ould provide access to bridge substituted BCPs in an asymmetric fashion. This would provide ac omplementary approach to that illustrated by Qin, and would not require the synthesis of enantioenriched precursors.
Efficient asymmetric access to densely functionalised bicyclo[1.1.1]pentanes is likely to be transformational within drug discovery.T he inclusion of such moieties into lead compounds increases their 3D character and stereochemical complexity without significantly increasing molecular weight, allowing ad eeper exploration of chemical space and the design of molecules which can better complement the spatial subtleties of target proteins. [12b,80] Increased potencya nd selectivity of lead compounds may reduce off-target effects: drug promiscuity commonly results in adverse toxicity,amajor cause of drug attrition in the clinic. [12b] "Escape from flatland" promises to be ap owerful solution to this widespread and intractable problem:weare apparently on the verge of anew era in the synthetic chemistry of the bicyclo[1.1.1]pentane core.