Acid–Base Free Main Group Carbonyl Analogues

Abstract Main group carbonyl analogues (R2E=O) derived from p‐block elements (E=groups 13 to 15) have long been considered as elusive species. Previously, employment of chemical tricks such as acid‐ and base‐stabilization protocols granted access to these transient species in their masked forms. However, electronic and steric effects inevitably perturb their chemical reactivity and distinguish them from classical carbonyl compounds. A new era was marked by the recent isolation of acid–base free main group carbonyl analogues, ranging from a lighter boracarbonyl to the heavier silacarbonyls, phosphacarbonyls and a germacarbonyl. Most importantly, their unperturbed nature elicits exciting new chemistry, spanning the vista from classical organic carbonyl‐type reactions to transition metal‐like oxide ion transfer chemistry. In this Review, we survey the strategies used for the isolation of such systems and document their emerging reactivity profiles, with a view to providing fundamental comparisons both with carbon and transition metal oxo species. This highlights the emerging opportunities for exciting “crossover” reactivity offered by these derivatives of the p‐block elements.


Introduction
Carbonyl compounds (R 2 C=O) containing the C=Ofunctional group are ubiquitous in organic chemistry and include aldehydes,ketones,esters,carboxylic acids,amides,ureas etc. (Scheme 1). Their prevalent nature is underscored by the thermodynamic stability of the C=Od ouble bond, which uniquely features s (392 kJ mol À1 )a nd p (399 kJ mol À1 ) components of approximately equal strength. [1] Nevertheless, the charge disparity within the C = Omotif (Table 1) [2] induces polarization in the sense C d+ À O dÀ ,w hich, coupled with the sterically open environment, facilitates nucleophilic attack at carbon and electrophilic attack at oxygen. Such reactions can be reversible due to energetically favourable regeneration of the C = O p bond (e.g.a ddition-elimination reactions at carbon through at etrahedral intermediate,p rotonation at oxygen etc.). Hence,t he C=Of unctionality is au nique platform that displays rich chemistry.A ss uch, they represent indispensable chemical building blocks and are cornerstones of organic synthesis.
Ar ecent aspiration in synthetic chemistry has been to incorporate p-block elements into classical organic molecules to construct main group analogues with diverse structural and reactivity profiles. [3] Thus,b ys ubstituting carbon with apblock element (E = groups 13 to 15), isoelectronic main group carbonyls of the form R 2 E=O( Ty pe I)c an be conceived (Scheme 1). Alternatively,replacing an Rgroup with aneutral donor Lgenerates main group carbonyls of the form R(L)E= O( Ty pe II), which conceptually bear some resemblance to acylium ions.W ith that said, main group carbonyls of these types (I and II)c ontaining terminal E = Od ouble bonds are thermodynamically unstable species,i ns tark contrast to classical carbonyl compounds.T he greater electronegativity difference between the main group element and oxygen, coupled with the weaker p overlap,l eads to pronounced polarization of the E = Om otif,r esulting in substantial contribution of the ylidic form E + À O À .T his inability to quench the charge disparity by p bond formation is reminiscent of frustrated Lewis pair (FLP) systems, [4] in which steric constraints impede formation of a s covalent bond. Hence, the significant "frustration" within the E=Ofragment and its Main group carbonyl analogues (R 2 E = O) derived from p-block elements (E = groups 13 to 15) have long been considered as elusive species.P reviously,employment of chemical tricks such as acid-and base-stabilization protocols granted access to these transient species in their masked forms.H owever,e lectronic and steric effects inevitably perturb their chemical reactivity and distinguish them from classical carbonyl compounds.Anew era was marked by the recent isolation of acid-base free main group carbonyl analogues,ranging from alighter boracarbonyl to the heavier silacarbonyls,p hosphacarbonyls and agermacarbonyl. Most importantly,their unperturbed nature elicits exciting new chemistry,spanning the vista from classical organic carbonyl-type reactions to transition metal-like oxide ion transfer chemistry.Inthis Review,wesurvey the strategies used for the isolation of such systems and document their emerging reactivity profiles,w ith aview to providing fundamental comparisons both with carbon and transition metal oxos pecies.This highlights the emerging opportunities for exciting "crossover" reactivity offered by these derivatives of the p-blocke lements.
sterically exposed position renders them highly reactive and prone to self-quenching processes (e.g.v ia head-to-tail oligomerization, C À Ha ctivation). As such, main group pblock carbonyl analogues are highly elusive species,o ften regarded as lab curiosities,a nd their chemistries have been little developed until recently.
Inspired by the rich chemistry of the C=Of unctional group,m ain group chemists have taken up the challenge to synthesize p-block mimics (Type I and II). Initial attempts were aimed at their in situ generation and chemical trapping. Amajor development in this respect was the employment of external acid and base-stabilization of the E=Of unctional group,w hich has enabled the isolation of bottleable main group carbonyl analogues from across the p-block. [5] However,such methods bring with them inevitable electronic and steric perturbation of the E = Of unctionality which distinguishes them from the classical C = Of unctional group.
In 2012, Tamao,Matsuo et al. reported the isolation of the landmark acid-base free germanone (R 2 Ge=O), representing a" genuine" germanium analogue of ak etone (Scheme 2). [6] This compound has subsequently inspired interest in the synthesis of related group 14 systems.I np articular, the corresponding silicon analogue has received significant attention due to its position as the lightest "heavy carbonyl". More than 100 years ago,Kipping attempted to synthesize silanones (R 2 Si=O), [7] producing instead what were later shown to be polysiloxanes (R 2 SiO) n ,a nd leading to the genesis of ak ey class of industrial polymers.Assuch, the isolation of adiscrete monomeric silacarbonyl (or silanone,R 2 Si=O) remained elusive for more than 100 years.Kippingsdream was fulfilled in 2017, when Kato et al. reported the breakthrough isolation of room temperature-stable silacarbonyl species featuring the "free" Si = Om otif. [8] Three months later, the groups of Inoue [9] and Kato [10] independently reported stable acyclic silacarbonyls and ab ora-ylide substituted silacarbonyl, respectively.T his work inspired other efforts from across the Periodic Table and soon after,i n2 018, Dielmann et al. isolated ab ase-free phosphacarbonyl analogue,t hat is,a n oxophosphonium ion, [R 2 P = O] + ,w hich is isoelectronic with the silacarbonyl. [11] In 2019, we reported the discovery of al ighter carbonyl analogue,i .e.b oracarbonyl, in the form of an acid-free anionic oxoborane [R 2 B=O] À ,r epresenting an entry point to unperturbed group 13 carbonyl analogues. [12] Within the same month, Iwamoto et al. reported aremarkable dialkylsilanone featuring a" genuine" Si = Od ouble bond. [13] In this Review,wesurvey main group carbonyl analogues (Type I and II)from across the p-block elements (Scheme 1). Thed iscussion is ordered according to position within the Periodic Table and includes (for each group) as ummary of the evolution of the field through an overview of preceding acid-and base-stabilized systems,b efore focusing on recent milestones in the isolation of acid-base free main group carbonyl analogues.W efocus on 1) their syntheses,i solation and stabilization strategies,2)the nature of multiple bonding as reported by structural, spectroscopic and DFT probes,and Ying Kai Loh is originally from Singapore and is an A*STAR scholar.H eobtained his BSc (Hons) in Chemistry from Nanyang TechnologicalU niversity (NTU) in 2015. He was awarded the Gold Medal in the Global Undergraduate Awards (2015) for research on multiply bonded main group systems. He received his DPhil in 2020 from the University of Oxford under the supervision of Prof. Simon Aldridge. He will soon embarkon postdoctoral research with Prof. Guy Bertrand at the University of California, San Diego. Simon Aldridge is originally from Shrewsbury (UK) and is Professor of Chemistry at the University of Oxford, and Director of the EPSRC/Oxford Centre for Doctoral Training in Inorganic Chemistry for Future Manufacturing. He has published more than 210 papers and is apast winner of the RSC's Main Group Chemistry (2010) and Frankland Awards (2018). His research interests include the development of p-block compounds with unusual electronic structure and their applications in small molecule activation and catalysis. 3) "unmasked" chemical reactivity and potential future applications.A ll examples are structurally authenticated unless otherwise stated. Formally triply-bonded main group monoxide analogous to CO are not included here.F or clarity, main group carbonyls are labelled N (N = number) or N(X) to account for simple variation of substituent X (X = R, LA, L etc.), precursors are labelled PN (P = precursor) and products of simple addition reactions are labelled N-S (S = small molecule).
While traditional Reviews pertaining to main group element chalcogen multiple bonds have tended to focus on elements in aparticular group, [5] we hope that the horizontal approach taken by this Review across the p-block might introduce an ew perspective.I np articular, recent milestone achievements in taming otherwise highly elusive main group carbonyl analogues in their unperturbed forms present au nique opportunity to use the "free" E = Om otif as ab asis for lateral comparison of elements across groups 13, 14 and 15 of the p-block, which might provide new insight into the nature of main group E=Od ouble bonds,a nd ultimately illuminate similarities and differences with classical C=O chemistry.

Group 13 Carbonyl Analogues
Group 13 elements have avalence electron count of three, and thus are typically tricoordinate species featuring atrigonal planar geometry.T heir reactivity profile is dominated by the presence of af ormally vacant porbital, making them archetypal Lewis acids.H ence,m ultiply-bonded species of the type RÀE=Of eaturing strongly Lewis acidic and basic sites adjacent to each other have am arked tendency to oligomerize in head-to-tail fashion. [5a] While these oxoboranes (R À B = O) and mono-alumoxanes (R À Al = O) might be regarded as carbonyl analogues with formal E = Od ouble bonds,t heir dicoordinate nature and the potential to engage in af urther (donor/acceptor) interaction with the terminal oxygen distinguishes them from classical carbonyl compounds.I ndeed, the microwave spectrum of gaseous HBO reveals alinear geometry,a nd aB À Obond length of 1.20 , that is confirmed computationally to indicate aB Ot riple bond. [14] Hence,R À E = Ospecies might be thought of as being most closely related to acylium ions (RÀCO + ). While "free" RÀE=Os pecies are hitherto unknown in the condensed phases,employment of neutral donor ligand Lhas facilitated isolation of neutral boracarbonyls and alumacarbonyls of the form R(L)E = O( Ty pe II). Alternatively,a nionic R À ligands have enabled isolation of anionic boracarbonyls and alumacarbonyls of the form [R 2 E=O] À (Type I), which are also isoelectronic with carbonyls.S uch systematic strategies (employing Lo rR À ligands) bridge the gap between group 13 elements and carbon by generating isolable group 13 analogues of classical organic functional groups. This notion is perhaps best exemplified by the group 13 alkene analogues,t hat is,n eutral R(L)E = E(L)R and dianionic [R 2 E=ER 2 ] 2À diborenes [15] and dialumenes. [16] Here,we present the evolution of doubly-bonded group 13 carbonyl analogues,f rom their initial isolation as acid-base stabilized entities,t oh ighly reactive dimer-stabilized alumacarbonyls and af irst acid-free boracarbonyl. Ar eview article by Inoue et al. on multiply-bonded group 13 element-chalcogen systems documents developments in the field up till 2016, and this material will not be extensively considered here. [5a]

Organoboron Oxides
Boron is unique as it is the only main group p-block element that is lighter than carbon, making boracarbonyls the only lighter analogues of carbonyl compounds.T hese analogies are also highly pertinent as boron and carbon are neighbours within the second period in the Periodic Table. Boron also has ag reat affinity for oxygen, affirmed by the thermodynamically strong B À Ob ond (809 kJ mol À1 ). [17] Its highly oxophilic nature has been widely exploited in organic chemistry to drive industrially important reactions,n otably the Suzuki-Miyaura coupling which garnered the Nobel Prize in 2010. [18] This reaction benefits from the use of bench-stable and non-toxic organoboron oxides (RB(OR) 2 )asequivalents for otherwise highly reactive carbanions,a nd their stability can be attributed to the presence of robust B À Olinkages.As such, these tricoordinate organoboron oxides containing B À O single bonds are firmly established as an indispensable building block in the organic chemistst oolbox. Although their intrinsic stability is partly attributed to ad egree of multiple bonding between boron and oxygen, well-defined organoboron oxides featuring formal B = Odouble bonds (i.e. boracarbonyls) have for al ong time remained elusive.
In the 1930s,itwas discovered that dehydration of boronic acids (RB(OH) 2 )yields not the simple monomeric oxoborane species (RÀB=O), but stable boroxines (RBO) 3 containing ac entral B 3 O 3 ring. [19] Oxoboranes were postulated to be generated as short-lived intermediates that rapidly cyclotrimerize to the corresponding boroxines.S ince then, significant effort has been focused on detecting these fleeting species in the gas phase and in low-temperature matrices. [5a] Pioneering studies by West demonstrated the extreme reactivity of the in situ generated Mes*ÀB=Om olecule through as eries of trapping experiments. [20] In this case, kinetic protection from the sole bulky substituent proved insufficient to tame the reactive oxoborane species.W hile these experiments gave preliminary evidence for their existence,f urther studies on this interesting compound class are very limited.

Acid-Stabilized Boracarbonyls
Neutral analogues:R (L)B=O( Type II). Ab reakthrough in the quest for an isolable oxoborane came in 2005 when Cowley et al. reported the stabilization of am onomeric BO fragment in 1 by simultaneous acid-base coordination (Scheme 3). [21a] This protocol delivers at rigonal planar boron centre featuring aB = Od ouble bond, structurally reminiscent of ac arbonyl compound. As such, 1 can be regarded as the first acid-stabilized boracarbonyl. X-ray diffraction analysis revealed ashort BÀObond of 1.304 (2) , establishing the notion of aB = Odouble bond, albeit capped by aLewis acid. While DFT calculations revealed that AlCl 3 coordination increases the B À Ob ond length by only 1.9 % compared to the (hypothetical) acid-free analogue,t he B=O p bond for 1 is significantly lower in energy (HOMOÀ16) as compared to the acid-free model (HOMOÀ6). Hence,w hile the Lewis acid coordination strategy is effective in isolating an ovel boracarbonyl 1,t he inherent electronic perturbation might be expected to moderate the reactivity of the B = O functionality.
In 2011, Cui et al. reported that boron-based Lewis acids such as B(C 6 F 5 ) 3 can also be employed to stabilize aneutral bdiketiminate-derived boracarbonyl, 2(BCF) (BCF=B(C 6 F 5 ) 3 ; Scheme 3). [21b] Although the BÀOH containing precursor does not spontaneously isomerize to aB = Od ouble bond (which stands in contrast with facile enol-keto tautomerization between C À OH and C = O), O-to-C(ligand) proton migration can be induced by Lewis acid coordination at oxygen.
In 2017, Rivard et al. reported that heating IDipp·BCl 2 -(OSiMe 3 )i nt he presence of Lewis acids (LA) such as B(C 6 F 5 ) 3 or BAr F 3 (Ar F = 3,5-(CF 3 ) 2 C 6 H 3 ) 3 leads to the liberation of ClSiMe 3 to afford Lewis acid-stabilized neutral bora-acyl chlorides 5(LA) (LA = B(C 6 F 5 ) 3 or BAr F 3 )( Scheme 3). [21e] Notably,t he B = O p*orbital of 5(BCF) can be located in the LUMO,incontrast with other systems,inwhich the corresponding orbital is located at higher energy.T his suggests highly electrophilic character,consistent with classical acyl chlorides.H owever,c arbonyl-like reactions exploiting functionalization of the labile B À Cl bond of 5(BCF) (to yield B À Ho rB À Rb onds) were unsuccessful, presumably ac onsequence of steric overcrowding, and electronic perturbation of the reactive B=Ou nit by the Lewis acid. On the other hand, successful Cl-for-OH hydroxylation was achieved using HOSiPh 3 to generate the corresponding Lewis acidstabilized neutral boracarboxylic acid 5(BCF)-HOSiPh 3 . DFT analysis revealed that the mechanism for this transformation involves at etrahedral boron intermediate which collapses via release of ClSiMe 3 ,r esembling the additionelimination mechanism of classical nucleophilic acyl substitutions for carbonyl compounds.
In 2019, Gilroy et al. reported af ormazanate-based neutral boracarbonyl 6 stabilized by AlCl 3 (Scheme 3). [21f] Remarkably,examination of the photoluminescent properties revealed that 6 exhibits as mall Stokes shift (50 nm) and has aphotoluminescence intensity enhancement of more than 36fold (F PL :3 6%), in comparison with P6,w hich has large Stokes shift (174 nm) and was essentially found to be nonemissive in solution (F PL : < 1%). DFT analysis revealed that P6 is highly bent in the ground state and perfectly planar in the excited state,w hereas the molecular geometry of photoexcited 6 undergoes little structural distortion and resembles its ground state.H ence,i tw ould seem that formation of the exocyclicB =O p bond in 6 significantly improves the rigidity of the system, mitigating non-radiative decay,t hus reducing the Stokes shift and turning on photoluminescence.This work opens up opportunities for main group carbonyls featuring rigid E=Od ouble bonds to be exploited in the design of materials with turn-on photoluminescence properties.  This reactivity is analogous to the chemistry of the isoelectronic CO 2 molecule,w hich is commonly employed as aC 1 source to access carboxylic acid derivatives.T he anionic boron dioxide motif in 7(L) (L = IPr 2 Me 2 /DMAP) is generated via an unusual extrusion of 2,3-dimethyl-2-butene from the Bpin moiety;DFT studies revealed that the release of 2,3dimethyl-2-butene from [Bpin] À to generate [BO 2 ] À is highly energetically favourable (À333 kJ mol À1 ). Interestingly [21b] X-ray diffraction analysis revealed 2(H-IPr 2 Me 2 ) to contain as hort B À Ob ond (1.296 (3) )a nd astrong hydrogen bonding interaction between the imidazolium moiety and the B=Od ouble bond, representing aBrønsted acid-stabilized anionic boracarbonyl isoelectronic with urea.
In 2019, Wang et al. reported an imidazolium-stabilized anionic boracarbonyl 8 supported by a b-diketiminate ligand featuring aC 6 fused ring across C1ÀC2 of the backbone, which was accessed via three-fold deprotonation of acationic borinic acid P8 with I t Bu (Scheme 4, middle). [22a] More recently,i n2 019, Kong et al. reported au nique hydrogen bond-stabilized boracarboxylic acid anion 9 (Scheme 4, bottom), featuring two sets of hydrogen bond interactions, 1) involving ab oronic acid as ab ifurcated hydrogen bond donor to the B = Of ragment (interaction energies:1 60 and 91 kJ mol À1 ); and 2) involving an imidazolium ion acting as as ingle hydrogen bond donor to the BÀOH fragment (interaction energy:31kJmol À1 ). [22b] While hydrogen bonding interactions with C=Ofragments have been widely exploited in organocatalysis or molecular recognition, employment of isoelectronic B = Of ragments for similar applications can be anticipated.
At ransition metal-stabilized boracarbonyl. In 2013, Yamashita et al. reported ad iamino boronato ruthenium complex 10,i nw hich ar igid pincer scaffold enforces an unusually bent B-O-Ru angle (93.8 (3)8 8), thereby weakening the O(pp) À Ru(dp)i nteraction and enhancing the B À O p bond (Scheme 5). [23] Accordingly,t he B À Ob ond is short (1.329 (6) ), which, taken together with aW BI of 1.04, indicates acertain degree of B=Odouble bond character.As such, 10 can be considered as aboracarbonyl anion stabilized by coordination to ruthenium. Carbonyl compounds are well documented to adopt either h 1 binding mode through oxygen or h 2 side-on binding. [24] In the case of 10,t he geometrical imposition of the B = Od ouble bond held by the rigid pincer scaffold hints at the possibility of an h 2 binding mode. However,t he RuÀBb ond (2.608 (5) )i sm arkedly longer than the sum of the covalent radii of Ru and B(2.10 )and is therefore more consistent with an h 1 binding mode through oxygen. This situation hints at the opportunity for fine-tuning the substituents on boron to bias an h 2 coordination mode, which is hitherto unknown for B=Of ragments.C onversely, one can also envisage harnessing the potential of the highly polarized B=O p bond inherent in the [(R 2 N) 2 BO] À unit to design strongly donating h 1 O-donor ligands.

An Acid-Free Boracarbonyl:T he First Lighter Carbonyl
In 2019, we reported an acid-free boracarbonyl 11 (Type I), representing the first lighter carbonyl analogue (Scheme 6, top). [12a] Its unique stability can be attributed to two factors,1)the 6p aromatic diazaborole framework which reduces the inherent Lewis acidity of the boron centre,a nd 2) encapsulation of the potassium counter-ion by [2.2.2]cryptand, freeing it from the coordination capabilities of the strongly basic oxygen atom. Starting from the borinic acid P11,d eprotonation with K[N(SiMe 3 ) 2 ]g ave rise to dimerstabilized boracarbonyl K 2 [12] 2 ,a nd subsequent sequestration by [2.2.2]cryptand allowed access to the stable monomeric boracarbonyl anion 11.X -ray diffraction analysis confirmed encapsulation of the potassium ion by [2.2.2]crypt-and, thereby distancing it from the terminal oxygen atom (OÀ K: 5.919(6) ;F igure 1). TheB ÀObond (1.273 (8) )isvery short compared to acid-stabilized boracarbonyls (1.287(4)-1.329 (6) ), reflecting the effect of acid liberation. It is also shorter by 0.10 (ca. 8%)c ompared to borinic acid P11, suggesting enhanced O-to-B p donation on deprotonation. DFT analysis revealed the WBI of the B=Ob ond to be 1.40, that is,c onsiderably greater than acid-stabilized boracarbonyls (1.04-1.21). NPAcharges (B: + 0.99, O: À1.03) suggest the presence of as trong ionic component within the B = Om otif. As compared with the acid-protected analogues,t he B = O p bond of 11 (HOMOÀ2) is more energetically accessible, which, coupled with its sterically exposed nature,hints at high levels of reactivity across the unperturbed B = Odouble bond.
With an acid-free boracarbonyl compound in hand, the possibilities for carbonyl-like reactivity of 11 were then probed (Scheme 6, top). Treatment with CS 2 leads to facile p bond metathesis to afford borathiocarbonyl 11(S) with concomitant evolution of gaseous COS.X -ray diffraction analysis revealed 11(S) to contain at erminal B = Sd ouble bond (1.774 (1) ). This represents the first anionic thioxoborane isoelectronic with thiocarbonyls (cationic and neutral thioxoboranes having been reported previously). On the other hand, exhaustive hydrogenation of the B=Od ouble bond with Me 2 HN·BH 3 as am ild hydride source affords the corresponding hydroborane 11(H) (akin to carbonyl hydrogenation to an alkane). This contrasts with the lack of reactivity observed for Lewis acid-stabilized bora-acyl choride 5(BCF) towards the relatively strong hydride source K[HB( s Bu) 3 ]a nd underpins the non-innocent role of the Lewis acid in altering the electronic and steric environment around the B=Of ragment. Facile chlorination of 11 can also be achieved with POCl 3 as the chloride source to afford the corresponding chloroborane 11(Cl).M ost remarkably,t reating 11 with Tf 2 Oi nt he presence of pyridine results in complete abstraction of the oxide ion (O 2À ), to afford an electrophilic borenium cation 11(Py) stabilized by pyridine.
Lastly,b oracarbonyl 11 can also take on the role of an oxide ion transfer agent to an organic substrate,i nasimilar fashion to the nitrogen transfer exhibited by the isoelectronic nitrene (NHI) 2 P = N. Employing (p-Tol)N = C = N(p-Tol) as the substrate resulted in its insertion into the B=Odouble bond of 11 to form 11-NCN (Scheme 6, bottom). Subsequently, addition of (COCl) 2 induces release of the O-functionalized substrate from boron to furnish aurea derivative.F inally,the synthetic cycle could be closed by simple hydrolysis of the chloroborane 11(Cl) to restore the B À Ob ond, and borinic acid P11 can subsequently undergo ad eprotonation/sequestration sequence to regenerate 11.O verall, the boron centre acts as aplatform for oxide transfer, mimicking the activity of transition metals.
While B À Ob onds are traditionally regarded as thermodynamic sinks and are widely exploited to drive industrially important chemical transformations such as the Suzuki reaction, chemical recycling of the resulting BÀOb onds is challenging,a nd generally involves the use of harsh conditions and reagents.T his work demonstrates that the reactivity enhancement for classical doubly-bonded carbonyl compounds (cf.i nert C À Ob onds in ethers vs.m ore reactive C = Od ouble bonds in carbonyls) can be extended to boracarbonyl 11,i ne ffect facilitating facile cleavage of robust BÀOs ingle bonds by exploiting the more reactive terminal B=Od ouble bond, opening new avenues for reversing BÀOb ond formation under mild conditions. Classical carbonyl compounds are only weakly basic at the oxygen atom, whereas the isoelectronic boron analogue is anionic, which, coupled with the potent basicity of its oxygen atom, should make them versatile ligands.I ndeed, as was demonstrated in another report, the acid-free boracarbonyl 11 can additionally assume the role of an O-based ligand Scheme 6. Acid-free boracarbonyl 11 (Type I)a nd its reactivity as an oxide ion transfer agent.

Angewandte Chemie
Reviews 8632 www.angewandte.org (Scheme 7, top). [12b] Thus this new class of N-heterocyclic boryloxy (NHBO) ligand features as trongly nucleophilic Nheterocyclic boryl (NHB) unit [25] as the O-bound substituent and is isoelectronic with the well-known N-heterocyclic imine (NHI) ligand. [26] Hence it can be anticipated to possess similar strong 2s,4p donor abilities,inaddition to ademanding steric profile-two attributes that are absent in classical O-based ligands.F or instance,a lkoxides (RO À )l ack suitable steric protection in the vicinity of the metal centre.A lthough aryloxides (ArO À )w ith ortho-substitution can screen the metal centre,t heir good leaving group abilities render them weakly binding ligands.Inmain group chemistry,ligands that possess simultaneously strong donor and huge steric profile are key to the stabilization of highly reactive low-valent and low-coordinate species.Indeed, DFT analysis revealed stronger s and p donor properties for the NHBO ligand in comparison to other O-donors (e.g.[ {(Me 3 Si) 2 HC} 2 BO] À , [(2,6-Dipp 2 C 6 H 3 )O] À )-although not as strong as the donor properties of the N-based NHI family.

Organoaluminium Oxides
Among the group 13 elements,aluminium is characterized by its highly electropositive nature (Table 1). It also has an oticeably larger atomic size than boron (ca. 44 %l arger), thus it tends to adopt coordination numbers above three. Hence,t ricoordinate organoaluminium species containing well-defined AlÀOf ragments are extremely rare.
Monoalumoxanes (R À Al = O) are the monomeric units of alumoxanes (RAlO) n .T he simplest derivative,m ethylalumoxane (MeAlO) n or MAO, has significant industrial importance as ac atalyst activator in olefin polymerization. However,e xact details of its structural composition are not definitively known. In 1997, Power et al. attempted to generate am onoalumoxane by employing the much bulkier Mes* substituent. [27] However,t his kinetic stabilization approach proved insufficient to circumvent head-to-tail oligomerization, and tetrameric (Mes*AlO) 4 -featuring an eight-membered Al 4 O 4 ring-was isolated instead. It is noteworthy to compare with Westse arlier studies of the lighter homologue Mes*BO,w hich forms ad imer,h inting at the greater challenge associated with the quest for an isolable monoalumoxane.
Renewed interest in aluminium chemistry has been partly due to the discovery by Aldridge and Goicoechea in 2018 of an ew class of low-valent anionic organoaluminium species featuring an ucleophilic Al I centre that is isoelectronic with carbenes. [28a,b] This report was subsequently followed by related compounds from the groups of Coles,Hill, Yamashita, Kinjo and Harder. [28c-g] Theu mpolung character of this class of Al I compound has enabled access to unusual organoaluminium compounds previously inaccessible via traditional methods,i ncluding alumacarbonyls.H ere we survey the literature on these isolable aluminium analogues of carbonyl compounds and focus on the most recent examples and their reactivity studies.

Dimer-Stabilized Alumacarbonyls
In 2019 Theintrinsic preference for heteroallenes to react with the intermediate alumacarbonyl anion [15] 2 while leaving [P15] 2 untouched suggests that the Al=Om oiety possesses av ery high degree of reactivity.Accordingly,while the elusive [15] 2 cannot be detected even when carrying out the reactions at À80 8 8C, changing the solvent to THF allows for the isolation of [15-THF] 2 from the reaction with N 2 Oatlow temperatures (Scheme 9). X-ray diffraction analysis revealed that the nominally five-coordinate aluminium centre adopts ad istorted trigonal bipyridamidal geometry,w ith the oxide and the ligand oxygen assuming apical positions,and the nitrogen donor atoms and THF occupying equatorial positions.T hus, its stability can be attributed to THF occupation of the remaining sterically and electronically exposed vacant porbital in the equatorial direction, which compensates for the reduced N-to-Al p donation, due to the significant distortion imposed by the puckering of the bis(amino)dimethylxanthene ligand. Strikingly,t he Al À Ob onds (mean:1 .6763 (12) )a re shorter than those found in the [(Nacnac)Al(Me)OLi] 3 trimer (1.698 (1) )w hich contains tetracoordinate aluminum centres,presumably reflecting the considerably weaker AlÀO···K interactions (cf.A l ÀO···Li). Hence,   2 can be regarded as ad imeric THF-trapped alumacarbonyl anion. DFT analysis revealed that the model anionic fragment in (monomeric) 15-THF features an Al À Od istance which is only circa 1% shorter than in crystallographically determined (dimeric)   2 ,g iving further evidence that the potassium ions have am inor effect on the Al=Of ragment. The WBI value of 0.64 for the Al=Obond and NPAcharges of Al (+ 2.07) and O( À1.52) suggest that the short Al À Ob ond is due largely to electrostatic interactions,w ith am inor contribution from the Al = O p component.
In spite of featuring as trongly Lewis acidic aluminium centre adjacent to astrongly Lewis basic oxygen, the inability to quench by p bond formation potentially generates significant chemical "frustration". Accordingly,e xposure of   2 to H 2 affords the 1,2-addition product [15-H 2 ] 2 and substantiates the hypothesis that highly polarized E = Obonds within main group carbonyl analogues can exhibit FLP-like reactivity (Scheme 9).
In 2019, an ear-simultaneous report by Coles et al. described the isolation of aremarkably stable planar tricoordinate alumacarbonyl anion [16] 2 (Type I). [30b] This was achieved by exposing dicoordinate potassium aluminyl [P16] 2 to N 2 Oa tr oom temperature (Scheme 10). X-ray diffraction analysis confirmed ad imeric structure with the bridging potassium ions sandwiched between the arene p systems.I ti sn oteworthy that this dimeric form resembles the lighter boron homologue of the NHBO potassium dimer K 2 [12] 2 .T he most dominant feature is the three-coordinate aluminium centre,w ith an extremely short Al À Ob ond (mean:1 .6362 (14) )-noticeably shorter than Roeskys four-coordinate Lewis acid-stabilized neutral alumacarbonyl (1.659 (3) )a nd significantly shorter than five-coordinate THF-trapped [15-THF] 2 (mean:1.6763 (12) ), indicating the sensitivity of the Al À Ob ond to the coordination number at the aluminium centre.The significantly more stable nature of [16] 2 derives from the planarization of the aluminium centre, enabling efficient p donation from the nitrogen atoms and oxide ion to quench its Lewis acidity.D FT analysis on the model anionic fragment in (monomeric) 16 revealed that the Al À Ob ond is only 0.5 %s horter than in experimentally determined (dimeric) [16] 2 ,suggesting the minimal influence of the Al À O···K interactions,h ence giving credence to the notion of [16] 2 being considered ad imeric alumacarbonyl anion approaching its acid-free form. TheWBI of 0.86 for the Al=Of ragment of 16 is greater than that for 15-THF (0.64), While DFT analysis might suggest only minor influence of the potassium ions in [16] 2 ,c omplete sequestration by [2.2.2]cryptand "unmasks" the true nature of the alumacarbonyl moiety and results in intramolecular C(sp 3 )ÀHa ctivation of the proximal methyl group on the flanking Dipp substituent across the Al = Ob ond to form 16-crypt (Scheme 10). Freeing the flanking Dipp groups from h 6 coordination with the potassium ions presumably facilitates free rotation, positioning the Me group close to the newly exposed Al=Ou nit. Hence,i tw ould appear that the influence of the potassium ions in maintaining ad egree of structural rigidity in the dimeric form contributes crucially towards the overall stability of [16] 2 .
Analogous to    (2) )a nd CÀOs ingle bonds (mean:1 .3842-(15) ), distinguishing it from the common oxalate [C 2 O 4 ] 2À ligand and indicating aformulation more consistent with the further reduced ethene-tetraolate tetra-anion [C 2 O 4 ] 4À .T his work represents proof of concept that an alumacarbonyl anion can mimic certain aspects of transition metal chemistry to promote the elaboration of C 1 sources via the construction of new CÀCb onds to access more complex molecules.

Organosilicon Oxides
Silicon and oxygen are the two most abundant elements in the Earthsc rust-Si (28 %) and O(46 %), and their great affinity for each other is reflected in the exceptionally strong SiÀO s bond (501 kJ mol À1 ). [1] Hence,r obust materials made of silicon oxides (e.g.g lass,p olymers and semiconductors) play an integral role in our everyday lives.Despite this,onthe molecular level, organosilicon oxide chemistry is still in its infancy. Organic chemists primarily exploit the strong oxophilicity and substantial steric bulk of tetracoordinate organosilanes to act as protecting groups.F rom af undamental perspective,a no rganosilicon oxide featuring ah igher bond order is of considerable interest, and silacarbonyls featuring Si = Od ouble bonds are regarded as the lightest "heavy carbonyl" compounds.H owever, theoretical studies predict that the strength of the Si=O p bond (245 kJ mol À1 )i so nly half that of the corresponding s bond, in sharp contrast with carbonyl compounds. [1] As such, the isolation of discrete monomeric silanones analogous with classical ketones is extremely challenging-reflectedi nt he fact that they remained elusive for more than 100 years.

Acid-Base Stabilized Silacarbonyls
Neutral analogues:R 2 Si = O( Type I). In 2007, seminal work by Driess et al. introduced the donor-acceptor strategy to tame the highly reactive Si=Om oiety and generate av ariety of bottleable silacarbonyl species (Scheme 11). Thef irst example to be reported was a b-diketiminatesupported silaformaldehyde 19 capped by B(C 6 F 5 ) 3 ,inwhich ashort Si = Odouble bond (1.552 (2)  protection. [31d-g] It is remarkable that these silacarbonyl compounds could all be derived from the same silylene precursor P17.
In 2011, Roesky et al. reported as ila-acid anhydride 20 featuring ac entral O = Si À O À Si = Ol inkage with Si = Ou nits stabilized by donor-acceptor interactions (Scheme 12, top). [32a] In 2012, the same group employed asimilar strategy to coordinatively trap silaformyl chloride 21 (Scheme 12, bottom). [32b] Formyl chloride is an organic building block with great synthetic value,however its use is limited by its unstable nature,a si tr eadily decomposes to CO and HCl at room temperature.Hence,this bottleable silaformyl chloride derivative can be anticipated to be auseful reagent to introduce the HSi=Of unctional group.

Angewandte Chemie
Reviews 8636 www.angewandte.org 28 which can undergo a[ 4 + +2] cycloaddition/rearomatization sequence with benzaldehyde to form 28-PhCHO,o ra lternatively,i nt he presence of DMAP effect an intramolecular olefin metathesis to furnish cis-stilbene and form basestabilized 1-silaketene 27(DMAP),a lluding to transition metal-like behaviour at the silicon centre. [33e] In 2015, Robinson et al. reported the isolation of remarkable NHC-trapped molecular silicon oxides Si 2 O 3 (29)a nd Si 2 O 4 (30), [34a,b] and am ixed silicon/carbon oxide Si 2 CO 6 (31) [34c] via the controlled oxygenation of NHC-stabilized disilicon P29 with N 2 O, O 2 or CO 2 ,respectively (Scheme 14). Hence,N HC-stabilized disilicon P29 presents au nique molecular platform to mimic silicon surfaces and examine their oxidation to silicon oxides-a process that is highly relevant for the semiconductor and aviation industries.
In 2019, Inoue et al. reported that an NHC-stabilized silyliumylidene P33(Ter) undergoes hydrolysis in the presence of GaCl 3 to afford acid-base stabilized sila-aldehyde 33(Ter)(H) (Scheme 16). [36]  In 2020, Kato et al. described asila-acyl rhodium complex 40 featuring athree-coordinate silicon centre with short Si=O double bond of 1.540(3) (Scheme 18). [38f] Interestingly, while 40 undergoes reversible uptake of H 2 at the Rh I centre at room temperature,h ydrogenation of the sila-acyl moiety can be achieved at 60 8 8Ctoform 40-H 2 .DFT analysis revealed hydrogen transfer to occur via as eries of H-migrations from rhodium to the Si = Ofragment, reminiscent of the mechanism of the Fischer-Tropsch process. [39] It should be noted that direct hydrogenation of as ilacarbonyl with H 2 has not been reported to date,a nd Aldridge and Goicoecheasa lumacar-bonyl [15-THF] 2 remains the only example of am ain group carbonyl analogue to showcase such FLP-type reactivity. Hence,this work hints at future opportunities for cooperative bond activation by pairing transition metals with main group carbonyls.
With au niquely unperturbed sila-amide in hand, the authors explored its reactivity with various small molecules. Its latent Lewis acidity was confirmed via adduct formation Three months later,i n2 017, Inoue et al. described the remarkable isolation of crystalline acyclic silanones (Type I).
[9a] While the precursor (imino)(silyl)silylenes are highly reactive species and "mask" themselves by inserting into aromatic C = Cd ouble bonds to form silepins P42(R) (R = t Bu/SiMe 3 ), the silylene form could be "unmasked" in the presence of N 2 Ot oa fford the acid-base free acyclic (imino)(silyl)silacarbonyls 42(R) (Scheme 20). These systems are remarkably stable,with room temperature half-lives of 7h for 42(SiMe 3 ) and 24 hf or 42( t Bu) in solution, and are indefinitely stable in the solid state at À30 8 8C(for 42(SiMe 3 )) and at room temperature (for 42( t Bu)). 29 Si NMR signals at 33.7 ppm for 42(SiMe 3 ) and 28.8 ppm for 42( t Bu) are similar to KatosN ,C-silacarbonyl 41( i Pr) (38.4 ppm). X-ray diffraction analysis confirmed the monomeric nature of 42( t Bu) with at hree-coordinate silicon centre and terminal oxygen atom ( Figure 3). TheS i = Od ouble bond length of 1.537 (3) is similar to 41( i Pr) (1.533 (1) ). Thes hortened Si À Nb ond (1.646 (3) )a nd elongated exocyclic C = Nd ouble bond suggest the strong influence of the NHI substituent, offering ylidic stabilization of the Si=Of ragment. This was further confirmed via DFT analysis of 42( t Bu),i nw hich the HOMOÀ10 represents the Si = O p bond, with ap ronounced contribution from the exocyclic nitrogen of the NHI ligand. WBI gives av alue of 1.13 for the Si = Ob ond, similar with 41( i Pr) (1.14), which seems to suggest as imilar degree of perturbation by the p donor substituents.N otably,t he positive charge on Si (+ 1.70) is drastically reduced as compared with 41( i Pr) (+ 2.16), which can be attributed to the strong s donating abilities of the silyl ligand. Overall, the NHI-silyl ligand pair, featuring the complementary action of NHI as astrong p donor offering outer-sphere protection and the silyl ligand as as trong s donor offering steric bulk in the immediate vicinity of the Si=Om otif,s eems tailor-made to tame acyclic silacarbonyls.
Investigation of the reactivity of silacarbonyl 42(SiMe 3 ) with small molecules has also taken place:u ptake of CO 2 occurs via [2+ +2] cycloaddition to form 42(SiMe 3 )-CO 2 ;1 ,2addition of MeOH to form the silanol 42(SiMe 3 )-MeOH is also reported (Scheme 20). It is interesting to note that the silanol is formed exclusively,d espite the possibility for protonation of the ylidic nitrogen, unlike Katos 41( i Pr)i PrOH,i nw hich the ylidic carbon has been protonated. The authors also investigated the decomposition pathways of acyclicsilanone systems.Silacarbonyl 42(SiMe 3 ) decomposes in C 6 D 6 within 14 ht oaseries of unidentified products, possibly via 1,3-silyl migration of aS iMe 3 group from the Si(SiMe 3 ) 3 ligand to the terminal oxygen to give an intermediate disilene,which undergoes further decomposition via activation of the NHI ligand. However,i nt he presence of weakly basic MeCN,abase-stabilized silanone 42(SiMe 3 )-  MeCN could be isolated and structurally characterized. When stronger bases such as IMe 4 or THF were employed, the proposed disilenes were immediately formed as their base adducts,a nd in the case of the more basic IMe 4 ,d isilene 43 was sufficiently stable to permit structural authentication. On the other hand, silacarbonyl 42( t Bu) undergoes aremarkable Brook-type 1,2-silyl migration of the Si t Bu 3 ligand to the terminal oxygen to furnish an acyclicd icoordinate (imino)-(siloxy)silylene 44.These silyl migrations,driven by the highly oxophilic nature of silicon coupled with its strong desire to form new Si À Os ingle bonds (rather than Si = Od ouble bonds), remarkably overcome the counter-intuitive oxidation state changes from Si IV to Si II .Furthermore,this N,O-silylene 44 has recently been shown to be able to undergo oxidation to furnish atransient N,O-silacarbonyl which can be trapped by an NHC to furnish 45(IMe 4 ). [9b] However,itd imerizes in the absence of an external base,h ighlighting the integral role of the silyl ligand to stabilize the Si=Om oiety.T he overall transformation from silepin P42( t Bu) to N,O-silacarbonyl 45(IMe 4 ) involves multiple oxidation state changes of the central silicon atom, i.e.S i IV -Si II -Si IV -Si II -Si IV ,h inting at the versatility of such NHI-supported silicon species for possible future catalytic processes involving silicon centres.
TheW ittig reaction in which ac arbonyl compound is converted to an alkene by ap hosphorus ylide is ap owerful synthetic tool in the toolbox of organic chemists which surpasses all other olefination methods.More recently,Inoue et al. reported as ila-Wittig reaction in which as ilacarbonyl can undergo heavier olefination with phosphorus ylides to generate aseries of silenes,elegantly mimicking the classical Wittig reaction. [9c] Then ature of the ylide plays an integral role to determine the selectivity of products (i.e.( E)/(Z)alkenes based on thermodynamic/kinetic control). Hence,the authors investigated the reactivity of acid-base free N,Sisilacarbonyl 42( t Bu) with stabilized, semi-stabilized and unstabilized ylides (Scheme 21). With the stabilized ylide Ph 3 P=C(H)COO( i Pr), no sila-Wittig reaction resulted, as evidenced from the lack of O=PPh 3 formation. Instead, the silacarbonyl-ylide adduct 46 was formed, in which the ylide acts as an amphiphilic electron donor and acceptor simultaneously.T his adduct is thermally unstable and collapses to aS i À Hc ontaining species,a sv erified by 1 HNMR. The authors further posit its formation as involving initial dissociation of the ylide followed by rearrangement of silacarbonyl 42( t Bu) to N,O-silylene 44,w hich is able to activate the ylidic CÀHb ond of the phosphorus ylide.T his assumption is further supported by the fact that 44-CH can be acquired directly by treating 44 with the free ylide.
NOESY experiments revealed high Z-selectivity,i na nalogy with the classical Wittig reaction, which tends to be highly selective for (Z)-alkenes when unstabilized ylides are employed. However,i nc ontrast, the removal of O = PPh 3 is impossible as the highly reactive (imino)(silyl)silene products 48(R) decompose upon work-up.H ence,t he authors attempted to isolate these transient silenes as their NHC adducts.E mploying IMe 4 resulted in intramolecular activation of the wingtip C(sp 3 )ÀHbond of the NHC across the Si= Cdouble bond to form asilyl-substituted NHC 48(Me)-IMe 4 , which was identified by X-ray diffraction. Theo bserved instability of the silene-NHC adduct is in line with the highly reactive nature of (imino)(silyl)silenes.T his work represents proof of concept that unperturbed aheavier silacarbonyl can not only exhibit carbonyl-like reactivity,b ut also mimic the activity of transition metals by acting as aplatform for oxide ion transfer chemistry,a se xemplified by the sila-Wittig reaction.

Angewandte Chemie
Reviews 8640 www.angewandte.org Substituting the carbo-ylide functionality with an exceptionally strong p donating bora-ylide leads to adramatic increase in the half-life from 0.5 hf or N,C-silacarbonyl 41( i Pr) to 4days for N,B-silacarbonyl 49.The 29 Si NMR spectrum of the latter shows ab road quartet at 71.3 ppm, which is slightly downshifted compared with 41( i Pr) (38.4 ppm) and Inoues N,Si-silacarbonyl 42( t Bu) (28.8 ppm). X-ray diffraction analysis confirms the tricoordinate nature of both the central silicon and the adjacent boron atom (Figure 4). TheS i = O bond (1.5432 (12) )i ss lightly elongated as compared with 41( i Pr) (1.533 (1) )and 42( t Bu) (1.537 (3) ). TheB ÀSi bond (1.899 (2) )israther short, approaching those of borasilenes containing B=Si double bonds (1.838-1.859 )w hile the NÀ Si bond (1.763 (2) )i ss ignificantly elongated as compared with 41( i Pr) (1.731 (2) ). Hence,the unique stability of N,Bsilacarbonyl 49 can be ascribed to the enhanced perturbation of the Si=Of ragment by an ylidic ligand based on the more electropositive boron. DFT analysis further confirms this assessment, as the Gibbs free energy for dimerization of 49 (À67 kJ mol À1 )i sa pproximately half that of 41( i Pr) (À126 kJ mol À1 ). TheW BI of 1.09 for the Si = Om oiety is reduced compared to other tricoordinate silacarbonyls 41( i Pr) (1.14) and 42( t Bu) (1.13), further verifying the strong electronic perturbation of the Si=Of ragment. While the HOMO is centred on the ylidic boron ligand, the Si = O p and p * orbitals can be located in the HOMOÀ8and LUMO + 6, respectively.
In spite of extensive electronic and kinetic stabilization, dimerization of N,B-silacarbonyl 49 still occurs at room temperature (albeit slowly;S cheme 22). However, X-ray diffraction analysis of [49] 2 revealed no simple cyclodisiloxane product. Thea uthors postulated that initial head-to-tail dimerization forms ahighly strained cyclodisiloxane,and the relief of steric congestion within the four-membered ring drives the onward ligand rearrangement. Interestingly,t his dimer features a( borylene)(methylene)phosphorane function with cumulative B=P=Cd ouble bonds and is the first example of such which incorporates agroup 13 element. The stability of 49 could be further increased by capping the terminal oxygen with aL ewis acid such as MgBr 2 ,a nd 49-MgBr 2 is stable in solution for more than three weeks.T he solid-state structure revealed only slight lengthening of the Si=Ob ond to 1.553(2) (0.6 %), while both the BÀSi (1.865 (3) )a nd NÀSi (1.744 (2) )b onds experienced substantial shortening.F urthermore, 49-MgBr 2 represents the first example of abase-free silacarbonyl coordinated only by aL ewis acid. This suggests that 49 has au nique nucleophilic character,asopposed to the more typical electrophilic nature associated with such systems. 49 can also undergo a [ 2 + +2] cycloaddition with small molecules such as CO (employing Fe 2 (CO) 9 as the CO source) to form af our-membered cyclic dioxocarbene Fe(CO) 4 complex 49-CO.
While the incorporation of exceptionally strong internal p donor functionalities (i.e.amino,ylide,NHI, bora-ylide) has replaced the need for external bases to tame base-free silacarbonyls,i nevitable electronic perturbation of the of the central Si = Ofunctionality distinguishes these silacarbonyls from silanones,w hich might be regarded as the true homologues of ketone." Genuine" silanones have remained elusive for more than 100 years,until recently when Iwamoto et al. described the breakthrough synthesis/isolation of acrystalline cyclic di(alkyl)silanone 50 (Type I)w hich is stable at room temperature. [13] This significant feat was achieved through an arduous six-step synthesis to the precursor silylene P50,f ollowed by final oxygenation with N 2 O( Scheme 23). The 29 Si NMR spectrum for 50 displays ar esonance at 90.0 ppm that is downfield-shifted compared with other silacarbonyls featuring strong p donor substituents (41( i Pr), 42( t Bu), 49:2 8.8-71.3 ppm). X-ray diffraction analysis unequivocally confirms the planar three-coordinate nature of the silanone 50 with aS i = Ob ond length of 1.518 (2) , which is noticeably shorter than other acid-base free silacarbonyls (41( i Pr), 42( t Bu), 49:1.533(1)-1.5432 (12) )and very close to H 2 Si=O( 1.515 )d etermined by rotational spectroscopy,hence hinting at the unperturbed nature of the Si=O moiety ( Figure 5). This hypothesis was affirmed by DFT analysis,w hich revealed the WBI of the Si = Of unction to be 1.35, that is,s ignificantly greater than other silacarbonyl systems (41( i   substituents,the charge on the terminal oxygen is determined by the p donor properties,thus making it agood measure for the p donating capacity of its substituents.For silanone 50 (Si: + 2.08, O: À1.10), the lack of p donors substantially reduces the charge on oxygen as compared with strong p donor substituted silacarbonyls (41( i Pr), 42( t Bu), 49: À1.23 to À1.27), hence supporting the notion that its stability is purely based on kinetic protection. While this huge steric profile underpins its remarkable stability in solution at room temperature,i ti si nsufficient to completely inhibit head-totail dimerization, which is still possible at 60 8 8Ct oa fford av ery congested cyclodisiloxane [50] 2 .
With a" genuine" Si = Om otif in hand, the authors began to explore its reactivity.Incontrast to KatosN,B-silacarbonyl 49,which prolongs its half-life by forming stable adducts with Lewis acids,the adduct of silanone 50 and B(C 6 F 5 ) 3 collapses via 1,3-C 6 F 5 migration from boron to silicon, in as imilar fashion to Roeskysa lumacarbonyl 13 and in line with the electrophilic character of 50 (Scheme 23). While an ene reaction with the polarized C = Od ouble bond of acetone is possible, [2 + +2] cycloaddition with the weakly polarized C = C double bond of styrene to form 50-H 2 CC(H)Ph confirms the strong ambiphilicity intrinsic in the Si=Om otif.T his work affirms the integral role of kinetic stabilization to tame acidbase free silanones,g iving rise to as pectrum of isolable silacarbonyl derivatives with avariety of substitution patterns. Hence,t his opens avenues for comparison with their lighter homologues,advancing the vision of Kipping and bringing his dream to reality.

Acid-Base Stabilized Germacarbonyls
In 2012, the landmark discovery of the first heavier carbonyl analogue by Tamao,M atsuo et al. was based on an acyclica cid-base free germanone (Type I). [6a,b] This remarkable achievement was enabled by the deployment of the extremely bulky and rigid Eind ligand scaffold (Scheme 25). Tr eating bis(Eind)germylene P53 with Me 3 NO as the oxygen atom transfer reagent yielded bis(Eind)germanone 53,which has remarkable thermal stability up to 200 8 8C. X-ray diffraction analysis unambiguously revealed the planar tricoordinate nature of the central germanium atom affixed with aterminal oxygen atom ( Figure 6). While the sheer size of the Eind ligand is apparent from the solid-state structure,another key aspect is the rigidity imbued by its fused ring structure, restricting motion of the Eind ligands on either side of the Ge = Om oiety,e ffectively obviating any potential side reactions (such as C À Ha ctivation) that have plagued previous attempted syntheses of such species.T he Ge=Ob ond length of 1.6468(5) is at the shorter end of the range defined by base-stabilized germaureas (51(L):1 .646(2)-1.672 (3) ). DFT analysis revealed aWBI of 1.25 for the Ge=Ofragment, with NPAc harges of + 1.80 (Ge) and À1.05 (O), indicating the important contribution of the ylidic germylene oxide (Eind) 2 Ge + À O À form. TheG e = O p and p * orbitals are represented in the HOMOÀ5a nd LUMO,r espectively,a nd the HOMO is composed of the non-bonding oxygen lone pair. An in-depth computational study by Pandey in 2015 highlighted the importance of non-covalent London dispersion interactions provided by the Eind substituents towards the overall stability of 53. [6c] Thesame team subsequently probed the reactivity of this germanone,with aview to elucidating similarities and differences from lighter ketone homologues.W hile analogous reduction with LiAlH 4 could be carried out, reduction by milder PhSiH 3 proceeds without the need for ac atalyst to afford 53-H-H and 53-PhSiH 3 ,r espectively (Scheme 25). Familiar nucleophilic addition reactions with MeLi or H 2 O yielded the expected methylated germanol 53-Me-H and hydroxylated digermanol 53-H 2 O,respectively.Most interestingly,s ignificant differences are observed in reactivity towards C=Oc ontaining compounds.W ith acetone,a ne ne reaction occurs,w hich is uncommon with two carbonyl compounds.This presumably reflects the much higher basicity of the terminal oxygen atom in 53.L astly,t he latent polarization within the Ge=Od ouble bond facilitates binding of CO 2 in a[ 2 + +2] fashion to form 53-CO 2 ,areaction that is energetically unfavourable for standard ketones.P erhaps most importantly,t his work sparked excitement by offering the possibility of bottleable acid-base free main group carbonyl analogues,a nd it might be said that it paved the way for more recent syntheses of carbonyl analogues from across the Periodic Table. 4

. Group 15 Carbonyl Analogues
Oxoammonium ions are the isoelectronic nitrogen analogues of carbonyl compounds.I np articular,t hose derived from [TEMPO] + are widely employed as catalytic oxidants in the dehydrogenation of alcohols to carbonyl compounds,due Scheme 25. Tamao and Matsuo's acid-base free germanone 53 (Type I)a nd its reactivity. to the unique stability of the TEMPO radical. [42a] However,it was not until 2007 that Nishide et al. reported the structural authentication of [TEMPO] + ,r epresenting the first isolation of an itrogen analogue isoelectronic with carbonyl compounds (Type I). [42b] This system features at rigonal planar nitrogen centre and ashort N=Odouble bond (1.184 (10) ). On the other hand, the only isolable heavier group 15 carbonyl analogues reported so far are the phosphacarbonyls.

Organophosphorus Oxides
Phosphorus and carbon share ad iagonal relationship in the Periodic Table,a nd organophosphorus compounds have been described as "carbon copies". In stark contrast with other main group elements,o rganophosphorus oxides containing PO single bonds and double bonds are equally prevalent. In fact, the high oxophilicity of phosphorus and the latent P=Odouble bond strength (536 kJ mol À1 )drive the Michaelis-Arbuzov reaction which transforms the PÀOsingle bonds of phosphites to the P=Od ouble bonds of phosphine oxides. [43] Thec lassical Wittig reaction also exploits the formation of aP = Od ouble bond to drive conversion of the C = Od ouble bonds of ketones to C = Cd ouble bonds of alkenes (which are less stable by ca. 126 kJ mol À1 ). [43] The inherent stability of these tetrahedral s 4 l 5 -phosphine oxides can be partly ascribed to steric protection of the electrophilic phosphorus centre.H ence, s 3 l 5 -phosphacarbonyl cations featuring at rigonal planar phosphorus environment, and which are isoelectronic with carbonyl compounds,are elusive synthetic targets.H ere,w es urvey the development of phosphacarbonyls from their initial isolation as base-stabilized entities to the recent discovery of base-free phosphacarbonyl cations.

Base-Free Phosphacarbonyls
In 2018, Dielmann et al. reported the breakthrough isolation of Lewis base-free oxophosphonium monocations, which represent the first phosphacarbonyl species Angewandte Chemie Reviews 8644 www.angewandte.org (Type I). [11a] To prevent unwanted further reaction with an oxygen atom transfer agent, they devised an alternative synthetic approach via chloride abstraction from phosphoryl chlorides P60(X) (X = N/CH) containing ap reformed P = O double bond (Scheme 28). X-ray diffraction analysis unambiguously revealed the planar tricoordinate nature of the phosphorus centres in N,N-phosphacarbonyl 60(N) and N,Cphosphacarbonyl 60(CH) (Figure 7). TheP =Ob ond lengths of 1.4603(9) and 1.463 (2) ,r espectively,a re in the range of base-stabilized tetracoordinate phosphacarbonyl monoand dications (54-59:1 .451(2)-1.4843 (7) ). Thes hort P À N and PÀCbonds allude to the possibility for multiple bonding between the ylidic NHI and NHO ligands,e ffectively acting as intramolecular electron donors to quench the strongly electrophilic nature of the central P = Omoiety and removing the need for external Lewis bases.D FT analysis on the P = O fragment reveals WBIs of 1.32 for 60(N) and 1.30 for 60(CH) with NBO charges at P/O of + 2.31/À1.03 and + 2.14/À1.03 for 60(N) and 60(CH),r espectively.T he HOMOs of 60(N) and 60(CH) are centred on the ylidic nitrogen and carbon, respectively,w hile the LUMOs correspond to the P=O p * orbitals.Furthermore,the computed fluoride ion affinities (FIA) of 60(N) (634 kJ mol À1 )and 60(CH) (618 kJ mol À1 )are intermediate between B(C 6 F 5 ) 3 (425 kJ mol À1 )a nd Stephans [PF(C 6 F 5 ) 3 ] + (795 kJ mol À1 ), hinting at their potential Lewis acidic properties.T he more electrophilic nature of the bis(NHI)-stabilized system 60(N) than (NHI)(NHO)-substituted 60(CH) is reproduced with the Gutmann-Beckett method which gives an acceptor number of 102 for 60(N), which is in the range of Lewis superacids,w hile that of 60(CH) is significantly lower at 33. Preliminary reactivity studies have also been carried out on these phosphacarbonyl species.I nterestingly, 60(N) reversibly binds pyridine,m imicking the addition-elimination mechanism of classical carbonyl compounds.Inaddition, 60(N) activates i PrOH via 1,2addition across the P À Nb ond (rather than the P = Od ouble bond) hinting at the greater latent basicity of the ylidic ligands than the terminal oxygen atom, in am anner reminiscent of KatosN ,C-silacarbonyl 41( i Pr).O verall, this work demonstrates that base-free phosphacarbonyls are potent Lewis acids that show promise towards catalytic applications,a s exemplified by the reversible binding of pyridine.
Thec arbonyl-yne reaction between ketones and alkynes proceeds via a[ 2 + +2] cycloaddition promoted by Lewis acid catalysis or photo-irradiation to form intermediate oxetenes.
However,the ring strain imposed by incorporation of aC =C double bond within af our-membered ring results in subsequent collapse by electrocyclic ring opening to furnish enones. Dielmann et al. most recently reported on ah eavier carbonyl-yne reaction between ab ase-free N,N-phosphacarbonyl 60(N) and av ariety of alkynes to afford isolable oxaphosphete cations (Scheme 28). [11b] Notably,c yclo-reversion is possible,e videnced by the complete liberation of phenylacetylene from oxaphosphete 60(N)-HCCPh at 300 8 8Ct o regenerate phosphacarbonyl 60(N),which could alternatively be trapped by DMAP or by exchanging phenylacetylene with 4-ethynyltoluene.S uch reversible P = Od ouble bond formation is also reminiscent of the classical addition-elimination mechanism of carbonyl compounds.Onthe other hand, X-ray diffraction analysis revealed that the PÀOb ond of 60(N)-HCCPh (1.677(2) )i sr ather long, hinting at possible electrocyclic ring opening to access ap hospha-enone cation with aP = C À C = Ol inkage.D FT analysis revealed the barrier to ring opening is 95 kJ mol À1 ,supporting the notion that P À O bond cleavage could be achieved at elevated temperatures. Indeed, the phospha-enone cation could be trapped by ah etero Diels-Alder reaction with dienophiles such as Figure 7. Solid-state structure of 60(N).F or clarity,hydrogen atoms are omitted, Dipp groups are simplified as wireframe.T hermal ellipsoids set at 50 %p robability.

Conclusions and Outlook
Main group analogues of the ubiquitous carbonyl functional group incorporating p-block elements have long been considered to be highly elusive entities.A major turning point in the quest for isolable main group carbonyls was the successful employment of acid/base protocols,w hich granted access to these transient species in their masked forms.H owever, the electronic and steric perturbation imposed by such chemical tricks impairs their chemical reactivity and contrasts with the rich chemistry displayed by classical carbonyl compounds.
An ew era has been marked by the recent isolation of crystalline acid-base free main group carbonyl analogues ranging from al ighter boracarbonyl to the heavier silacarbonyls,p hosphacarbonyls and ag ermacarbonyl, completely free from acid/base interference ( Table 2). These synthetic achievements have been enabled by the employment of electron-rich substituents (e.g.y lides,a mino groups) with huge steric profiles,hence relinquishing the need for external acids and bases,and enabling (close-to) unbiased comparison with classical carbonyl compounds.M ost importantly,t heir "unmasked" nature elicits exciting new chemistry.F rom carbonyl-type reactions to transition metal-like oxide ion transfer chemistry,t hese systems offer to bridge the gap between carbon and transition metals,o pening up to the possibility for unique "crossover" reactivity.Furthermore,the variation in overall charge from anionic group 13 to cationic group 15 main group carbonyls imbues them with additional properties as exemplified by the strong ligating abilities of boracarbonyl and potent Lewis acidity of phosphacarbonyl, while charge-neutral silacarbonyls and germacarbonyls maintain more ambiphilic character.H ence,t hese main group carbonyl systems have journeyed al ong way from their humble beginnings as lab curiosities,t ob ottleable trophy compounds,t ot heir present status as potentially versatile reagents in chemical synthesis.