f‐Block Phospholyl and Arsolyl Chemistry

Abstract The f‐block chemistry of phospholyl and arsolyl ligands, heavier p‐block analogues of substituted cyclopentadienyls (CpR, C5R5) where one or more CR groups are replaced by P or As atoms, is less developed than for lighter isoelectronic C5R5 rings. Heterocyclopentadienyl complexes can exhibit properties that complement and contrast with CpR chemistry. Given that there has been renewed interest in phospholyl and arsolyl f‐block chemistry in the last two decades, coinciding with a renaissance in f‐block solution chemistry, a review of this field is timely. Here, the syntheses of all structurally characterised examples of lanthanide and actinide phospholyl and arsolyl complexes to date are covered, including benzannulated derivatives, and together with group 3 complexes for completeness. The physicochemical properties of these complexes are reviewed, with the intention of motivating further research in this field.


Introduction
The f-block elements, the lanthanides( Ln) and actinides (An), exhibit remarkable physicochemicalp roperties that have spurred numerous curiosity-driven investigations and technological applications. [1] Organometallic f-blockc hemistry is predominated by cyclopentadienyll igands (Cp, C 5 H 5 )a nd their derivatives (Cp R ,C 5 R 5 ), where one or more of the ring Ha toms are substituted by aw ide variety of alkyl, aryl or heteroatomic Rg roups; much of their success owes to:( i) straightforward ligand synthesis andi nstallation at metals by well-developed and robust synthetic routes;(ii)occupation of the equivalent of three coordination sites at large f-blocki ons in their most common h 5binding mode;a nd, (iii)facile tuning of ligand steric and electronic properties by Rg roup variation to provide additional kinetic and thermodynamic stabilisation andf ine-control of metal coordination spheresa nd redox chemistry. [2] Cp and Cp R ligandsh ave supported seminale xamples of f-blockc hemistry in both as pectatorl igand role and in controlling the physicochemicalp roperties, including rare examples of f-block-metal-(loid) bonds [3] and terminal unsupported multiple bonds between f-blocka nd p-blocke lements, [4] rich single-electron transfer (SET) chemistry, [5] the discoveryo fh itherto unknown + 2o xidations tates in solution for aw ider ange of Ln and An, [6] and Ln single-molecule magnets (SMMs)w ith high blocking temperatures. [7] Given the huge influenceo fs ubstituents in f-blockC p R chemistry,t he comparatived earth of examples of isoelectronic heterocyclopentadienyl f-block complexes, where one or more of the ring Ca toms is substituted by other p-block atoms, is noteworthy. [8] Of these related ligand families, phospholyls (C 5Àn R 5Àn P n )a nd arsolyls (C 5Àn R 5Àn As n )h ave proved popular, with the lighter congenersm ore widely investigated, and their group 3a nd f-blockm etal chemistry was reviewed several times between 1998 and 2006. [3a, 8, 9] The relatively restricted development of f-block phospholyl anda rsolyl chemistry compared with that of Cp and Cp R analoguesi sm irrored in the s-, p-and d-blocks. [9a, b] The reasons for this disparity are the same as for other heterocyclopentadienyls:t he well-documented benefits of Cp and Cp R ligandss ummarised above, together with their widespread renown, maket hem naturalp rimary choicesf or chemists in exploratory synthesis fields. [2a, b] However,f or more nuanced and specific applications, the introduction of ring heteroatoms can provide electronic fine-tuning to maximise physicochemical properties, providing rich and diverse chemistry. [8] The relative popularity of phospholyls and arsolyls in f-block chemistry compared with other heterocyclopentadienyls can be attributed to both pragmatic (i-ii)a nd ligand design (iii-vi) considerations: [3,8,9] (i)synthetic routes to monophospholyls and -arsolyls are maturea nd are relativelys traightforward; (ii) 31 Pn uclei are I = 1/2 with 100 %n aturala bundance ( 75 As I = 3/2, 100 %a bundant), providing au sefulN MR/EPR spectroscopich andle;( iii)phospholyl and arsolyl ligands are relatively soft compared with Cp R analogues anda re thus well-suited for stabilising low oxidation state f-block ions;( iv) phospholyls and arsolylsa re able to bind in a h 1 -fashion through their P/As lone pairs but are more likely to exhibit an h 5 -binding mode than lighter congenersw ith harder heteroatom donora toms, for example, pyrrolyl (C 4 R 4 N) and pyrazolyl/imidazolyl (C 3 R 3 N 2 ), thus they more effectively mimicC p R ligands in occupying a large proportion of metal coordination spheres;( v) the Pa nd As lone pairs provide ar ange of alternative binding modes over Cp R ;f or example, for monophospholyls m:h 5 ,h 1 -a nd m:h 1binding modes increase the likelihood of formation of multinuclear complexes;a nd, (vi)phospholyls and arsolyls are poorer p-donors and stronger p-acceptors than analogousC p R ligands,i nfluencing metal reduction potentials and redox chemistry.
Since the first rare earth phospholyl complexesw ere reported by Nief and Mathey in 1989, [10] < 100 monophospholyl, monoarsolyl and polyphospholyl complexes of the group 3 metals,L na nd An (including benzannulated derivatives) have been structurally authenticated to date;t his contrasts starkly with the corresponding Cp/Cp R chemistry,w here the first reportedexamples were by Birminghama nd Wilkinson in 1954 [11] and there are now thousands of structurally characterisedcomplexes. [12] As noted above,various books and reviewshave cov-ered group 3a nd f-block metal phospholyl and arsolyl chemistry prior to 2006 as part of wider subjecta reas. [3a, 8, 9] In the last fifteen years, there have been significant discoveries that we believe now warrant areview solely dedicated to this topic.
Here, we will firstly presenta no verview of the ligand design criteria and binding modes of phospholyls and arsolyls, followed by general synthetic routes to these ligandsa nd metal complexes;w ef ocus on selected examples for brevity as this material is covered in detail elsewhere. [3a, 8, 9] We then review all structurally authenticated group 3a nd f-blockm onophospholyl and -arsolyl complexes, divided intos eparate sections for Ln and An, and subdivided by formal metal oxidation state;i nt he case of the most developed Ln III chemistry this is furthers plit by Ln startingm aterials and ancillary ligands. The small number of examples of polyphospholyl Ln and An complexes are covered together at the end in ad edicated section, subdivided by the number of Pa toms in the rings. As we focus on structurally characterisede xamples, we compile thesec omplexes and salient data at the end (Table 1) and we only provide ligand binding modesw here these have been authenticated in the solid state. When appropriate, we cover interesting physicochemical properties that the complexes have been shown to exhibit, within individual sections. We conclude with remarks on the current and predicted future state of group 3 and f-blockp hospholyl and arsolyl chemistry.W ew illi nclude the group 3e lements Sc, Yand La under the heading of Ln in this review,a st hey can be considered as diamagneticM III mimics of Ln III ions, although we appreciate the term "rare earth" is the preferred nomenclaturef or the group 3a nd Ln metals combined. [1a]

Ligand Design Criteria and Binding Modes
The phospholyl and arsolyl ligandst hat have been employed in f-block chemistry to date are compiled in Figures 1a nd 2; acronyms are provided for monophosphoyls and -arsolyls, whereas polyphospholyls are labelled A-D.M onophospholyl and -arsolyl ligandsa re variously substituted at the 2,5-, 3,4-, DavidP .Mills is aR eader at the Department of Chemistry in the University of Manchester, whereh eh as spent his independent career to date focusing on non-aqueous synthetic chemistry, mainly in f-block chemistry.H is researchi nterests are centred around the synthesisa nd studyo fc omplexesw ith atypical oxidation states, geometries, and bonding regimes.
Peter Evansr eceived his MChema nd PhD from the Newcastle University under the supervision of Dr Keith Izod, where he researched the stabilisation of heavier carbene analogues with bulky phosphides. He is interested the unusual reactivity of low oxidation state complexes and their synthesis.  [56] or all four C-positions of the C 4 E( E = P, As) rings, apart from the parent phospholylC 4 H 4 P, Hhp. Substituents include R = Me, tBu, SiMe 3 and Ph, and benzannulated derivatives; the currently available selectiona nd ring positions are intrinsically linked to the common synthetic routes to these ligands, as can be deduced fromo nly tBu substituents being seen in the polyphospholyls A-C (see Section 3). As stated previously,t he introduction of Po rA si nto carbocyclic rings influencesb oth the strength of metal-ligand bindingand the resultant redox properties of complexes. [3a, 8, 9] As with Cp R chemistry, [2] the substituents affect complex solubility and both the thermodynamic and kinetic stability of f-block complexes;t he electron density of the rings is also influenced by donating (Me, tBu) and withdrawing (SiMe 3 ,P h, fused carbocyclic rings) Rg roups. Ligands with the largest Rg roups tend to give the most kinetically stable complexes,w hich are less likely to oligomerise owing to ac ombination of steric bulk about the metal coordination sphere and buttressing of the heteroatom lone pairs. The most commonb inding modes of mono-a nd polyphospholyls are compiled in Figure 3, with analogoush apticities seen for monoarsolyls. The introduction of heteroatoms with lone pairs increases the flexibility of ligand coordinationo ver the most common h 5 -, h 3 -a nd h 1 -binding of Cp R ligands, where bridging modes are rare for the f-block. [2] As stated previously,t he h 5 -binding mode is the most common binding mode for phospholyls anda rsolylsw ith f-blocke lements as P and As atoms are relatively soft. Substituents, available space at metal coordination spheres and ancillary ligandsa re all contributory factorsa st ow hether or not the heteroatom lone pairs form dative bonds with f-blocki ons. [3a, 8, 9] Although this review focuses on binding modes observedi ns olid-state structures, it must be appreciated that dynamic fluxional behaviour in solution is common, and the presence of 31 Po r 75 As nuclei can provide an additional usefulh andle to study this behaviour by NMR or EPR spectroscopy. [Sc{h 5 -C 2 tBu 2 P 3 )} 2 (m:h 6 ,h 6 -C 3 tBu 3 P 3 )] [Sm(Cp*) 2 (h 2 -C 2 tBu 2 P 2 E)(THF)] (E = P, Sb) [Li(THF) 4 ][Yb(h 5 -C 2 tBu 2 P 2 E) 2 (h 2 -C 2 tBu 2 P 2 E)] (E = P, Sb) [Sc(h 5 -C 2 tBu 2 P 3 ) 2 (h 2 -C 2 tBu 2 P 3 )] [Y(h 5 -C 2 tBu 2 P 3 ) 2 (h 2 -C 2 tBu 2 P 3 )] [Tm(h 5 -C 2 tBu 2 P 3 ) 2 (h 2 -C 2 tBu 2 P 3 )] [U( h5 -C 2 tBu 2 P 3 ) 2 (h 2 -C 2 tBu 2 P 3 )] [(h 5 -C 2 tBu 2 P 3 ) 2 Sc(m:h 2 ,h 5 -C 2 tBu 2 P 3 )Sc(h 5 -C 2 tBu 2 P

Synthetic Routes to Complexes
An umber of synthetic strategies have been developedf or the synthesis of f-block phospholyl, arsolyl and polyphospholyl complexes,w ith the route depending upon the nature of both the metal and ligand employed, as well as the metal oxidation state. [3a, 8, 9] The four most commons trategies to monophospholyls and -arsolyls and practical considerations will be briefly outlined in this section, in decreasing order of their frequency of application; examples will be providedt hroughout Sections 5-6. Various synthetic routes to f-block polyphospholyl complexes will be discussed separately with dedicated schemes in Section7.

Salt metathesis
Salt metathesis reactions between Ln and An halides or pseudo-halides with alkali metal ligand transfer agents are by far the mostcommon route for synthesising f-block phospholyl and arsolyl complexes. [3a, 8, 9] This is typically due to the commercial availability and facile synthetic routest oa nhydrous and donor-solvent coordinated trihalidesf or all the Ln (with the exception of PmX 3 ), which can be used to prepare pseudohalide complexes,f or example, Ln(BH 4 ) 3 (THF) 3 and Ln(AlMe 4 ) 3 ; some Ln diiodides are also readily available (Sm, Eu, Tm and Yb here;a lthough DyI 2 and NdI 2 are known, [5b-f] they have not been utilised successfully in Ln II phospholyl chemistry to date). [1a] Conversely,t he two An with lowest radiological hazard and highest naturala bundancies, thorium and uranium, have well-developeds ynthetic routes to readily solvated halide (e.g.,A nCl 4 ,U I 3 )a nd pseudo-halide (e.g.,U (BH 4 ) n , n = 3, 4) precursors from nitrate (Th), and oxide and metallic( U) starting materials. [1b] Mostf -blockh alide and borohydridep recursors can be converted to donor solvent adducts, typicallyT HF or DME, to endow improved solubility,w hich facilitates their salt metathesis reactions, but the presence of these solvents can also lead to unwanted side-products, for example, diprotonation andring-opening reactions of THF.
The choice of ligand transfer reagent and reaction solvent are crucial for determining the compositiono fp roducts because of the highly electropositive nature of the f-block elements. [1,16] In the majority of cases where lithium phospholyls are used as transfer reagents salt-occluded complexes tend to form, where Li is trapped in the coordination spheret hrough contactsw ith several Ln-/An-bound halides. Ac ombination of Ln or An di-/tri-iodides and sodium or potassium transfer reagents often yields discrete f-block complexes by assisting the precipitation of alkali metal iodide by-products;p otassium iodide is only sparingly soluble in THF and is therefore ad esirable by-product. Althought he occlusiono fs uch salts is suppressed with these reagents, ah andful of potassium 'ate' fblock phospholyl complexes have been isolateda nd are included in this review.T he high affinity of f-block ions for binding ethereal solvents can make it challenging to synthesise solvent-free phospholyl and arsolyl complexes as diethyl ether or THF are typically used as the reaction solvents for solubility reasons. Whilst some metal-bound solventm olecules can be removedf rom f-blockc omplexes upon exposure to vacuum, the elevated temperatures often required to facilitate the dissociationo fs trongly bound N-and O-donors olvents can be greater than the temperature of complex decomposition. As a consequence, some solvent-free f-block phospholyl and arsolyl complexes are synthesised by reactingb inaryL no rA nh alides with sodium or potassium ligand transfer agentsi nt oluenea t reflux for extendedp eriods to circumvent the low solubility of the reactants in aromatic solvents. [3a, 8, 9] These reactions are moderate-to high-yielding and have facilitated numerous studies of the resultant rare earth phospholyl complexes.

Redox transmetallation
Whilst salt metathesis is often the most convenient synthetic strategyf or preparing f-block phospholyl and arsolyl complexes, several alternative approaches have been developed, which in some cases can be more suitable. Redox transmetallation reactions using ligand transfer reagents of readily reducible metal ions such as Tl I andP b II have provedu seful for concomitantl igand installation and f-blockm etal iono xidation, in cases where the Ln or An ions have suitable redox potentials, for example, Sm II and U III . [1]

Bond insertion
Biphospholes and biarsolesc ontaining EÀEb onds, and phospholesa nd arsoles containing EÀXb onds (e.g.,X= halide, Ph), may react directly with metallic Ln and An by af ormal bond insertionw ith metal oxidation and ligand reduction.T his has provedm ost useful to date for Ln II phospholyl and arsolyl chemistry for Sm and Yb. [3a, 8, 9]

Redox
As phospholyl and arsolyl ligands have proven utility for stabilising metals in low oxidations tates it is unsurprising that when these ligands have been installed on f-block metals in intermediate oxidation states, the resultant complexes can often be straightforwardly oxidised or reduced, for example, Ln II to Ln III or U III to U IV ,and vice versa.
Also in 1994, Nief and Ricard reported the syntheses of the heteroleptic dinuclear Yb II complexes [{Yb(h 5 -Tmp)(m-Cl)(THF) 2 } 2 ]( 4)a nd [{Yb(h 5 -Tmp)(m-SPh)(THF) 2 } 2 ]( 5;F igure 5) by the oxidative insertion reactions of Yb powder with the respective reagent XPC 4 Me 4 (X = Cl or SPh) in THF with at race amount of HgCl 2 .C omplex 4 could also be prepared by the ligand scrambling reactiono f[ Yb(Tmp) 2 (THF) 2 ]w ith YbCl 2 in THF;t he salt elimination reactiono f4 with two equivalents of NaSPh in THF also gave 5. [21] Complexes 4 and 5 exhibit similar geometries in the solid state, with half-sandwich motifs at Yb with h 5 -Tmp ligands (YbÀP: 2.911(1) for 4;2 .931(4) and 2.955(5) for 5), Yb 2 X 2 cores, and each Yb coordination sphere completed by two bound THF molecules. The 31 PNMR spectra of 4 (81.4 ppm) and 5 (82.5 ppm) each exhibited one signal, with no J YbP couplingc onstants reported. Desmurs [22] Green-brown crystalso f6 were analysed by single-crystal XRD to revealadistorted octahedral Ru centre with trans-hydrides and the two PPh 3 ligands mutually cis-, with the h 1 ,h 1 -P,P'-chelating{ Yb(THF) 2 (Tmp) 2 }m etalloligand completing the Ru coordination sphere. The coordination of the Tmp Pl one pairs to Ru enforces an ear-eclipsed configuration of the two C 4 Pr ings, which are boundi na nh 5 ,h 5 -fashion to Yb in ab ent metallocenem otif (YbÀP: 2.930(2) ), with two mutually cis-THF molecules completing the Yb coordination sphere. The 31 PNMR spectrum of 6 exhibited doubletsf or both the PPh 3 and Tmp Pa toms, with the latter signal of interest at 103 ppm, confirming that the 220 Hz splitting is due to a trans-2 J PP coupling, with no J YbP coupling constants disclosed.
The homoleptic mononuclear Dy III complex [Dy(h 5 -Dtp) 2 ] [Al{OC(CF 3 ) 3 } 4 ]( 41)w as synthesised in 2019 by Chilton,M ills and co-workers by the sequential salt metathesis and protonolysis reaction of 37-Dy with allyl magnesium chloride and [NEt 3 H][Al{OC(CF 3 ) 3 } 4 ], with the respective elimination of magnesiumd ihalides,t riethylaminea nd propylene providing thermodynamic driving forces (Scheme 6). [39] The installation of a sufficiently weakly coordinating anion provided an isolated bent [Dy(h 5 -Dtp) 2 ] + cation in the solid state, with DyÀPd istances of 2.7880(8)a nd 2.7981 (8) .T he axial ligand field and rigidity of the aromatic ligandsof41 are both conducive to enhance the SMM properties for Dy III ,a nd the effective barriert o magnetic reversal (1760 K) and maximum hysteresis temperature (48 K) for 41 are both competitive with leadingC p R Ln SMMs. [7]
In 1992, Ephritikhine and co-workers reported the salt metathesis reaction of UCl 4 with three equivalents of K(tmp) in toluene to yield [U(h 5 -Tmp) 3 (Cl)] (58,F igure 18);t he corresponding reactionw ith stoichiometric K(tmp) gave [U(Tmp) 2 (Cl) 2 ]. [53] Several derivatives of 58 were synthesised by salt metathesis protocols with variousr eagents:  [54] All complexesw erec haracterised by 1 HNMR spectroscopy and elemental analysis (except [U(Tmp) 2 (CH 2 SiMe 3 ) 2 ]), but as olid-state structure was only disclosed for 59 ( Figure 18). The U IV centre in 59 exhibits a pseudo-octahedral geometry with a mer-configuration of Cl ligands;t he two Od onor atoms of DME and an h 5 -bound Tmp ligand complete the coordination sphere,w ith aU ÀPd istance of 2.926(4) .T he synthesis of such al arge number of Tmp U IV complexes and analogousC p* complexes allowed the authors to compare the steric and electronic effects of these ligands on complex spectroscopic data and redox chemistry. [54] In were assigned by 1 HNMR spectroscopy and elemental analysis, although as olid-state structure was determined for 60 to reveal aU IV centre coordinated by k 3 -BH 4 ,T HF, h 8 -COT and h 5 -Tmp, with at ypicalU ÀPd istance of 2.970 (8) .
The  [56] Singlecrystal X-ray diffractions tudies of 61 and 62 revealed that the pseudo-tetrahedral U IV centres were bound by two terminal cis-chlorides and two Tmp ligands in an h 5 -binding mode in each case (61:r ange UÀP: 2.823(7)-2.862(7) ; 62:U ÀP: 2.851(9) and 2.86(1) ). The assignment of U IV centres in 61 and 62 is made through analysis of NiÀPd istances in the former complex being in line with Ni 0 tetrakis-phosphines, and as hort NiÀNi distance in the latter complex( 2.546(9) )b eing consistentw ith the presence of am etal-metal bond and formal Ni I centres;l ong mean U···Ni distances in these complexes( e.g.,3 .38(2) for 61)a re not in line with 5f/3d metalmetal bonds. The 1 Ha nd 31 PNMR spectra of 61 provided additional characterisation data (d P :1 99.2 ppm), whilst those of 62 were broad and could not be interpreted;c rystalso f62 could not be separated easily from the NaCl by-product, hence no additional characterisation data were obtained. 7. Lanthanide and Actinide Polyphospholyl Complexes 7.1. C 3 P 2 and C 2 P 3 complexes Ah andful of examples of Ln and An polyphospholyl complexes have been reported, whilst there have been no reports to date of corresponding polyarsolyls. The first structurally characterised rare earth polyphosholyl complex, [Sc{h 5 -C 2 tBu 2 P 3 )} 2 (m:h 6 ,h 6 -C 3 tBu 3 P 3 )] (63), was reported in 1996 by Cloke,N ixon andc o-workerst of orm in 5-10 %i solated yield from the cyclooligomerisation reaction of tBuCP with Sc vapour in a1 0:1r atio (Scheme 12). [57] This noteworthy triple decker complexr epresented the first structurally authenticated example of formal Sc I centres, together with an ovel instance of ligated1 ,3,5-triphosphabenzene in the solids tate;f -block complexes with Ln or An centres in formal + 1o xidation states are unknownt od ate. The total valence electron count of 63 is only 22 e À ,w hich is also remarkably low for at riple-decker sandwich complex. Ther eactionm ixture that yielded 63 was furtheri nvestigated by Cloke,N ixon and co-workers, and the sandwich complex [Sc(C 3 tBu 3 P 2 ) 2 ]w as isolated in 5-10 %c rystalline yield after sublimation (Scheme 11). [58] Unfortunately, this complex could not be structurally authenticated, but all characterisation data were in line with aS c II formulation.T o the besto fo ur knowledge, no Ln and An diphospholyl complexes have been structurally authenticated to date, but it is noteworthy that [Yb(C 3 tBu 3 P 2 ) 2 ]w as made by analogous procedures and has been spectroscopically characterised. [59] The diuranium complex [{U[HC(SiMe 2 NC 6 H 4 Me-4) 3 ]} 2 {m:h 4 ,h 4 -C 2 tBu 2 P 2 }],r eported by Liddle andc o-workersi n2 014 to form from the reductivec oupling of two molecules of tBuCP by a U III precursor,i sa lso worthy of mention at this point as the sole example of aL n/An complex containing a cyclo-C 2 P 2 ring that has been structurally characterised to date; [60] Liddle has recently reviewed f-block complexes containing dianionic fourmembered aromatic rings. [61] The solid-state structure of 63 revealed that the planar bridging 1,3,5-triphosphabenzene has elongated ring PÀC bonds compared with unbound 1,3,5-C 3 tBu 3 P 3 ,t ogether with remarkably short Sc···C 3 P 3centroid distances of 1.787(5) ;this contrasts with the relatively long Sc···C 3 P 2centroid distances to the capping anionic C 3 tBu 3 P 2 -1,3 rings of 2.338(6) (ScÀP: 2.802(2), 2.843(2) and2 .877(2) ). [57] These unusual metrical parameters indicatet hat significant charge transfer is present in 63, making the formal oxidation state am oot point, but the assignment of Sc I centres is au seful formalism to rationalise experimental data. Crystals of 63 exhibit af orest-greenc olour, and the intense absorption in the UV/Vis spectrum of ad ilute toluene solution (l max = 426 nm, e = 12 000 dm 3 mol À1 cm À1 )w as assigned to am etal to ligand charget ransfer band, which is typical of low oxidation state scandium; [57] ap entanes olution of dark-purple [Sc(C 3 tBu 3 P 2 ) 2 ]s imilarly exhibits am aximum absorbance at 571 nm and e = 15 000 dm 3 mol À1 cm À1 . [58] Solutions of 63 were determined to be EPR silent between 298 Ka nd 77 K, [57] whilst at oluene glass of [Sc(C 3 tBu 3 P 2 ) 2 ]a t 120 Kw as shown to exhibit ar ich and well-resolvedE PR spectrum with hyperfine coupling of the Sc-based unpaired electron to a1 00 %a bundant I = 7/2 45 Sc nucleusa nd additional splitting by four equivalent 31 Pn uclei (100 %a bundance, I = 1/ 2);t hese features weres imulated with g ? = 2.0098, g k = 1.9273, A ? ( 45 Sc) = 29.9 G, A k ( 45 Sc) = 52.9 Ga nd A( 31 P) = 7.2 G. [58] As olution of 63 was additional probedb yE vans method magnetic susceptibility,w here the value at 295 K( 3.98 m B )i sl ower than the predicted value of 4.47 m B for four unpairede lectrons arising from two isolated Sc I centres. [57] In contrast, the magnetic susceptibility measured at room temperature for at oluene solution of [Sc(C 3 tBu 3 P 2 ) 2 ]( 1.70 m B )i sf ully in accord with the expected value of 1.73 m B for a3 d 1 system with no orbital contribution, and amore clear-cut Sc II centre;however,the stability of this complex was attributed to the capability of the diphospholyl ligands to accept electron density from the metal. [58] In 2000, Deacon et al. reported the syntheses of co-crystallised [Sm(Cp*) 2 (h 2 -C 2 tBu 2 P 2 E)(THF)] (64-E,E = P, Sb;F igure 20), in 10 %y ield from the SET reaction of [Sm(Cp*) 2 (THF) 2 ]w ith [Tl(C 2 tBu 2 P 2 E)],w here the Sb/P ratio of Ei nt he Tl I precursor was approximately 4:1. [62] In the same paper,amixture of [Tl(C 2 tBu 2 P 2 E)] and Yb metal in THF was sonicatedf or 48 h, and upon work-up co-crystals of [Li(THF) 4 ][Yb(h 5 -C 2 tBu 2 P 2 E) 2 (h 2 -C 2 tBu 2 P 2 E)] (65-E,E= P, Sb;F igure 20) were isolated, with the Tl I precursor presumably contaminated with as ignificant amount of Li-containing compounds. [62] The authors made valiant efforts to determine the Sb/P ratios of Ei n64-E and 65-E by 1 Ha nd 31 PNMR spectroscopy,a nd found that for the former mixture Pw as in excess, whereas for the latter the Sb/P ratio was 2:1. This indicates that if pure [Tl(C 2 tBu 2 P 3 -1,2,4)] could be obtainedt hen it may react with [Sm(Cp*) 2 (THF) 2 ]t o give 64-P cleanly,b ut there are more variablest oe xplore for the synthesis of pure 65-P in the future.
Recrystallisation of 64-E gave several crystalso fa ntimonyfree 64-P,w hich were analysed by single-crystal XRD, whereas the SC-XRD dataset for 65-E showed the presence of both Sb and P, as wella so ther products containing C 2 tBu 2 P 3 rings. [62] The Sm III centre in 64-P is coordinated by two h 5 -Cp* ligands, a molecule of THF,a nd the 1,2,4-C 2 tBu 2 P 3 ring in an h 2 -fashion, with relatively long SmÀPd istances of 3.135(2) and 3.164(2) attributed to steric buttressing. The angle between Sm, the PÀ Pb ond mid-point and the mean plane of the 1,2,4-C 2 tBu 2 P 3 ring in 64-P is 143.88,w hich contrastsw ith the analogous approximately 1808 angle shown by similar h 2 -bound pyrazolyl complexes in the same paper such as [Yb(Cp*) 2 {C 3 HPh 2 N 2 -3,5}]; however,a lthough noteworthy,t he differing coordination spheres of thesec omplexes precludes am eaningful comparison. The Yb II centre in the anion of 65-E exhibits two h 5 -a nd one h 2 -bound C 2 tBu 2 P 2 Er ing, with the major Sb-containing dentate and the two alkynyl ligandsb ridge to form an asymmetric Eu 2 C 2 core. From the context of this review,the most interesting structural feature of 68 is that the two [C 2 tBu 2 P 3 ] À anions do not bind to the Eu II centres;i solated[ C 2 tBu 2 P 3 ] À rings had not previously been observedi nt he solid state. Remarkably, 68 is the only Eu phospholyl or arsolyl complex that has been structurally authenticated to date.

Planar cyclo-P 5 complexes
Although pentaphospholyls form au nique familyo fc omplexes that are somewhat independent of the organophosphorus derivatives in the rest of this review,w einclude them here for completeness;t here have been no reports to date of Ln or An cyclo-As 5 complexes. Metal cyclo-P 5 complexes are typically synthesised via the direct activation of white phosphorus or reactions with various P n -transfera gents; aromatic cyclo-P 5 anions are one of an umber of potential outcomes of these reactions along with ar ange of P n -boundf ragments, including Zintl clusters andr elateda romatic cyclo-P 4 dianions. [66] To the best of our knowledge only one Ln and one An complex that contain planar cyclo-P 5 ligands have been structurally authenticated to date ( Figure 21). [67,68] The sole example of as tructurally authenticated Ln complex containing ap lanar cyclo-P 5 ring, [{(Sm(Cp*) 2 } 3 (m:h 1 ,h 1 ,h 2 ,h 2 -cyclo-P 5 ){Mo(Cp)(CO) 2 } 3 ]( 69), was reported in 2015 by Roeskya nd co-workers to form as am inor product from the reduction of the P 2 unit in [{Mo(Cp)-(CO) 2 } 2 (m:h 2 ,h 2 -P 2 )] by [Sm(Cp*) 2 (THF) 2 ]. [67] Owing to disorder the metrical parameters from the single-crystal X-ray diffraction data for 69 are unreliable, butt he connectivity is clear-cut, with the planar cyclo-P 5 ring h 2 -bound to two Mo centres and h 1 -bound to at hird, with one of the Pa toms additionally h 1bound to as ingle Sm centre (SmÀP: 2.978(11) ). Unfortunately,o wing to the low yield of 69 and co-crystallisation with another reaction product, no additional characterisation data could be obtained.
Also in 2015, Liddle and co-workersr eported the synthesis of the dinucleari nverted sandwich uranium complex [{U[N(CH 2 CH 2 NSiiPr 3 ) 3 ]} 2 (m:h 5 ,h 5 -cyclo-P 5 )] (70)f rom the reduction of P 4 by the U III precursor [{U[N(CH 2 CH 2 NSiiPr 3 ) 3 ]}] in a1 :1 U/P ratio. [68] The planar cyclo-P 5 ring is disordered over two positions in the solid-state structureo f70,w hich again prevents meaningful analysis of PÀPd istances, and the UÀPd istances (range 3.250(6)-3.335(6) )a re relatively long owing to the bulky ancillary ligands. Surprisingly,t he UÀNd istancesi nt he ancillary ligandsa re in line with the presence of two identical U IV centres rather than the expected mixed U(III/IV) system for a cyclo-P 5 anion. All other analytical data for 70 (NMR and UV/ Vis/NIR spectroscopy,S QUID magnetometry) are also consistent with the formal presenceo ft wo U IV ions and a cyclo-P 5 dianion, although such formalismsa re often moot in systems with significant covalency.DFT studies of 70 showeds ignificant d-donation from filled uranium 5f orbitals of appropriate symmetry to the vacant p*e 2 orbitals of cyclo-P 5 ,w hich were again in line with significant charge transfer from uranium to the cyclo-P 5 ring. This is ac onsequence of both the ability of uranium to donate d-electron density using 5f orbitals and the superior electron accepting capability of cyclo-P 5 over Cp;t he isolation of 70 versus the absence of Cp from the family of bridging cyclo-C n R n ligands (n = 4, 6-8) in inverted sandwich An chemistry is significant. [69] It is noteworthy that non-planar cyclo-P 5 fragments wereobserved as part of P 10 moieties in the Sm complexes [{Sm(C 5 Me 4 R) 2 } 2 {Fe(Cp*) 2 } 2 {m:k 2 ,k 2 ,h 4 ,h 4 -P 10 )}] (R = Me, n Pr), where aP ÀPs ingle bond connects the two P 5 sub-units that are k 2bound to Sm and h 4 -bound to Fe;t hese complexes were prepared in 2013 by Scheer,Roesky andc o-workersf rom the reactions of parent [Sm(C 5 Me 4 R) 2 (THF) 2 ]w ith [Fe(Cp*)(P 5 )]. [70] Planar aromatic cyclo-E 4 dianions (E = P, As) have also been observed in f-blockc hemistry,a nd structurally characterised examples have been shownt oe xhibit ar ange of binding modes when bridging between metal centres,with the steric effects of ancillary ligandsdictating how these rings coordinate. [67,71]

Summary and Outlook
Althoughg roup 3a nd f-blockm etal phospholyl and arsolyl chemistry is immature compared with cyclopentadienyls and their derivatives, some differences between the families of complexes are already evident, which provide perspectives for future exploration. Firstly,t he ability of monophospholyls to stabilisel ow oxidation states has been demonstrated by the isolation and reactivity studies of rare examples of molecular Tm II complexes; [25] given the crucial role of Cp R ligands in the development of low oxidation state Ln and An chemistry, [6] the furthere xploitation of monophospholyl and -arsolyl ligands in synthesis and reactivity studies of analogous complexes is an obvious pathway to explore.F ine-tuning of reduction potentials by multiple heteroatom substitution in polyphospholyl and -arsolyl complexes could be au seful tool in stabilising more exotic low oxidations tate group 3a nd f-blockc omplexes, as has been demonstrated in the isolation of aS c I complex. [57] Polyphospholyl substituents are limited to tert-butyl groups to date owing to the current reliance on tBuCP to generate these ligands;t he development of facile syntheticr outes to aw ide range of polyphospholyl and -arsolyl ligandsw ould be transformative in developing their f-block chemistry to the same degree as monosubstituted analogues. Secondly, some interesting SMM properties have already been reported for Ln monophospholyl complexes; [39,48] in view of recent reports of Figure 21. Complexes 69 and 70. [67,68] Chem. Eur.J. 2021, 27,6645 -6665 www.chemeurj.org 2021 The Authors. Chemistry -AEuropean Journalp ublished by Wiley-VCH GmbH high blockingt emperature Ln SMMs containing Cp R ligands, [7] it is unsurprising that Ln SMMs containing polyphospholyl ligands have already been predicted and are targets for the synthetic community. [72] The optical properties of Ln phospholyl and arsolyl complexes will also vary from Cp R derivatives and one can speculate that these can also be tuned by variation of the ligand field to suit specific applications.
There is considerable chemical space to explore in An phospholyl and arsolyl chemistry.C urrently,t here are only structurally characterised examples of such complexes for An = U; the lack of Th complexes to date is surprising given the relatively low radiological hazard of Th,t he similarity of Th IV and U IV chemistry,a nd that solvated Th IV startingm aterials are readily synthesised from commercially available precursors. [73] For transuranic elements, the increasing radiological hazard across the An series limits investigations to specialist facilities, [1b] but the recent extension of Cp R chemistry to as tructurally authenticated Am III complex [74] indicatest hat phospholyls and arsolyls can also find successf or Np, Pu, Am and even beyond. Investigationsi nto An phospholyl and arsolyl redox chemistry is also currently limited to U III and U IV examples, where there are a wide range of An oxidation states to explore; [1b] for example, for U, Cp R complexes have been structurally authenticated from the + 2t ot he + 6oxidation state. [2d-g] There are also pathways for future exploration that are of relevance to both Ln and An phospholyl and arsolyl chemistry, which have not yet been fully exploited. Firstly,w es peculate that the heteroatom lone pairs in phospholyl and arsolyl rings in h 5 -bound complexes could be actively involved in reactivity profiles. Ln andA nC pa nd Cp R complexes have well-established applicability in aw ide range of hydroelementation and polymerisation reactions, including catalytic processes, [75] and low oxidation state Ln andA nc omplexes of these ligands have shown rich small molecule activation chemistry. [2d-g] We anticipate that future investigations with analogousL na nd An phospholyl anda rsolyl complexes will furnish results that complementa nd contrast with established Cp/Cp R chemistry,w ith the possible involvement of Pa nd As lone pairs in these reactions an exciting prospect. Secondly,t he presence of 100 % abundant spin-active 31 Pa nd 75 As nuclei in phospholyl and arsolyl rings provides new opportunities for quantification of fblock covalencyb yN MR [76] and pulsed EPR [77] spectroscopy.I n the latter case, this has already been achieved for Th and U Cp R complexes with 1.1 %a bundant 13 Cn uclei, thus the presence of 31 Po r 75 As would provide improved sensitivity,a sh as been shown in NMR spectroscopy covalency measurements for heteroatom-containing ligands. [76] Ta king into consideration the importance of minor differences in covalency between fblock elements to their technological applications, obtaining such data is crucial for future developments. [78] To conclude, althought he field of f-blockp hospholyl anda rsolyl chemistry is in its relative infancyi th as already provided important results that juxtapose with those of derivatised cyclopentadienyl f-block complexes. Given these past successes and the potential for wide variations in chemistry with heteroatom substitution,w er ealistically anticipate that other exciting resultswill surely follow in future investigations.