Expanding the Scope of Aluminum Chemistry with Noninnocent Ligands

Conspectus Aluminum is the most abundant metal in the earth’s crust at 8%, and it is also widely available domestically in many countries worldwide, which ensures a stable supply chain. To further the applications of aluminum (Al), such as in catalysis and electronic and energy storage materials, there has been significant interest in the synthesis and characterization of new Al coordination compounds that can support electron transfer (ET) and proton transfer (PT) chemistry. This has been achieved using redox and chemically noninnocent ligands (NILs) combined with the highly stable M(III) oxidation state of Al and in some cases the heavier group 13 ions, Ga and In. When ligands participate in redox chemistry or facilitate the breaking or making of new bonds, they are often termed redox or chemically noninnocent, respectively. Al(III) in particular supports rich ligand-based redox chemistry because it is so redox inert and will support the ligand across many charge and protonation states without entering into the reaction chemistry. To a lesser extent, we have reported on the heavier group 13 elements Ga and In, and this chemistry will also be included in this Account, where available. This Account is arranged into two technical sections, which are (1) Structures of Al–NIL complexes and (2) Reactivity of Al–NIL complexes. Highlights of the research work include reversible redox chemistry that has been enabled by ligand design to shut down radical coupling pathways and to prevent loss of H2 from unsaturated ligand sites. These reversible redox properties have in turn enabled the characterization of Class III electron delocalization through Al when two NIL are bound to the Al(III) in different charge states. Characterization of the metalloaromatic character of square planar Al and Ga complexes has been achieved, and characterization of the delocalized electronic structures has provided a model within which to understand and predict the ET and PT chemistry of the NIL group 13 compounds. The capacity of Al–NIL complexes to perform ET and PT has been employed in reactions that use ET or PT reactivity only or in reactions where coupled ET/PT affords hydride transfer chemistry. As an example, ligand-based PT reactions initiate metal–ligand cooperative bond activation pathways for catalysis: this includes acceptorless dehydrogenation of formic acid and anilines and transfer hydrogenation chemistry. In a complementary approach, ligand based ET/PT chemistry has been used in the study of dihydropyridinate (DHP–) chemistry where it was shown that N-coordination of group 13 ions lowers kinetic barriers to DHP– formation. Taken together, the discussion presented herein illustrates that the NIL chemistry of Al(III), and also of Ga(III) and In(III) holds promise for further developments in catalysis and energy storage.

of group 13 3+ ions in the overall aromaticity of the molecule is described along with the f irst reported squareplanar Ga complexes.

INTRODUCTION
Recent decades have seen extensive research and development in main-group reaction chemistry, where accessible oxidation states and bonding modes have been greatly expanded as a result of efforts in synthetic chemistry. 4,5The element aluminum, Al, has many features that suggest it would be appealing for large-scale processes if its chemistry were developed in a way that enabled those applications.For example, Al has a very high natural abundance; with the earth's crust comprising 8% Al, it is the third most abundant element. 6t is domestically available in many countries and provides security in supply chains.Also, Al is relatively inexpensive with a current price of just $2,213/ton. 7The purification and extraction of Al are notoriously high-energy processes driven electrochemically, but recent innovations have both lowered the energy inputs through advances in process engineering and reduced their costs with renewably derived electricity.This Account focuses on the chemistry of noninnocent ligand (NIL) Al complexes, particularly how electronic structures and reaction pathways can be expanded using NILs.To a lesser extent, this Account will also discuss the NIL chemistry of gallium, Ga, and indium, In, where the comparison is useful.Both the lightest and the heaviest of the group 13 elements, boron, B, and thallium, Tl, respectively, have not been characterized as M(III) ions with NILs, to the best of our knowledge.We begin with a brief overview of the known NIL chemistry of Al.Sections 1.1 and 1.2 include strategies for the syntheses of iminepyridine−Al compounds with reversible electron transfer (ET) and proton transfer (PT) chemistry.Section 1.3 summarizes and explains some of the electronic properties of the iminepyridine−Al complexes that can be accessed.The reaction chemistry of NIL−Al complexes can be categorized into three main types, as outlined in Section 2: ET, PT, and proton-coupled ET (PCET) reactions.The subsections discuss each of these reaction types.Section 2.1 describes ligand ET chemistry; section 2.2 describes Al−ligand cooperative bond activation for catalysis, which is enabled by ligand PT chemistry; and section 2.3 focuses on ligand-based PT coupled with ET chemistry, which is also organohydride chemistry.The reader is also referred to several excellent reviews on the coordination chemistry of Al, which have focused on the synthesis of coordination compounds, 8−11 Al-mediated catalysis, 12−17 and metal−ligand cooperative bond activation. 18,19This Account contains very little overlap with those prior reports.
ET and PT elementary reaction steps are common to many chemical transformations, but they are difficult to accomplish with many of the main-group elements, including Al. Redox cycling through Al(I) and Al(III) requires significant energy input and accesses highly unstable and air-sensitive compounds of Al(I).Despite these difficulties, several elegant reports of redox chemistry mediated by the Al(I/III) couple have been reported (Scheme 1). 20,21In these examples, a β-diketiminate ligand supports both the Al(I) and Al(III) redox states without participating directly in the redox chemistry via ligand-based ET.

Overview of NILs That Have Been Reported with Al
−26 Some examples of NILs that have been characterized with Al(III) are highlighted in this section as an introduction to the remainder of the Account.Iminepyridine (abbreviated as IP, Scheme 2) was initially reported in complexes with Mg, and diiminepyridines (abbreviated as I 2 P, Scheme 2) were reported in metal−ligand complexes with firstrow transition elements by Wieghardt and co-workers. 27Maingroup I 2 P complexes first appeared with Ga, in work by Richeson and co-workers. 28Work by Chirik in this area has developed the reaction chemistry of transition-element I 2 P complexes, which includes the 2 + 2 cycloadditions of dienes and olefin catalysis, among many others. 29,30Other NILs that have been studied with Al(III) include dipyrazolylpyridines (generally abbreviated here as pz 2 P), acenaphthene-1,2diimines (BIAN), and bis(3,5-di-tert-butyl-2-phenol)amine (ONO) (Scheme 3); the chemistry of these NILs will be mentioned in the appropriate sections of this Account.imine C atom. 23,22,27,29,30This is denoted in the ligand abbreviation as a preceding superscript when needed (Scheme 2).Substituents on the imine N atom can be denoted as a preceding subscript.For example, the most commonly used I 2 P ligand has Me on the imine C atoms and dipp (defined in Scheme 1 and 2) on the imine N atoms: dipp Me I 2 P. The analogous ligand with only H on the imine C, dipp H I 2 P, is rare in the literature, likely because of the potential for radical C−C coupling (Scheme 4).When coordinated to a metal ion (transition metal or group 13), Me I 2 P ligands commonly lose H 2 in a chemically irreversible process, and the resulting CH 2 I 2 P ligand form can support extensive metal-based chemistry, as described in section 1.1.
When discussing the chemistry of Ar Ph I 2 P in this Account, we often leave off the superscript of the understood phenyl group (Scheme 4).To simplify notation, similarly, the Ar group on the imine N is commonly dipp and is also left off.Exceptions are when variation of the Ar group is a key aspect of the work, and subscripts are then included.Conveniently, all the I 2 P derivatives can be easily modified at the imine N. Ar groups are incorporated via a condensation reaction with aryl amines as the final step in I 2 P synthesis.Larger Ar groups, such as dipp in dipp I 2 P, support the formation of 1:1 ligand−Al complexes.Smaller substituents at the 2 and 6 positions of the aryl ring (e.g., 2-MeOPh) support the formation of 2:1 ligand−Al complexes, which are coordinatively saturated and sixcoordinate at the Al center.
Neutral ligand complexes are commonly prepared by the addition of two equivalents of a group 13 metal chloride, MCl 3 , to one equivalent of a neutral ligand, L, to produce a salt of the form (LMCl 2 )(MCl 4 ) 2,31 (Scheme 5).Reduced ligand complexes can then sometimes be prepared by reduction of the (LMCl 2 )(MCl 4 ) salt with chemical reductants. 2,32As examples, dipp I 2 P complexes of Al can be synthesized by addition of two equivalents of AlCl 3 to one equivalent of dipp I 2 P, which affords [( dipp I 2 P)AlCl 2 ](AlCl 4 ).In a second common approach, reduced ligand complexes are prepared by reduction of the ligand with an alkali-metal reductant.For example, Na can be used to afford Na + L − or (Na + ) 2 L 2− , which can then undergo salt metathesis with MCl 3 to afford NaCl and (L n− )MCl (3−n) . 33In some instances, a one-pot method involving a mixture of L, MCl 3 , and a reductant is more effective for the formation of (L n− )MCl (3−n) , as in the syntheses of complexes with the fluorinated PhFd 5 I 2 P ligands (introduced in Scheme 2). 34A third common synthetic approach is where the starting ligand is already isolated in a reduced and protonated state: in that case, deprotonation by a base such as NaH or KH affords the alkali-metal ligand salt, which will react with MCl 3 .This deprotonation approach has been employed in the synthesis of bis(enol)amine, 35 and in the synthesis of Al complexes with an iminocatecholate pincer ligand, 36 as examples.The three approaches described in this section have been employed extensively across many examples.However, there are some examples in which radical ligand intermediates generated by the reduction reactions are unstable toward dimerization, 37,38 or decomposition.Ligand design can play an important role in the successful metalation of NILs, such as the inclusion of steric bulk to physically block  radical reactions 39 or the addition of stabilizing organic functional groups like phenyl in place of an H atom. Once obtained, NIL complexes can sometimes be oxidized or reduced to obtain another charge state in a series of compounds.
Solvent choice plays a role in the outcomes of the reactions described above.More polar solvents, such as tetrahydrofuran (THF) or dimethoxyethane (DME), promote faster ligand reduction than diethyl ether (Et 2 O) or alkane or aromatic solvents.Coordinating solvents, such as THF, may also coordinate to the metal.An example here is the formation of five-coordinate ( dipp I 2 P 2− )AlCl(THF) and four-coordinate ( dipp I 2 P 2− )AlCl; they are obtained from the same reaction performed in THF or Et 2 O, respectively. 35,40When the group 13 ion is coordinatively saturated, as in octahedral complexes with tridentate ligands, then some kinetic stability toward strongly donating solvents, such MeCN and H 2 O, has been observed, but these are rare examples. 1,34

Electronic Properties of NIL−Al Complexes
While the unique reactivity afforded by NILs bound to Al(III) [or Ga(III)] is the main subject of this Account, this reactivity can sometimes be rationalized or predicted if an appreciation of the electronic structures of the compounds is presented first.The metal−ligand bonding has covalent character between the metal valence p-orbitals and the ligand-based p-orbitals that make up the π-bonding in the ligands.Those interactions impart properties, such as metalloaromaticity, exchange coupling pathways, and Class III delocalization, and they enable tuning of the Lewis acidity of the Al(III) or Ga(III) ion by modification of the I 2 P ligand's electronic structure.
We first consider the geometric and electronic structures of the square planar (SP) Al and Ga complexes (I 2 P 2− )AlCl, (I 2 P 2− )AlH, (I 2 P 2− )AlI, (I 2 P 2− )GaCl, and (I 2 P 2− )GaH (Chart 1). 2 The other four-coordinate molecules, (I 2 P 2− )Al-(NHdipp), 41 (I 2 P 2− )Al(PHPh), and (I 2 P 2− )Al(PHMes) (Chart 1) 42 are distorted from SP. NHdipp, PHPh, and PHMes are 2,6-diisopropylphenylamide, phenylphosphide, and mesitylphosphide, respectively.The NIR absorption spectra for all of these compounds display broad and intense electronic transitions around 1000 nm, which have been assigned as ligand−metal charge transfer (LMCT) bands because the reduced ligands donate electron density into the metal-based p-orbitals.Observations from single-crystal X-ray diffraction studies suggest that the coordination of the monodentate ligands to the metal center influence the electronic structures of the I 2 P ligand.Four-coordinate metal centers have a delocalized I 2 P electronic structure, and five-coordinate metal centers show alternating single-and double-bond character (Chart 1).Another interesting feature of all the fourcoordinate group 13 structures is the N py −M−X angle (where N py is the pyridyl N and X is a monodentate ligand), which deviates further from θ = 180°with stronger π-donor ligands.Presumably, the filled π-symmetry orbitals on X interact with the empty p-orbital on M, which makes this interaction energetically favorable.The results from 1 H NMR and NIR studies show that Lewis acidity does not change greatly as the ion is varied from Al(III) to Ga(III) or when the X ligands vary.According to the measurement performed with the Gutman−Beckett method, the Lewis acidity of (I 2 P 2− )AlX (X = Cl, H or I) and (I 2 P 2− )GaX is lower than the Lewis acidity of other reported three-and four-coordinate Al compounds, and this presumably stems from the metalloaromatic character.
The structures and properties described in the paragraph above are consistent with metalloaromaticity.Metalloaromatic molecules are roughly defined as those with a planar structure, near-equal bond lengths for equivalent bond types, a broad and low-energy LMCT band, and negative values for the nucleusindependent chemical shift (NICS) for all nuclei in the molecule. 43The data we collected for four coordinate compounds in Chart 1 are consistent with a metalloaromatic structure.Theoretical studies of I 2 P and Al or Ga have reported negative NICS values, which supports the assignment of metalloaromaticity. 44Dianionic metalloles of Al(III) and Ga(III), as well as the tropylium cation analogues (aluminepin and gallepin), are other groups of metalloaromatic compounds in which empty atomic Al and Ga p-orbitals participate in πbonding with planar organic ligands. 45SP complexes comprising a d-block metal and I 2 P ligand exhibit antiferromagnetic coupling between ligand radicals and an Fe center rather than delocalization. 46A comparison of group 14 and group 4 cyclopropene analogues showed that group 4 complexes are closer to classical organometallic π complexes while group 14 complexes are aromatic. 47These examples highlight the unique aromatic electronic properties afforded when the metal center has valence p-orbitals as in main-group coordination complexes, in contrast to the bonding and electronic structures found in d-block coordination compounds.
1.3.1.Exchange Coupling and Ligand Mixed Valency.When two ligands coordinate to a single group 13 ion, unpaired electrons on the ligands can interact with each other.These interactions occur via exchange coupling in the case of two unpaired electrons or by varying degrees of delocalization for a single unpaired electron.The earliest investigation of these possibilities was reported in the work of Pierpont and coworkers with tri-and dicatecholate complexes of Al(III) and Ga(III) (Chart 2). 48For IP − and I 2 P − ligands paired with group 13 centers, magnetic exchange coupling is observed between the ligand-based unpaired electrons.For (IP − ) 2 MX (M = Al or Ga), the exchange coupling is antiferromagnetic, and the proposed pathway for exchange is through the M−X σ* orbital.−51 Magnetic exchange coupling has also been observed between two radicals on [( o-PhOMe I 2 P − ) 2 M] + complexes (M = Al, Ga).Variable-temperature EPR spectroscopy revealed a net ferromagnetic interaction between the ligand radicals with J coupling values of 56  The N py −M−X angle is marked as θ, and the electronic structure of the I 2 P ligand is distinct for the two geometries.zoquinone biradical complexes with group 13 metals range from −68.6 to −128 cm −1 for Ga complexes and from −122 to −179 cm −1 for Al. 52Coupling of octahedral tridentate diaryl amine Ga complexes display a coupling of ∼200 cm −1 . 53hen two ligands bound to a metal have different charge states, there is the possibility of mixed-valent character and electron delocalization.Both [(IP)(IP − )Al(Me)][BAr 4 ] and [(IP 2− )(IP − )Al(OH)][Na(THF)(DME)] 38,49 have localized Class I electronic structure, and that assignment was made based on the absence of an intervalence charge transfer (IVCT) band in the NIR region.Complexes ( Ar I 2 P 2− )( Ar I 2 P − )-M are Class III delocalized when M = Al or Ga, and delocalization is lowered to Class II for M = In (Chart 3). 1,34e proposed that the near-octahedral coordination geometry of these I 2 P Al and Ga complexes provides an orbital pathway through metal-based p-orbitals to support electron delocalization.Delocalization effects are diminished with the larger In ion, which permits non-octahedral geometry about the central ion.In addition, delocalization is most efficient when the overall molecular charge is lowest and the fewest electrons populate the π* ligand framework.For Al(III), an analogous series of compounds was prepared with electron-withdrawing and -donating ligands, PhFd 5 I 2 P and p-PhNMed 2 I 2 P, respectively, where delocalization was reduced and unchanged, respectively.Access to 5 charge states separated by one electron in octahedral I 2 P Al complexes also lends to their use as flow battery analytes which can store many electrons to achieve high charge density. 54

REACTION CHEMISTRY AND APPLICATIONS OF NIL (GROUP 13) COMPLEXES
This section explores the reaction chemistry and electrochemical energy storage methods that are enabled by our access to the many charge states in NIL (group 13) complexes.For ease of reference, this section has been divided into three parts: section 2.1 discusses ligand-based ET for reaction chemistry and energy storage; section 2.2 features ligand-based PT that enables catalysis via metal−ligand cooperative bond activation; and section 2.3 presents ligand-based PT coupled with ET, and this includes hydride transfer chemistry.

Ligand-Based ET for Reaction Chemistry
There are a number of reaction classes that traditionally proceed with the support of oxidation-state changes at a transition element center. 55,56One of these is group transfer chemistry, where a two-electron oxidation-state change of a metal center is concomitant with the formation of a new metal−ligand bonding interaction.This reaction type is important in biological enzymes, e.g., cytochrome P450, 57 and in synthetic organic chemistry where epoxidation of alkenes and aziridination chemistry via N atom transfer has been widely studied.A variation of this common reaction type in NIL complexes involves the ligand supporting the ET events that enable the bond-breaking and -making events at the metal, as can be witnessed in an early report with Zr. 58,59 The first report of this type for group 13 appeared in 2011. 50The oxidation of [(IP  50 Along with the benefit of reproducible oxidations, the use of metal salts allows for a wide variety of X groups to be employed to form Al−X bonds.We report reactions of X = Cl, CCPh, N 3 , SPh, and NHPh (Scheme 6).Reactions of either [(IP 2− )2Al] − or [(IP 2− )2Ga] − with tetramethylthiuram disulfide and other oxidants also afford two-electron group transfer products. 51,60,61n I 2 P-supported Al chemistry, reaction of the (2,2,6,6tetramethylpiperidin-1-yl)oxyl radical, TEMPO, with (I 2 P 2− )-AlCl(THF) yields (I 2 P − )AlCl(TEMPO).Reduction of this complex by an equivalent of Na produces (I 2 P 2− )Al(TEMPO), providing a unique example of ligand substitution by redox cycling. 62Subsequent isolation of [( H I 2 P 3− )Al(TEMPO)] − represents a rare example of a singly ligated (I 2 P 3− )−Al coordination complex.This was made possible by group transfer chemistry, which afforded a reductively stable Al−O interaction between the Al center and TEMPO.Ligand-based redox chemistry has also been reported with the BIAN and the ONO ligands (Scheme 3).As examples, reduction of aromatic ketones, R 2 CO, to pinacolates is reported with (BIAN

Ligand-Based PT: Catalysis via Metal−Ligand Cooperative Bond Activation
Having discussed ligand-based ET chemistry, we now discuss PT chemistry that is enabled by the Lewis-basic sites available in reduced NIL complexes of Al(III).Ligand O and N atoms can function as PT sites, enabling metal−ligand cooperative bond activation (MLCBA).In this reaction type, Al(III) behaves as an electrophile, and the Lewis-basic reduced ligand site is a nucleophile resulting in heterolytic bond cleavage of the substrate (Scheme 7).Metalloaromaticity serves to reduce the Lewis acidity of the formal Al(III) ion in complexes such as (I 2 P 2− )AlH, (I 2 P 2− )AlCl, and (I 2 P 2− )AlI, and this feature of the electronic structure is important in promoting MLCBA instead of substrate coordination to the Al(III) ion.Polar bond activation is favorable only in some instances, such as in the work of Milstein and co-workers using Ru and Fe complexes (Scheme 7). 64In those examples, the MLCBA reaction results in de-/rearomatization of the catalyst-supporting ligand.Interestingly, in all of the examples of MLCBA described below with Al(III), the polar bond activation step also induces a de-/rearomatization of the catalyst-supporting ligand.Others have also recently shown that metalloaromatic group 13 compounds can exhibit MLCBA type reactivity which varies based on the nature of the substrate and the metallocycle. 65,66sing I 2 P 2− as the supporting ligand for Al(III), MLCBA toward weakly acidic RNH 2 or ROH substrates results in O−H or N−H bond activation, respectively. 41,67−69 As a specific example, consider the reaction of an amine or alcohol with (I 2 P 2− )AlH(THF) (Scheme 8).The reaction of (HI 2 P 2− )AlH-(THF) with RNH 2 or ROH affords (HI 2 P − )AlH(NHR) or (HI 2 P − )AlH(OR), respectively.Protonation of the amido group of the reduced I 2 P 2− ligand has occurred and is shown by an italicized superscript on the ligand label (Scheme 8). 67In the case where the substrate is weakly acidic such as when using RNH 2 to produce (HI 2 P − )AlH(NHR), a follow-up reaction produces H 2 and (I 2 P 2− )Al(NHR) when mild heat is applied.The products obtained from more acidic substrates such as alcohols or formic acid, (HI 2 P − )AlH(OR), do not lose H 2 even at elevated temperatures.However, (HI 2 P − )AlH(OR) does react with further equivalents of acid to give products of the form (HI 2 P − )Al(OR) 2 with the release of H 2 .With sufficiently low pK a , a further reaction produces (H 2 I 2 P)Al-(OR) 2 , such as when formic acid is employed (Scheme 8, second line). 41,68The stoichiometric MLCBA reactions described above do in some cases initiate catalysis, for example, in the dehydrogenative coupling of amines when benzylamine is the substrate 41 and in the dehydrogenation of HCOOH. 70eactions of (I 2 P 2− )AlCl follow a different pathway after the initial MLCBA step (Scheme 8, bottom).Following MLCBA, there is no option for the release of H 2 because the Al center does not have a hydride ligand.Reactions of (I 2 P 2− )AlCl with alcohols including iPrOH or Ph 2 CH(OH) afford ( H 2 I 2 P)Al-(OR) 2 Cl; whereas the reaction of (I 2 P 2− )AlCl with an amine such as aniline produces an equilibrium mixture of the substrates and ( H I 2 P − )Al(NHR)Cl.A further difference is Scheme 8. Reactivity Summary of (I 2 P 2− )AlH(THF) and (I 2 P 2− )AlCl with Alcohols or Amines a a H H 2 is added to the label; this denotes the ligand is once or twice protonated, respectively.that (I 2 P 2− )AlCl reacts more readily with alcohols whereas (I 2 P 2− )AlH reacts more readily with amines.A recent report described (I 2 P 2− )AlCl(THF) as a precatalyst for the Meerwein−Ponndorf−Verley oxidation of isopropanol to acetone. 3,71roup 13 complexes with other anionic ligands have also been reported to support PT reactions.Fedushkin and coworkers have reported several MLCBA examples with (BIAN)−Al and (BIAN)−Ga complexes (BIAN = example shown in Scheme 3) for the activation of RE−H 72 or C−C π bonds. 73An illustrative example of this MLCBA reactivity is the reaction of (BIAN 2− )AlEt(Et 2 O) with amines and alcohols, RE−H (Scheme 9, bottom).This MLCBA reactivity was instrumental in the ring-opening polymerization (ROP) catalysis performed by dimeric (BIAN)Al−Al(BIAN). 74In another example, calix [4]pyrrolato Al complexes, [(c 4 p 4− )Al] − , undergo protonation at a carbon atom of the pyrazole upon activation of the O−H bond in alcohols, ROH, where R = EtOH, iPrOH, tBuOH, 4-methylbenzyl alcohol, 4-nitrobenzyl alcohol, bis(4-methoxyphenyl)methanol 4-bromophenol, and benzoic acid (Scheme 9, top right). 75This selectivity, C versus N protonation, is attributed to the strained cyclic geometry of the ligand.N protonation of [(c 4 p 4− )Al] − would result in an energetically unfavorable distortion from planarity.Greb and co-workers have also shown that [(c 4 p 4− )Al] − can bind reversibly with aldehydes and ketones in a bidentate fashion to form C−C and Al−O bonds (Scheme 9, top left). 76witching between the σ and 1,2-adducts was controlled by the presence or absence of a Li + ion (Scheme 9, left).[(c 4 p 4− )Al] − was found to be an active catalyst for the hydroboration of 4nitrobenzaldehyde, 4-dimethylamino benzaldehyde, and acetone with pinacolborane.Slower reaction rates were obtained for aldehydes with electron-withdrawing substituents, presumably resulting from the preference for activated aldehydes to adopt catalytically deactivated 1,2-binding.This observed rate is opposite the general trend that activated aldehydes react faster with hydrides, and it provides an exciting prospect for how substrate binding and MLCBA can influence selectivity in future reactions involving noninnocent Al complexes.

Ligand Proton-Coupled Electron Transfer (PCET): Hydride Transfer (HT) Chemistry
When a hydride ion is transferred to a C atom of pyridine, a dihydropyridinate (DHP − ) is formed.The resulting DHP − features a tetrahedral C with two H atoms on the C of the former pyridine ring.DHP − chemistry has a long history that is primarily derived from efforts to model the organic cofactor nicotinamide adenine diphosphate (NAD + ) which is known to be an efficient HT agent. 77Those model systems suffer almost universally from large overpotentials needed for the reduction of the NAD + model analog to generate the NADH model equivalent via a transfer of two electrons and one proton. 78,79ork on pyridyl (group 13) compounds has previously taught us that the p-valence orbitals of Al and Ga have a good energy match and alignment with N-donor reduced ligands.Interactions of these p orbitals with the ligand orbitals lower the potential needed to inject multiple electrons into the I 2 P ligands featured in much of this Account.This section describes our ongoing efforts to employ this strategy of using group 13 ions with organic ligands to lower barriers to pyridine reduction and the energy for cycling through DHP − intermediates for hydride transfer (HT) to substrates.Our goals in this work are 2-fold: to model the redox chemistry of NAD + and to develop functional HT catalysts for applications in organic synthesis.
To facilitate DHP − formation, we turned to the dipyrazoylylpyridine ligand platform (pz 2 P as defined in Scheme 3 and related ligands).Weakly basic pyrazole groups on pz 2 P allow ET and PT chemistry to be localized at the pyridyl ring in group 13 complexes of pz 2 P-type ligands (Scheme 10). 80This is in contrast to the I 2 P chemistry discussed above where the imino donors are reduced before the pyridyl donor of the pincer ligand.In one report a structural analog to NAD + , [(pz 2 P)CH 3 ]OTf, was synthesized by alkylating pz 2 P at the central pyridyl N. Cyclic Voltammetry (CV) of [(pz 2 P)CH 3 )]OTf displayed irreversible reduction events that have a potential E p that is ∼900 mV more anodic relative to ligand-based reduction events of pz 2 P-type complexes of divalent metals. 81Experiments probing the origin of the irreversibility of the reduction event suggest the formation of DHP − species: (pz 2

Accounts of Chemical Research
several contributing factors are possible.These include two main ideas: (1) (pz 2 H P − )CH 3 , (pz 2 H P − )AlCl 2 , and (pz 2 H P − )-GaCl 2 are all cations, whereas the divalent metal complexes are uncharged; and (2) AlCl 2 + , GaCl 2 + , and CH 3 + interact with pz 2 P via valence p-orbitals, whereas valence d-orbitals are involved for the Zn II and Fe II complexes.Greater delocalization through the p-orbital manifold is likely based on our prior observations with metalloaromatic Al and Ga I 2 P complexes.
Electrocatalytic activity of [(pz 2 P)AlCl 2 (THF)]AlCl 4 for the H 2 evolution reaction (HER) offers insight into how complexes may function as HT catalysts (Scheme 10 bottom).Other redox-inactive metal complexes of NILs have also since been reported to exhibit HER and carbon dioxide reduction reaction (CO 2 RR).CO 2 RR behavior has been reported in work by Brudvig, 82 Grapperhaus, 83 Kumar, 84 and their coworkers.Reports of catalytic HT from (DHP − )−Al complexes are also known, but those may proceed via Al−hydride or Al− (DHP − ) intermediates. 85

SUMMARY AND OUTLOOK
This Account outlines the chemistry of Al(III) when combined with NILs.NIL−Al(III) complexes have been isolated in 6 charge states, and these have led to insights into various electronic properties of the compounds.As an example, the importance of the empty Al(III) valence p-orbitals in facilitating electronic coupling and delocalization between Al and ligand and between two ligands that are connected by a single Al center has been demonstrated.Metalloaromaticity has been used as a framework to rationalize some of the structures and reactivity properties, and probes of metalloaromatic (I 2 P 2− )MX complexes demonstrate that more electron density in the M p-orbital leads to decreased Lewis acidity which can be used to tune future metalloaromatic complexes to control reaction pathways and substrate scope.Another material property that can be influenced by I 2 P is stability toward coordinating or protic solvents.Octahedral NIL−Al I 2 P complexes are stable in MeCN or in the presence of trace water despite highly reduced oxidation states. 54he unusual electronic properties of the new molecular Al complexes described above have given rise to new reaction chemistry.NIL−Al(III) complexes support ET reactions, including those where both 1 and 2 electron oxidation and reduction reactions occur.Ligand-based PT chemistry supports mechanisms following MLCBA pathways as an entry point to many reaction types.
For example, formic acid and benzyl amine dehydrogenation and transfer hydrogenation have all been demonstrated.Along with ET and PT reactivity, control of ligand pK a values can influence the PCET vs PT reaction chemistry, ligand-based PCET chemistry offers an attractive method to lower the potentials needed to drive the formation of ligand-based C−H hydrides, and future work will expand HT reactivity of C−H hydride species.Through ligand design and correct reaction conditions, the chemistry of molecular Al continues to expand through the aid of noninnocent ligand complexes.We foresee NIL continuing to expand the chemistry of Al by providing new material properties, substrate scope, and catalytic reaction pathways.

Scheme 6 .
Scheme 6. Group Transfer Chemistry of Selected IP and I 2 P Complexes a
not specifically mentioned above, the foregoing examples of I 2 P-supported transition-element chemistry employ primarily I 2 P ligands that are Me substituted at the a Ligands are shown here with the commonly observed charge states that are accessible with group 13 ions, Al(III) and Ga(III).