Diazaphospholene and Diazaarsolene Derived Homogeneous Catalysis

Abstract The past 20 years has seen significant advances in main group chemistry and their use in catalysis. This Minireview showcases the recent emergence of phosphorus and arsenic containing heterocycles as catalysts. With that, we discuss how the Group 15 compounds diazaphospholenes, diazaarsolenes, and their cationic counterparts have proven to be highly effective catalysts for a wide range of reduction transformations. This Minireview highlights how the initial discovery by Gudat of the hydridic nature of the P−H bond in these systems led to these compounds being used as catalysts and discusses the wide range of examples currently present in the literature.


Introduction
During the course of this century,t here has been ag rowing surge in using main-group compounds to replicate the roles of transition metals. [1,2] This is driven in part by the ever growing need to find more economically viable and environmentally sustainable alternatives to these metals,b ut also by scientific curiosity.I nt he past six years, diazaphospholenes (DAPs) have emergeda sa ni nteresting class of heterocycle that has proven to be effective at catalyzing ap lethora of reduction-based transformationsu nder mild conditions. The diazaphospholene heterocycle may be simply defined as an N-heterocyclic phosphine contained within af ive membered unsaturated ring. Diazaphospholenes startedt og arner attention in the late 1990s where it was discovered that they could act as precursors to forming diazaphosphenium cations (NHPs), [3,4] which were themselves receiving significant focus. [5] Initially independently reported by both Fleming and Hutchins in 1972, [6,7] NHPsa re cationic,d ivalent phosphorus(III) species which possess al one pair of electrons and av acant p-orbital. These properties mean NHPs have ambiphilic character and can act as both aL ewis acid and Lewis base. However, althoughc omparisons can be made between NHPs and the familiar Arduengo N-heterocyclicc arbenes (NHCs), NHPs have inverse electronic properties ( Figure 1). That is, NHPs are weaker s-donors but much stronger p-acceptors;aconsequence of the formal positive chargea nd + 3o xidation state at phosphorus. [3,[8][9][10][11][12] In the early 2000s the structure andr eactivityo fd iazaphospholenes were extensively studied by Gudat, who has since reviewed this. [13] Gudat's studies revealed that DAPs possess 6pdelocalization in the five-membered ring unit, but to achieve this the s*(P-X)-antibonding orbital is required. This in turn reduces the bond order of the PÀXbond and transfers additional negative charge on the X-atom. Thus, ac ompromise is reached where greater energetic stabilization in the DAP ring is achieved but at the cost of al oss of the degree of covalency in the PÀXb ond. [14] It was observed that when X = H, hydridicb ehaviorw as observed, contrasted to the classically observed protic character of the hydrogen atom in the PÀHb ond. This was exploited by stoichiometrically reducing benzaldehyde. [15] This observation of hydridic behavior would be key for the use of DAPs in catalytic reduction reactions. Furthermore, Gudat and colleagues reported that diazaphospholenes may be used as organocatalysts for phosphorus-carbon bond formation from the condensation of silyl phosphine with alkyl chlorides. [16] Another key discovery on the road to DAP assisted catalysis was from Radosevich and colleagues, who in 2012 first reported the reversible two-electron redox cycling of P III /P V ,which enabled it to be used for transfer hydrogenation of ammonia borane to reduce azo benzene. This was achieved by using a three-coordinate phosphorus species with an NO 2 type pincer ligand that forced as trained, planar T-shaped geometry (Scheme 1). [17] In contrast to the vast studies and attention diazaphospholenes and phosphenium cations have received, their arsenic Scheme1.Proposed catalytic cycle for the reduction of azo benzene via P III / P V redox cycling. counterparts have remained largely unexplored. The arsenic analogueo ft he diazaphospholene is termed diazaarsolene. Early examples of diazaarsolidines (five membered ring heterocycle containing arsenic but as aturated backbone) were reported by Wolf and Cowley, [18,19] but al iterature search into diazaarsolenesg ave few results. Minkin and colleagues computationally lookeda tt he energy barrier of pyramidal inversion in diazarsolenes, [20] but synthetic work is limited. Examples include work from Nieger et al.,w ho synthesized 2-halogeno-1,3,2-diazarsolenes, [21] as well as reports from Gudat and Ragogna. [22,23] The first isolateda nd structurally characterized arsenium cations were reportedb yBurford in 1992, [24] and although much rarer than phospheniums, an umber of examples do exist. [19,23,25,26] Althoughalone pair and av acant p-orbital are still present, their bondingt ot ransition metals is typically confined to Lewis acid chemistry,w here there is little to no s-donation from the lone pair. [27,28] Thisi sd ue to the heavier pnictogen elements having ag reater reluctance to form at rigonal planar geometry and so the lone pair adopts more s-orbital character. [10,29] Herein, this review looks to explore the examples currently present in the literature of diazaphospholene, diazaarsolene, and their cationic counterparts in performing reduction-based organic transformations, and to highlight the versatility these systemsh ave. Furthermore, the catalytic cycles are discussed and mechanistic differencesb etween the catalysts debated.

Transfer hydrogenation
The journeyt ou sing DAPs as catalysts was first paved by the discoveryo ft he hydridicn ature of the PÀHb ond [15] and the catalytic reduction of azobenzene using P III $P V redox cycling. [17] These two observations led to the Kinjo group in 2014 to use 2-H-1,3,2-diazaphospholene (1)f or the first time as a catalystf or the reduction of azobenzenes using ammoniaborane as the hydrogen source. After optimization,5mol %o f the diazaphospholene 1 with four equivalents of ammoniaborane were used for the reduction of ar ange of (E)-azo-compounds, giving the corresponding hydrazine product.U nlike in the P III $P V redox cycling case (Scheme 1), mechanisticallyt his catalysis proceeded firstly by the addition of the PÀHb ond in 1 to the N=Nb ond to give ap hosphinohydrazine. This then undergoes hydrogenolysis of the exocyclic PÀNb ond by hydrogen transfer from ammonia-borane to give the desired product and regenerate 1 (Scheme 2). Investigating the mechanism further using deuterium kinetic isotopee ffect (DKIE) found that cleavage of the BÀHa nd NÀHb onds takes place via ac oncerted double cleavage pathway in the rate-determining step. [30] Reduction of carbonyl groups Since the report on azobenezene reduction, [30] as eries of additional reductions have been reported. Although aldehydes and ketones have previously been reduced stoichiometrically by diazaphospholenes, [14,31] in 2015t his was performed catalytically in the first metal-free catalytic hydroboration of carbonyl derivatives with pinacolborane (HBpin). [32] Here catalytic amount of the same diazaphospholene as used for azobenezene reduction (1)w as ablet or educe aldehydes (using 0.5 mol % 1)a nd ketones (using 10 mol % 1)w ith HBpin( 1.0 equiv and 1.3 equiv,r espectively). Aw ide substrate scope was performed and 1 was found to be tolerant to both aliphatic anda romatic aldehydes, as well as av ariety of ketones. This catalytic reaction proceeds by the formation of an alkoxyphosphine intermediate from the addition of 1 to the carbonyl substrate, where subsequentc leavage of the PÀOb ond and the BÀH bond in HBpin gives the hydroboratedp roduct and regenerates catalyst 1.K inetic studies along with DFT calculations found that the bond dissociation is involved in the rate-determining step in the transition state and that the process is stepwise, albeit almost concerted (Scheme 3).
Given our groups previousi nterest in arsenic chemistry, [33,34] we lookedtodetermine whether arsenic couldmimic this reactivity by performing hydroboration of aldehydes with HBpin. Although our systems included the fusing of ab enzene ring on the backbone, recentw ork by Yang and Chen on the nucleophilicity of different diazaphospholenes showed that these should still be hydridic. [35] Ar ange of diazaarsolenes and dithiaarsolenes were synthesized, including the chloro-and benzyloxy-derivativesa sw ell as their cations. Optimization studies found that 5mol %o fd iazaarsolene 2 proved to be the most effective pre-catalyst (Scheme 4). Proceeding with the substrate scope, 2 was shown to be an efficient pre-catalyst for this catalysis, reducing both electron withdrawing and electron donating substrates;a lbeit 10 mol %c atalyst loading was required for the latter.M echanistic investigations found that the catalysis proceeds in an analogous fashion to Kinjo. [32] The diazaarsolene pre-catalystr eacts with HBpin to form the proposed active arsenic-hydride catalyst (Scheme 4) via s-bond metathesis, where the mechanism then follows the proposed catalytic cycle shown in Scheme 3. [36] This reactivity was then compared to that of the phosphorus derivative. As eries of diazaphospholene, dithiaphospholene, and dioxaphospholene pre-catalysts were produced,a sw ell as using their cationic counterparts (Scheme 5). In this case optimizationr eactions found that the diazaphosphenium triflate cation 3 was the best performing pre-catalyst. Using 10 mol % 3 with one equivalent of HBpin,aseries of both electron withdrawing and electron donating aldehydesw ere smoothly reduced.M echanistically we proposed that this catalysis did not perform in as imilar fashion to the carbonyl reduction described above, [32] and insteadinvolved the formation of ab oronium species. However,a ttempts to attain mechanistici nsight were thwarted by the detectiono ft he decomposition product PH 3 at d = À238.5 ppm. [37] From here an umber of comparisons could be made between the arsenic and phosphorus systems (Scheme 6).
The neutral arsenic compounds showed greater catalytic activity than their phosphorus analogues;f or example, the dithiaarsolene pre-catalyst gave 64 %c onversion of 4-(trifluoromethyl)-benzaldehyde to the hydroborated product after 12 hours,w hereas the phosphorus analogue only achieved9% Scheme3.GeneralScheme and proposed catalytic cycle for carbonyl reduction. 0.5 mol % 1 for aldehyde reduction and 1.0 equiv HBpin;10mol % 1 for ketonereduction and 1.3 equiv HBpin.
Scheme4.To p: Arsenic pre-catalysts used in the optimization study. Bottom:F ormation of the proposed active arsenic hydride catalyst.
conversion after 24 hours in CDCl 3 .O nt he other hand, al ess clear picturee merged from the comparison of the cationic complexes but comparing the diazaarseniumt riflate to the diazaphosphenium triflate showed higher reactivity for the latter. For the former,5 0% product conversion of the hydroborated product wasd etected after 24 hours in CH 2 Cl 2 ,w hereas for the latter > 95 %p roduct conversion in CDCl 3 was observed (Scheme 6).

Reduction of imines
Reduction reactions of imines is another area which hasb een exploredu sing diazaphospholene based homogeneous catalysts. In 2017, Speed and colleagues lookeda tt he reduction of imines with HBpin to produce amines (Scheme 7). Ad iazaphospholene similart o1 wasu sed in which the hydrogen atom is replaced by an eopentyloxy group (4). The purpose behind this was the PÀHb ond is sensitive to oxygen/moisture, thus the inclusion of the neopentyloxy group offersm ore stability to the system, making handling the diazaphospholene much more convenient for use in organic transformations. Thus 4 is ap re-catalyst which generates the active catalyst 1 via addition of HBpin.S creening results for the optimum conditions found that 2mol % 4 with one equivalent of HBpin at room temperature were best. Proceeding with the substrate scope, ar ange of imines were explored, with sterically hindered indanone-derived imine and aldimines with different steric demand tolerated. AL ewis basic pyridyl ring was found to give no detrimental effect and, using a p-methoxybenzyl (PMB) protecting group gave the expected reduced product. Aqueous (acid/base) work-up theng ave the amine product. Mechanistically,t he formation of the activec atalyst 1 from 4 occurs, which is then able to deliver ah ydride and reduce the imine substrate. [38] This reduction of imines was speedily followed up by the reporto ft he first example of enantioselective reduction using ac hiral diazaphospholene.N eeding as ource of chirality,a chiral diimine was used. This was reacted withP Br 3 and cyclohexene to produce ac hiral diazaphospholene, bearingaP ÀBr bond, whichw as then reactedw ith neopentyl alcohol to produce the pre-catalyst (Scheme 8). For the catalysis, the same optimized conditions were used as above,a lbeit with THF not CH 3 CN as the solvent. Asymmetric reductiono fi mines with HBpin was then undertaken using 2mol %o ft he chiral diazaphospholene pre-catalyst 5.Ab road substrate scope of imines flankedb ya romatic groups revealed high enantiomeric ratios of up to 88:12. These resultsa tt he time were the best reported for alkyl imine hydroboration with HBpin. [39] The mechanism for this reduction is proposed to proceed as above.
Further work on asymmetrici mine reduction later led to a chiral diazaphosphenium triflate species( 6)t hat could perform the catalysis. Althought he use of diazaphosphenium cations as catalysts forr eduction chemistry had previously been reported, [40] this was the first example of using them for asymmetricc atalysis. To make the diazaphosphenium chiral, the same ligand scaffold that was used in diazaphospholene 5 was again employed. With that, optimization reactions found that 1mol %o fd iazaphosphenium 6 with 1.2 equivalents of HBpin were sufficient for the reduction. Expanding the scope, cyclic imines were found to undergo reduction,g iving aryl pyrrolidines as products, with enantiomeric ratios of up to 97:3. Moreover, iminesi ncorporating functional groups such as pyridyl rings and thiophenes,w hich are traditionally challenging for transition metal catalysts, were efficientlyr educed.
Owingt ot he cationicn ature of 6,t he mechanism is found to be dissimilar to that with diazaphospholene 4 andi sp roposed to be similar to other phosphenium based reduction (e.g. see reduction of pyridines later). [40] The first step is the phosphenium cation abstracts ah ydride from the activated imine-HBpin complex, where it is then redelivered to the subsequent boranyl-substituted iminium cation species. This gives the desired reduced imine and regenerates the catalyst. The proposed catalytic cycle is shown in Scheme 9. [41] 1,2,4,3-triazaphospholenes have likewise been employed as catalysts. The triazaphospholene ring is similar to ad iazaphospholene, except it contains three nitrogen atoms instead of two. Synthesis of the triazaphospholene pre-catalyst is similar to diazaphospholenes but uses amidrazones as the starting ligand.S creening studies of ar ange of triazaphospholenes with varying steric properties found that 7 and 8 (Scheme 10) were the most suitable to proceed with as ubstrate scope. Using 10 mol %pre-catalyst with 1.1 equivalents of HBpin, avariety of imines were found to undergo hydroboration readily, but more interestingly imines derived from aniline werea lso readily reduced. This is of interest as theses ubstrates do not undergo reduction using diazaphospholene catalysts. Mechanistically this catalysis is intriguing since, unlike the catalytic examples discussed so far,n oe videnceo fP ÀHb ond formation was observed. Instead it is proposed that the pre-catalysti s ionized in CH 3 CN, giving the cation, leading to an interaction between the positively charged phosphorus and Na tom from the imine substrate. Hydridet ransfer via as ix-membered transition state (I)t hen occurs, after which the active catalyst is regenerated by releasing the borylateda mine via II. [42] The proposed catalytic cycle, as found from DFT studies, is given in Scheme10.
In af urthera ttempt to develop stable main group catalysts, Speed employeda ir and water stable phosphine(V) oxide precatalysts in the reduction of imines. In these systems the precatalystw ill be reduced into the catalytically actived iazaphospholenes upon addition with HBpin.T his work initially resulted from the observation that the diazaphosphole pre-catalyst 4 undergoes hydrolysis to the phosphine oxide 9 over time. It was also observed that when HBpinw as present, reduction of 9 to generate active catalyst 1 occurred (Scheme 11). Inspired by this, 9 was prepared from the addition of the bromide precursor to 1 andt riethylamine, followed by addition of water. With 9 in hand, its suitability as ap re-catalyst was tested by performing reduction catalysis that diazaphospholenes were known to catalyze. Note that this catalysis is av ariation of the Scheme9.Proposed catalytic cyclicfor imine reduction using ac hiral diazaphosphenium cation.
Scheme10. Proposed catalytic cycle for the reduction of iminesu sing at riazaphospholene pre-catalyst.
Scheme11. Top: Synthesisofd iazaphospholene from secondaryphosphine oxide. Bottom:Secondary phosphine oxides used in catalysis. above-mentioned imine reduction with pre-catalysts 4 and 5 (Scheme 7). With that, 1mol %o f9 was used to catalyzet he reduction of imines with 1.1 equivalents of HBpin (Scheme12). Developing this further,e nantioselective imine reduction was performed using ac hiral secondary phosphine oxide (10). For the asymmetric catalysis,5mol % 10 was used, which could reduce selected imines to the corresponding amine with comparable enantioselectivity to using the chiral diazaphospholenes previously discussed (Scheme 12). [43] Conjugate reduction Having previously shown that diazaphospholene 1 can promote transfer hydrogenation and reduce carbonyl bonds, [30,32] it was then shown that 1 can also enablet he reduction of a,bunsaturated esters. To begin with, two initial stoichiometric reactions were performed:( i) reduction of methyl methacrylate using 1 to afford the 1,4-addition product and, (ii)subsequent addition of ammonia borane to give the C=Cr educed ester product (Scheme 13).
Twoc atalytic variants based on the above stoichiometric reactions were explored,t he first using ammonia borane as the reductant, affording saturated esters, and the second using HBpin to afford b-ketoesters after af ollow-up reaction with a nitrile. In the first case, 1mol %o f1 was used alongw ith stoichiometric ammonia borane (Scheme 14, top). 1,4-hydroboration of a,b-unsaturatede sters required 10 mol % 1 at 90 8Ca nd the resulting boryl enolate intermediate was then reacted with nitriles to form the b-ketoester product following hydrolysis (Scheme 14, bottom).
Both reactions proceed via the formation of phosphinyl enol ether from 1,4-hydrophosphination of the a,b-unsaturated ester (first step Scheme 13). Addition of ammonia borane then cleaves the PÀOb ond, generating an enol intermediate which tautomerizes to saturated esters. On the other hand, addition of HBpin again affords PÀOb ond cleavage, but through sbond metathesis, generating ab oryl enolate intermediate. This then undergoes coupling with nitriles. [44] The Cramer group have previously had interestinthe closely relatedd iazaphospholidine heterocycle (diazaphospholene but with as aturatedb ackbone), which they have used as ligands for metal-based catalysis. [45,46] Therefore, given the groups interest in phosphorush eterocycles and chiral ligand design, in 2018 Cramera nd colleagues reported the enantioselective conjugate reduction of a,b-unsaturated carbonyl derivatives using diazaphospholene catalysis. To begin with, an umber of chiral pre-catalysts were synthesized, but screening results found that pre-catalyst 11 (Figure 2), which contains 3,5-xylyl substituents and am ethoxy group in the backbone, gave the best performance for the conjugater eduction of acyl pyrrole (reactiont ype shown in Scheme 15). Performing as ubstrate scope on ar ange of a,b-unsaturated acyl pyrroles using the conditions 5mol % 11 and1 .5 equivalents of HBpin in toluene solventg ave reduced products in yields and enatiomeric ratios of up to 97 %a nd 95.5:4.5 respectively.I na ddition, chalcones were found to reduce smoothly to the corresponding ketone and the more challenging a,b-unsaturated amides were tolerated, with an enantiomeric ratio of up to 86:14.
Upon explaining the origin of enantioselectivity in the catalysis, knowing that the PÀHb ond in the active catalyst is in a Scheme12. Reduction of iminesu sing pre-catalysts 9 and 10.
Scheme13. Stoichiometricaddition of methylm ethacrylate to diazaphospholene 1 followed by stoichiometric addition of ammonia borane.  perpendicular positiont ot he ring (a consequence of the pyramidal local geometry), two accessible quadrants are available away from the bulky aromatic backbone. This led to Cramer proposingt he depicted stereoselectivitys hown in Figure 2. Twop otentialc atalytic cycles were proposed;P ath Aa nd Path B. In Path A, the diazaphospholene hydride is the active catalyst, where the hydride is delivered upon addition of the conjugated substrate, after which regenerationo ft he active catalyst occurs via addition of HBpin.T his in turn gives ab oron enolate, whicha fter hydrolytic work-up generatest he final product. Alternatively,i nP ath Bt he first part of the catalytic cycle is the same, but the coordinated intermediate II undergoes sbond metathesis with pinBOMe (produced from the earlier sbond metathesis step), regenerating 11 and giving the boron enolate( Scheme 15). [47] Note Path Ai sthe same as that reported with catalyst 1. [44] Phosphine oxide pre-catalyst 9 could also enablec onjugate reduction,w here chalcone was smoothly reduced using 1mol %o f9 and 1.1equivalents of HBpin (Scheme 16).
Using catalytic diazaphospholene 12 for this transformation, the optimization studies exposed allyl 2-phenylacrylate to an array of terminal reductants,w here HBpin provedm ost effective for the transformation in combination with 1mol % 12.A substrate scope followed, where awide array of allylic acrylates bearing various functional groups were found to be tolerated for the rearrangement, which was also enantiospecific for substrates with existing stereogenic centers. Investigations into the diasteroselectivity found it could be tuned by varying the solventa sw ell as changing the diazaphospholene catalyst, suggesting several pathways exist depending on the nature of the pre-catalyst and substrate. Thus, two possible mechanisms are proposed for the reaction. In the first proposed pathway (Scheme 18), the addition of the active catalyst 1 gives intermediate I,w hich reacts with HBpin to form boron enolate III via s-bond metathesis. In turn, intermediate III rearranges to V. Scheme15. Proposed catalytic cycle for conjugate reduction with HBpin. Where Y = pyrrole fragment. Diazaphospholene shown is asimplified representation of 11.
On the other hand, as econd mechanistic pathway may take place (Scheme 19), where addition of the active catalyst gives intermediate II.F rom here two options are possible and both involvea [3,3]-sigmatropicr earrangement and eliminationo f catalyst 1 and differ only in their ordering. Intermediate II forming intermediate IV is most desirable as this would allow greater controlo ft he diastereoselectivity and enantioselectivity by the bound diazaphospholene.

Reduction of pyridines
Dihydropyridines are commonly found in biological molecules such as NADH (nicotinamide adenine dinucleotide) and are also usefuli ns ynthetic chemistry (e.g. Hantzsch esters). Their synthesis from the corresponding pyridines is nevertheless challenging owing to the stability of the aromatic ring and usually preactivated systemsa re required. In 2018, diazaphospheniumc ations were found to serve as an effective pre-catalyst for the reductiono fp yridines with HBpin.A fter as eries of screening reactions with different cations of varying steric properties, the diazaphosphenium 13 proved to be the most effective pre-catalyst to proceed with. Along with 1.05 equivalents of HBpin,5mol % 13 was used for the substrate scope, where av ariety of substituted pyridines were found to be smoothly reduced with both regio-and chemo-selectivity. Good functional group tolerancew as observed when the pyridine ring was substituted in the meta-position, howevers ubstitution in the ortho-a nd para-position proved more challenging. Given the cationic nature of 13 this catalysis does not proceed in an analogous fashion to that with the neutrald iazaphospholene 1.I nstead, investigations found that the first step involves hydride transfer from HBpin to 13,g enerating diazaphosphenium-hydride and the boronium salt [(py) 2 Bpin]OTf. The second step is then reduction of the activated pyridine via hydride delivery from the diazaphosphenium-hydride (Scheme 20). [40] It was found that neutral diazaphospholenes can also be used fort his reduction, with 2.5 mol %p re-catalyst 4 effective for reducingp yridinesw ith HBpin (1 equiv). Substrates bearing electron-withdrawing groups in the meta-position workedw ell, but again ortho-a nd para-substitutedp yridinesw ere more challenging. Mechanistically this pyridine reduction is different to the example reported above. The first step is postulated to be formation of the active catalyst 1 via s-bond metathesis, after which pyridine reduction takes place from hydride delivery.F rom here BÀPh ydride transfer is speculated to occur, giving the desired hydroborated pyridine product and regenerating catalyst 1 (Scheme 21).
Comparing reductions of neutrald iazaphospholenes with cationic diazaphospheniumss hows that the latter is able to tolerate more electronr ich pyridines, whereas the former requires more electron withdrawing groups attached to the pyridine ring for smooth reduction to take place. On the other hand, the diazaphospholenep re-catalyst operates well in low polarity solvents (such as [D 6 ]benzene), whereas the diazaphospheniumcation does not. [49] Finally,t he phosphine oxide pre-catalyst 9 described earlier was also used for pyridine reduction, which when using 1mol %p re-catalyst with 1.1 equivalents of HBpin,n iconitrile was found to be effectively reduced. Interestingly,3 -acetylpyridine was selectively reduced, with the ketone moiety remaining untouched (Scheme 22). [43] Reduction of CO 2 The use of CO 2 as aC 1s ource is potentially very powerful as it offers an on-toxic way to build more synthetically useful products in ac heap manner,b ut also gives au se for this harmful greenhouse gas. [50] As ar esulto ft his, the catalytic reduction of CO 2 has been investigated using the diazaphospholene catalyst 1.T he diazaphospholene was found to undergo ah ydrophosphination reactionw ith CO 2 (1 atm), producing ad iazaphospholenes pecies with af ormate group attached (Scheme 23, top). This transformation is ac onsequence of the oxygen group from CO 2 insertingi nto the PÀHb ond of 1 along with hydride transfer to the carbon atom of CO 2 .I tw as postulated that the formate group should readily transfer to an acceptor.T hus, the formate intermediate was reacted with half an equivalent of Ph 2 SiH 2 .P h 2 Si(OCHO) 2 resulted as the major product and the siloxane (Ph 2 SiO) 3 as am inor product (2.3:1 respectively) (Scheme 23, bottom). Moreover, it was later discovered that the formate transfer step can be accelerated by adding5mol % 1.
Subsequently,t he one-pot N-formylation of amines with CO 2 ,u sing 5mol % 1 as ac atalyst (Scheme 24) was performed. For the catalysis, aw ide substrate scope of both primary and secondary amines was used. For the secondary amines,l esshindered aliphatic aminesg ave the N-formylamine in excellent yields of > 90 %, but an increase in sterics afforded N-methylated amines. Secondary amines containing aryl substituents were found to be tolerated. Expanding the scope, all aliphatic and aromatic primary amines tested were found to work well, with yields in the range of 72 %t o9 9%. [51]

Conclusions and Outlook
In this Minireview,t he use of the heterocyclic diazaphospholenes, diazaarsolenes and their cationicc ounterparts as catalysts for organic reduction transformationsh as been evaluated. In 2014, the catalytic reduction of azobenzene using 2-H-1,3,2diazaphospholene was first reported, making use of the hydridic PÀHb ond these complexesp ossess.S ince then the reduction of carbonyls, imines, a,b-unsaturated esters, pyridines and CO 2 have all been reported.I nt hesec ases an umber of diazaphospholene speciesh ave been utilized, with the use of an alkoxide derived co-ligand providing an advancement in the field due to increased moisture/oxygen tolerance compared to 2-H-1,3,2-diazaphospholene. Further advances have come from the inclusiono fachiral ligand scaffold allowing for enatioselective catalysis. Halide abstraction from diazaphospholenes results in cationic phosphenium formation,a nd these cations have provedt ob ehighly effective for these reductions, and in certain cases outperforming the neutrald iazaphospholene. In addition to this, heavierG roup 15 arsenic pre-catalysts have been developed, including diazaarsolenes and diazaarsenium cations.H owever,i ng eneral, the reactivity and tolerance were diminished compared to the phosphorus counterparts.
Althoughs everal similarm echanisms operate in these reactions, ak ey feature is the formation of aP ÀHb ond in the catalytic cycle. Importantly,t he hydridic nature of the PÀHb ond opens the possibility fort hese phosphorus containingh eterocycles to be used for av ast array of reduction reactions. We are only at the beginningo ft he field and it is likely that many more catalysts and differing reactivity will be uncovered in the near future.