Coordination- and Redox-Noninnocent Behavior of Ambiphilic

Article pubs.acs.org/accounts

Coordination- and Redox-Noninnocent Behavior of Ambiphilic Ligands Containing Antimony J. Stuart Jones and François P. Gabbaï* Department of Chemistry, Texas A&M University, College Station, Texas 77843-3255, United States CONSPECTUS: Stimulated by applications in catalysis, the chemistry of ambiphilic ligands featuring both donor and acceptor functionalities has experienced substantial growth in the past several years. The unique opportunities in catalysis offered by ambiphilic ligands stem from the ability of their acceptor functionalities to play key roles via metal−ligand cooperation or modulation of the reactivity of the metal center. Ligands featuring group 13 centers, most notably boranes, as their acceptor functionalities have undoubtedly spearheaded these developments, with remarkable results having been achieved in catalytic hydrogenation and hydrosilylation. Motivated by these developments as well as by our fundamental interest in the chemistry of heavy group 15 elements, we became fascinated by the possibility of employing antimony centers as Lewis acids within ambiphilic ligands. The chemistry of antimony-based ligands, most often encountered as trivalent stibines, has historically been considered to mirror that of their lighter phosphorus-based congeners. There is growing evidence, however, that antimony-based ligands may display unique coordination behavior and reactivity. Additionally, despite the diverse Lewis acid and redox chemistry that antimony exhibits, there have been only limited efforts to explore this chemistry within the coordination sphere of a transition metal. By incorporation of antimony into the framework of polydentate ligands in order to enforce the main group metal−transition metal interaction, the effect of redox and coordination events at the antimony center on the structure, electronics, and reactivity of the metal complex may be investigated. This Account describes our group’s continuing efforts to probe the coordination behavior, reactivity, and application of ambiphilic ligands incorporating antimony centers. Structural and theoretical studies have established that both Sb(III) and Sb(V) centers in polydentate ligands may act as Z-type ligands toward late transition metals. Although coordinated to a metal, the antimony centers in these complexes retain residual Lewis acidity, as evidenced by their ability to participate in anion binding. Anion binding events at the antimony center have been shown by structural, spectroscopic, and theoretical studies to perturb the antimony−transition metal interaction and in some cases to trigger reactivity at the metal center. Coordinated Sb(III) centers in polydentate ligands have also been found to readily undergo two-electron oxidation, generating strongly Lewis acidic Sb(V) centers in the coordination sphere of the metal. Theoretical studies suggest that oxidation of the coordinated antimony center induces an umpolung of the antimony−metal bond, resulting in depletion of electron density at the metal center. In addition to elucidating the fundamental coordination and redox chemistry of antimony-containing ambiphilic ligands, our work has demonstrated that these unusual behaviors show promise for use in a variety of applications. The ability of coordinated antimony centers to bind anions has been exploited for sensing applications, in which anion coordination at antimony leads to a colorimetric response via a change in the geometry about the metal center. In addition, the capacity of antimony Lewis acids to modulate the electron density of coordinated metals has proved to be key in facilitating photochemical activation of M−X bonds as well as antimony-centered redox-controlled catalysis.

■

INTRODUCTION AND BACKGROUND Over the past decade, the field of Z-ligand chemistry1 has experienced a renaissance prompted by the development of new ambiphilic ligand platforms combining both L-type and Ztype ligands.2−9 These platforms have been used to generate complexes in which the L-type ligands serve to position a metal ion or atom in close proximity to the Lewis acidic Z-type ligand site (Scheme 1). The success of this approach is illustrated by a diverse series of complexes that incorporate a group 13 or group 14 element as a Lewis acidic site. When appropriately designed, these molecules display a strong donor−acceptor © XXXX American Chemical Society

Scheme 1. Known Types of Ambiphilic Complexes

interaction involving the metal as the donor and the Z-type ligand as the acceptor. Received: December 15, 2015

A

DOI: 10.1021/acs.accounts.5b00543 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 2. Syntheses of 1 and [2][PF6]

The most developed class of such complexes are those containing a tricoordinate boron atom as the Z-type ligand.5,7,8,10 In addition to displaying unique metal → boron donor−acceptor interactions, complexes supported by such ligands are also emerging in the domain of catalysis,10 where the boron center can promote heterolytic bond activation processes,11 facilitate redox processes at the adjacent transition metal center,12 or increase the Lewis acidity of the metal center by σ-inductive effects.13 Fascinated by these advances, we questioned whether a related chemistry based on ambiphilic ligands featuring a heavy group 15 element as the Lewis acidic site could be developed.14 Our interest in this class of ligands was reinforced by the rich coordination and redox chemistry that heavy group 15 elements display.15,16 In this Account, we describe the results that we have obtained using polydentate ligands incorporating antimony. While stibines are usually regarded as weaker σ-donating phosphine analogues,15 our contributions to this chemistry underscore the noninnocence of these ligands. Stibines supported by donor frameworks are readily converted into Ztype ligands by oxidation and participate in coordination events without dissociation of the adjacent metal center. In addition to investigating how this noninnocent behavior affects the manner in which the antimony center interacts with coordinated transition metals, we have also exploited the noninnocence of these ligands in the context of anion sensing and redoxcontrolled catalysis.

Scheme 3. Coordination Chemistry at the Mercury Center in [2]+

■

ANTIMONY ONIUM SPECIES AS σ-ACCEPTORS Our investigation into the coordination behavior of ambiphilic antimony ligands was initiated by a series of studies on compounds featuring Sb(V) centers tethered to mercury or gold centers by rigid 1,8-naphthalenediyl linkers.17−19 Having previously employed the 1,8-naphthalenediyl platform to enforce the proximity of Lewis acidic centers in bidentate Lewis acids,20 we questioned whether tethering a late, electronrich d-block metal to a Lewis acidic Sb(V) center using this platform would promote the formation of otherwise unstable metal → Sb interactions. The gold− and mercury−stibonium species 1 and [2][PF6] were prepared in one-pot syntheses by stepwise reaction of 1,8-dilithionaphthalene with Ph3SbBr2 and AuI or HgII synthons (Scheme 2).17,18 Structural studies showed that the Au−Sb distance in 1 (avg. 2.7616(8) Å) is significantly shorter than the Hg−Sb distance in [2][PF6] (3.0601(7) Å), supporting the greater metallobasicity of gold(I) versus mercury(II). The juxtaposition of an electron-deficient tetraarylstibonium with an electron-rich, heavy d-block center results in an enhancement of the Lewis acidity of the d-block center.19 The organomercury−stibonium compound [2]+, in contrast to typical organomercury species, is able to engage anionic (Cl−, Br−, I−) and neutral ligands (N,N-dimethylaminopyridine (DMAP), tetrahydrofuran (THF)) at the mercury center (Scheme 3 and Figure 1).19

Figure 1. Structures of [2]+ (left) and 2-I (right).

Although electrostatic effects figure to be a strong component in the Sb−Au and Sb−Hg bonding in these compounds, we sought to investigate the extent of charge transfer between the antimony and metal centers. Electron localization function (ELF) studies on 118 and derivatives of [2]+ 17 suggest that the stibonium moiety in these species engages the proximal metal center via orbital-based interactions (M → σ*(Sb−CPh)), which are enhanced in the case of 2-I upon binding of I− trans to the antimony center (Figure 2). Despite computational evidence for bonding interactions between the antimony and metal centers in these compounds, the Au and Hg L3-edge X-ray absorption near-edge structure (XANES) spectra of 1 and 2-I show no discernible oxidation of their respective metal centers relative to reference compounds (Figure 3).18,19 A similar lack of significant charge transfer has been observed in the XANES spectra of metal-only Lewis pairs between zero-valent group 10 isocyanides and Tl+ cations.21 In addition to their heavy d-block centers, the tetraarylstibonium groups in compounds 1 and [2]+ also readily engage Lewis basic substrates, including fluoride anions. Coordination of the latter affords the fluorostiborane complexes [1-F]− and 2-F, respectively (Scheme 4 and Figure 4).18,19 The ability of B


Article

Accounts of Chemical Research Scheme 4. Reactions of 1 and [2]+ with F−

Figure 2. ELF plots for 1, [2]+, and 2-I. The ELF plot for 1 is reproduced from ref 18. Copyright 2011 American Chemical Society. The ELF plots for [2]+ and 2-I are reproduced with permission from ref 17. Copyright 2010 Wiley-VCH.

the stibonium moiety to coordinate additional ligands suggests that despite being in the coordination sphere of the d-block center, the antimony center retains considerable Lewis acidity. It is also noteworthy that fluoride coordination does not strongly affect the length of the Au−Sb bond (2.7694(8) Å in [1-F]−) or the Hg−Sb bond (3.0495(18) Å in 2-F), which are almost equal to those in 1 (avg. 2.7616(8) Å) and [2][PF6] (3.0601(7) Å), respectively. Taken together, these seminal studies demonstrated two important concepts: (1) incorporation of a Lewis acidic antimony center into the coordination sphere of a d-block center modifies the behavior of the d-block metal by polarizing its electron density, and (2) the antimony center itself may engage neutral and anionic Lewis bases, potentially allowing for modification of the main group−transition metal interaction.

Figure 4. Structures of 1 (left) and its fluoride adduct [1-F]− (right).

Scheme 5. Coordination of 3-R to Gold

■

TRIVALENT ANTIMONY SPECIES AS σ-ACCEPTORS In addition to oxidized heavy group 15 species acting as σacceptors toward metal centers, we have also demonstrated that Sb(III) moieties may act as Lewis acids toward late transition metal centers.22,23 Reaction of (tht)AuCl (tht = tetrahydrothiophene) with the bis(phosphinyl)stibine ligands 3-Ph and 3Cl affords complexes 4 and 5 (Scheme 5).22 The structures of 4 and 5 show that the antimony centers in both complexes adopt disphenoid rather than tetrahedral geometries, indicating that they are not strongly engaged with the gold center as Lewis

bases (Figure 5).15 Despite the overall structural similarity between 4 and 5, replacement of a phenyl ligand on antimony with chloride results in a significant distortion of the geometry of the gold center, as indicated by their respective Sb−Au−Cl1 angles of 115.09(2)° (4) and 141.73(4)° (5). Such distortions toward square-planar geometry in tetracoordinate gold(I) complexes are associated with the incorporation of a Z-type ligand in the gold coordination sphere, as seen in ClAu[(o-

Figure 3. (a) Solid-state XANES spectrum of the Au L3-edge of 1 compared to those of Au(III) and Au(I) reference compounds. Reproduced from from ref 18. Copyright 2011 American Chemical Society. (b) Solid-state XANES spectrum of the Hg L3-edge of 2-I compared to a Hg(II) reference. Reproduced with permission from ref 19. Copyright 2011 Royal Society of Chemistry. C


Article

Accounts of Chemical Research Scheme 7. Synthesis of [8]2+, [9]+, and 10

presence of cyclohexyl isocyanide to afford [8]2+ as the hexafluoroantimonate salt. Dication [8]2+ readily binds fluoride under biphasic (H2O/CH2Cl2) conditions to give fluorostiboranyl cation [9]+ as a hexafluoroantimonate salt. Finally, treatment of 7-Cl with 2 equiv of TlF in the presence of cyclohexyl isocyanide yields difluorostiborane 10. Compounds [8]2+, [9]+, and 10 constitute a series separated by formal stepwise addition of two fluoride anions to the antimony center with retention of coordination geometry at the platinum center. Inspection of the 31P NMR spectra of the series found that the 1JPt−P coupling constants increase across the series, consistent with the platinum center becoming more reduced (Table 1).31 The increase in the 1JPt−P coupling constant

Figure 5. Front and side views of the structures of (a) 4 and (b) 5.

(iPr2P)C6H4)2BPh] (∠Cl−Au−B = 168.7°).24 Natural bond orbital (NBO) analysis of 4 breaks the Sb−Au bonding into distinct lp(Sb) → 6p(Au) and lp(Au) → σ*(Sb−CPh) interactions. In contrast to the “confused” σ-donor/acceptor behavior displayed by the stibine ligand in 4, NBO analysis of 5 finds only an lp(Au) → σ*(Sb−Cl2) interaction, indicating that the chlorostibine moiety acts as a pure Z-type ligand. Similar σacceptor behavior has also been documented in related bis(phosphinyl)bismuthine complexes.25,26

Table 1. Selected Spectroscopic Data for [8]2+, [9]+, and 10

■

complex

COORDINATION-NONINNOCENCE IN ANTIMONY−TRANSITION METAL COMPLEXES As observed in the gold and mercury stibonium complexes 1 and [2]+, organoantimony(V) species may retain residual Lewis acidity despite being in the coordination sphere of a metal. In addition to Sb(V) centers, coordinated Sb(III) centers have also been shown to engage in secondary interactions with Lewis basic substrates.14,27 This phenomenon, which we term coordination-noninnocence, is typically manifested through the presence of low-lying Sb−X or Sb−C σ* orbitals. Drawing an analogy between four-coordinate stibonium species, which are well-known Lewis acids,28,29 and four-coordinate, cationic transition-metal-coordinated stibines, we questioned whether an Sb−M σ* orbital could also impart Lewis acidic properties in a direction trans to the metal site (Scheme 6). In a fundamental effort to investigate this behavior, we prepared and characterized a series of platinum complexes featuring the tris(phosphinyl)stibine ligand 6 (Scheme 7).30 Reaction of 6 with (Et2S)PtCl2 afforded compound 7-Cl, which was subsequently treated with 2 equivalents of AgSbF6 in the

2+

[8] [9]+ 10

JPt−P (Hz)

ν(CN) (cm−1)

dSb−Pt (Å)

2330 2964 2888, 3351

2214 2196 2181

2.4706(5) 2.6236(3) 2.6568(6)

1

observed across the series is not monotonic, with the largest increase occurring upon introduction of the first fluoride ligand at antimony. This difference can be rationalized by noting that the first fluoride equivalent binds trans to the Sb−Pt bond, whereas the second equivalent binds trans to an Sb−C bond. Similarly, the solid-state CyNC CN stretching frequency diminishes across the series, further suggesting an increase in electron density at the platinum center. The structures of [8]2+, [9]+, and 10 provide further insight into the perturbation of the Sb−Pt bond by anion coordination at antimony (Figure 6). Notably, the Sb−Pt separation increases dramatically upon binding of fluoride to the antimony center, indicating that fluoride binding weakens the Sb−Pt interaction (Table 1). This view is supported by NBO calculations, which show a marked increase in the polarization of the Sb−Pt bonding pair toward Pt upon fluoride anion coordination. Taken together, these experimental and computational results suggest that anion binding at a coordinationnoninnocent antimony center results in a net push of the bonding pair toward the metal center, increasing the electron density at the metal center and weakening the Sb−M interaction (Scheme 8). With anion coordination at a noninnocent stibine ligand being shown to affect the antimony−metal bond, coordinationnoninnocence may be envisioned as a means to induce

Scheme 6. Illustration of the Isolobal Relationship Between [R3SbX]+ and [(R3Sb)M]+

D


Article

Accounts of Chemical Research

Figure 6. Structures of [8]2+ (left), [9]+ (center), and 10 (right).

two-electron oxidant, we surmised the possibility of three distinct modes of oxidation (A−C), with B and C highlighting the potential redox-noninnocence of the antimony ligand (Scheme 10).

Scheme 8. Effect of Anion Binding on the Sb−M Bond

Scheme 10. Possible Two-Electron Oxidation Products of (R3Sb)[M] reactivity or otherwise alter the electronic properties of the complex. This possibility is supported by the reaction of 11-Cl, the nickel analogue of 7-Cl, with in situ-generated catecholate dianion (Scheme 9).32 In the absence of an auxiliary ligand, 11Scheme 9. Reactions of [o-C6H4O2]2− with 11-Cl

In a first attempt to access such species, we targeted a gold complex featuring the tris(phosphinyl)stibine ligand 6.34 Reaction of (tht)AuCl with 6 afforded compound 15-Cl (Scheme 11), which underwent a clean antimony-centered Scheme 11. Synthesis and Subsequent Oxidation of 15-Cl

oxidation upon treatment with PhICl2, leading to complex 16. The formation of 15-Cl was found to be reversible by treatment of 16 with excess NaI as the reductant to afford the gold iodide 15-I. The structure of 16 confirms that PhICl2 oxidizes the stibine ligand to produce a dichlorostiborane moiety in the coordination sphere of gold. Although this oxidation is antimony-centered, the gold atom shows a noticeable response in its coordination geometry, which changes from distorted tetrahedral in 15-Cl to square-planar in 16. This geometry change occurs concomitant with a shortening of the Sb−Au separation from 2.8374(4) to 2.7086(9) Å. The square-planar geometry of the gold center in 16, which is typically associated with the trivalent state, suggests that oxidation of the coordinated stibine effects a significant electronic perturbation of the gold center. As indicated by NBO calculations, conversion of 15-Cl into 16 is accompanied by an umpolung of the Au−Sb donor−acceptor interaction, which switches from Sb → Au in 15-Cl to Sb ← Au in 16 (Figure 7). Similar two-electron oxidations of gold-coordinated stibines were observed upon treatment of bis(phosphinyl)stibine−gold complexes 4 and 5 with tetrachloro-1,2-benzoquinone (o-

Cl reacts with catecholate dianion to produce the orthometalated species 13. Compound 13 is speculated to stem from nickel insertion into an Sb−C bond from a putative catecholatostiborane species (12), which features a nominal Ni0 center as a result of the polarization of the Sb−Ni bonding pair upon coordination of the catecholate dianion. Repeating the reaction in the presence of cyclohexyl isocyanide as a secondary, π-acidic ligand afforded catecholatostiborane complex 14, which is valence isoelectronic with 10.

■

REDOX CHEMISTRY OF ANTIMONY−TRANSITION METAL COMPLEXES In addition to exploring the Lewis acidity of coordinated antimony centers, we have undertaken an investigation of the redox properties of heterobimetallic antimony−transition metal complexes. While oxidations of metal−stibido compounds with halogens had been previously reported,33 oxidations of coordinated stibines were unknown prior to our efforts. Upon treatment of a transition metal−stibine complex with a E


Article


Figure 7. (a) Front and side views of the structures of 15-Cl (top) and 16 (bottom). (b) NBO plots of the principal Sb−Au interactions in 15-Cl (top) and 16 (bottom).

chloranil) to form catecholatostiborane complexes 17 and 18, respectively (Scheme 12).22 Congruent with the structural

Scheme 13. Synthesis and Subsequent Oxidation of 19

Scheme 12. Oxidation of 4 and 5 by o-Chloranil

changes associated with the conversion of 15-Cl into 16, comparison of the structures of 4 and 17 finds a significant shortening of the Au−Sb bond (2.8669(4) Å in 4 to 2.6833(3) Å in 17) as well as a change in the coordination geometry of the gold center from trigonal-pyramidal to distorted squareplanar. NBO analyses performed on 17 and 18 describe the Sb−Au bond in both cases as a natural localized molecular orbital (NLMO) with elevated orbital contributions from the gold atom (16.2% Sb/83.8% Au for 17, 16.3% Sb/83.7% Au for 18), consistent with the presence of a polar covalent Au → Sb bond. Having demonstrated that polyphosphinylstibine−gold complexes may undergo two-electron oxidation at their antimony centers, we next sought to investigate whether other transition metal−stibine complexes might display similar reactivity. With this goal in mind, we turned our attention to the synthesis of zero-valent group 10−stibine complexes, which would provide a valence isoelectronic comparison to the gold complexes previously investigated. Reaction of 6 with Ni(PPh3)4 in THF afforded the Ni0−stibine complex 19 as an orange solid (Scheme 13).32 In contrast to the reactivity of 15-Cl, treatment of 19 with PhICl2 produced the chlorostiboranyl complex 11-Cl, which results from formal addition of a Cl2 equivalent across the Sb− Ni heterobimetallic core (Figure 8). Despite the increase in the

Figure 8. (a) Structures of 19 (top) and 11-Cl (bottom). (b) NBO plots of the principal Sb−Ni interaction in 19 (top) and the Sb−Ni bond in 11-Cl (bottom).

oxidation state of the Ni−Sb heterobimetallic core, the Ni−Sb distances observed in the structures of 19 (2.4575(10) Å) and 11-Cl (2.4852(5) Å) show little variation. Although the Ni−Sb bonding in complex 19 can be described in similar terms as that F


Article


Inspection of the structures of [21]+ and [22]+ shows that in both cases the palladium center adopts a square-planar geometry. Although the 31P NMR spectrum of [22]+ contains a single resonance, its structure shows that the third phosphine arm is not coordinated in the solid state, suggesting that the coordination of the phosphine arms is fluxional in solution. Treatment of CH2Cl2 solutions of both [21][BPh4] and [22][BPh4] with tetra-n-butylammonium fluoride (TBAF) results in the formation of fluoride complexes 21-F and 22-F, as indicated by the appearance of resonances featuring coupling to phosphorus in their 19F NMR spectra (21-F: −121.61 ppm, 3 JF−P = 26.7 Hz; 22-F: −129.72 ppm, 3JF−P = 20.7 Hz). In the case of [22]+, formation of the fluoride adduct is accompanied by a rapid color change from pale yellow to deep orange (Figure 9). Comparison of the structures of 21-F and 22-F reveals that while the geometry of the palladium center in 21-F remains square-planar, the palladium center of 22-F attains a trigonalbipyramidal geometry via coordination of the third phosphine arm. It is this change in the palladium coordination geometry, along with the corresponding change in ligand-field transitions, that is responsible for the colorimetric turn-on response to fluoride anion binding. The coordination-noninnocent antimony center in [22]+ thus acts as an allosteric site, with fluoride coordination at antimony triggering coordination of the pendant third phosphine arm to the palladium center. While the insolubility of [22]+ in water limits its use in pure water, CH2Cl2 solutions layered with aqueous solutions having fluoride concentrations as low as 4 ppm display fluoride binding, as monitored by UV−vis and 31P NMR spectroscopy. In addition to utilizing coordination-noninnocent stibine complexes as anion sensors, we have also explored the use of transition metal complexes featuring Sb(V) centers in the same capacity.37 Treatment of (Et2S)2PtCl2 with bis(phosphinyl)stibine ligand 20 afforded complex 23 (Scheme 15), which is characterized by a singlet 31P NMR resonance flanked by 195Pt satellites (1JPt−P = 2706 Hz). Subsequent oxidation of 23 with

in classical stibine complexes, the presence of an antimonybound chloride in 11-Cl (Sb−Cl: 2.6835(9) Å) makes the classification of the Ni−Sb bonding more ambiguous. While NBO analysis finds an Sb → Ni donor−acceptor interaction for 19, the Ni−Sb bond in 11-Cl is described as an NLMO with distinct polarization toward the nickel center (Sb 36.3%/Ni 57.7%). Thus, while the antimony fragment in 19 can be considered to act as an L-type ligand, oxidation across the Ni− Sb core causes the antimony fragment in 11-Cl to act as an Xtype ligand, which forms a polar covalent bond with the Ni atom.

■

APPLICATION OF COORDINATION-NONINNOCENT STIBINE COMPLEXES IN ANION SENSING As part of our ongoing interest in anion sensing,29,35 we investigated the use of coordination-noninnocent ligands for sensing of aqueous fluoride. As binding of hard anions such as fluoride at coordination-noninnocent antimony centers affects the electronics of the transition metals to which they are coordinated,30 we questioned whether this property could be harnessed to provide a “turn-on” response to anion binding. Our first effort toward this goal was the investigation of cationic bis- and tris(phosphinyl)stibine−palladium species [21]+ and [22]+, prepared by complexation of (cod)PdCl2 (cod = 1,5cyclooctadiene) with the corresponding polyphosphinylstibine ligands 20, 6, and followed by subsequent chloride abstraction with NaBPh4 (Scheme 14).36 Scheme 14. Synthesis of [22][BPh4] and [22][BPh4]

Figure 9. (a) Colorimetric and geometric changes in [22]+ upon binding of fluoride to form 22-F. (b) Changes in the 31P and 19F NMR upon formation of 22-F. (c) Changes in the absorption spectrum of [22]+ upon incremental addition of TBAF. G


Article


relevance to light-driven HX splitting reactions. In this context, a number of binuclear transition metal species have been investigated because of their ability to sustain reversible twoelectron oxidation.38 As heavy main group elements readily undergo two-electron redox processes,39,40 we have investigated the possibility that their incorporation into heterobimetallic transition metal−main group complexes may provide platforms that also sustain reversible two-electron redox processes at their dinuclear cores.41,42 Complexation of (Et2S)2PtCl2 with the bis(phosphinyl)chlorostibine ligand 20-Cl leads to the formation of platinum pincer complex 26 (Scheme 16).43 Complex 26 undergoes oxidation by PhICl2 to afford the corresponding tetravalent platinum species 27. We surmised that with five electron-withdrawing chlorine atoms decorating the central Sb−Pt core, the complex may be destabilized and thus prone to photoreduction. Upon irradiation at 320 nm in the presence of 2,3-dimethyl-1,3-butadiene (DMBD) as a radical trap, 27 undergoes clean elimination of a Cl2 equivalent to regenerate the reduced species, as indicated by both UV−vis and 31P NMR spectroscopy (Figure 11). Upon optimization of the DMBD concentration, a quantum yield of 13.8% was obtained for 27, which is lower than the value of 38% obtained for a related Pt2 complex.44 Remarkably, complex 27 also evolves chlorine when irradiated in the solid state at atmospheric pressure in the absence of a trap. Evolution of chlorine or chlorine radicals was confirmed using sodium metal as a trapping agent, accounting for ∼70% of the predicted chlorine released. Having demonstrated that a change in the redox state of an antimony center held in proximity to a transition metal center affects the electronics of the heterobimetallic bond, we speculated that redox changes in antimony-containing ligands may provide the means to control the Lewis acidity and possibly the catalytic properties of the adjoining transition metal. Such a concept is illustrated by the report of a gold− borane complex featuring a [L2Au → BAr3]+ core that acts as a catalyst for enyne cyclization.13 As [L2Au]+ species are typically not active in catalysis, we questioned whether a redox-active antimony fragment may provide similar activation of an otherwise inactive [L2Au]+ fragment (Scheme 17). To this end, we chose to examine gold complexes featuring the bis(phosphinyl)chlorostibine ligand 20-Cl, which because of its electron withdrawing chloride substituent, features a Lewis acidic antimony center even in the Sb(III) state.23 Complexation of (tht)AuCl with 20-Cl followed by chloride abstraction with AgSbF6 afforded the cationic gold species [28][SbF6] (Scheme 18). To investigate whether the mildly

Scheme 15. Oxidative Synthesis of 24 and Its Reaction with F−

o-chloranil resulted in the formation of complex 24, whose diminished 1JPt−P value (2192 Hz) suggested that the platinum center had undergone oxidation. Inspection of the structure of 24 shows that oxidation of the antimony center induces a rearrangement in which the phenyl and chloride ligands are transferred from antimony to platinum. Hypothesizing that the metallastiborane fragment of complex 24 may function as a unique Lewis acid,35 we next investigated its reactivity toward fluoride anions. Treatment of 24 with an acidified aqueous solution of KF under biphasic conditions resulted in the rapid formation of complex 25, characterized by a singlet 31P NMR resonance (1JPt−P = 2855 Hz) and a 19F resonance at −77.1 ppm. The elevated 1JPt−P value of 25 relative to 24 suggested that the platinum center underwent reduction, an assignment that was borne out by the square-planar platinum coordination geometry in the structure of 25. As a result of an exchange of a platinum-bound chloride for an antimony-bound fluoride, the Sb−Pt bond in 24 is cleaved, as indicated by the large increase in Sb−Pt separation (24, 2.5906(5) Å; 25, 3.0868(11) Å) (Figure 10). This structural change can be reversed by addition of fluoride scavengers such as AlCl3 to the aqueous phase. The observed structural rearrangement is also accompanied by a change in solution color from yellow to orange, which can be used as a colorimetric turn-on response to indicate the presence of aqueous fluoride anions.

■

APPLICATIONS OF REDOX-ACTIVE ANTIMONY−TRANSITION METAL PLATFORMS Complexes that support the photoreductive elimination of halogens are attracting increasing interest because of their

Figure 10. Colorimetric and geometric changes in 24 upon undergoing F−/Cl− exchange to form 25. H


Article

Accounts of Chemical Research Scheme 16. Synthesis and Photoredox Chemistry of 27

Figure 11. (a) Absorption spectra obtained during the photolysis of 27. The inset shows a plot of quantum yield vs DMBD concentration. (b) 31P NMR spectra of the photolysis of 27.

Scheme 17. Proposed Activation of an [L2Au]+ Fragment by Oxidation at Antimony

center, we questioned whether oxidation of the antimony center might allow for stronger activation of the gold center and thus improve the catalytic performance. Treatment of 28-Cl with PhICl2 cleanly afforded 29-Cl, which was then converted into the more soluble 30-Cl with TBAF (Scheme 19). Subsequent chloride abstraction with Scheme 19. Synthesis of [30][SbF6]

Scheme 18. Synthesis of [28][SbF6]

Lewis acidic Sb(III) center present in [28][SbF6] is sufficient to activate the bis(phosphinyl)gold cation fragment for catalysis, its ability to catalyze the hydroamination of phenylacetylene with p-toluidine was tested (Table 2).45 The salt [28][SbF6] displays poor but non-negligible activity in this hydroamination reaction relative to [(Ph3P)2Au][SbF6] (no conversion), with only 2.7% conversion into the imine product after 3 h. Although the catalytic activity of [28][SbF6] demonstrates that a chlorostibine ligand may activate an otherwise inert gold

AgSbF6 afforded the cationic bisphosphinyl gold species [30][SbF6]. Comparison of the structures of [28][SbF6] and [30][SbF6] (Figure 12) reveals that oxidation of the antimony center results in a notable shortening of the Au−Sb separation from 2.9318(5) to 2.8196(4) Å, indicating the stronger Au → Sb interaction present in [30][SbF6]. In line with the strengthening of the Au−Sb interaction, the acidity of the gold center in [30][SbF6] increases, as evidenced by its ability to form adducts with adventitious water, a trait typically associated with trivalent gold species.46 In an attempt to quantify the acidities of these cations relative to typical bis(phosphine)gold(I) species, we carried out Gutmann−Beckett-type measurements on [28][SbF6], [30][SbF6], and [Au(PPh3)2][SbF6] using triphenylphosphine oxide (Ph3PO) (31P NMR (CH2Cl2): δ = 27.3 ppm) as the Lewis base. While no change in the chemical shift of Ph3PO

Table 2. Activities of Selected Gold Catalysts for Hydroamination

[Au]

reaction time

conversion (%)

[(Ph3P)2Au][SbF6] [28][SbF6] [30][SbF6]

3h 3h 40 min

0 2.7 98 I


Article


such species may also display coordination-noninnocent behavior, engaging in secondary interactions while in the coordination sphere of a metal. Additionally, we have shown that the antimony centers of these ambiphilic ligands are redoxactive, which allows electronic tuning of their metal complexes. These developments constitute a significant advance in the fundamental understanding of antimony coordination chemistry as well as the entry of antimony-containing ambiphilic ligands into the growing catalog of Z-type ligands. Having established the fundamental coordination behavior and reactivity of these tethered antimony ligands, we have begun to exploit their properties in a variety of applications. On the basis of our results, we envision that the coordinationnoninnocent properties of antimony ligands may be used to aid in the activation of transition metal−halide bonds, generating active sites for catalysis in situ. Furthermore, we anticipate that the Lewis acidity enhancement provided by positioning a highly acidic Sb(V) center in the coordination sphere of a metal may be utilized in a variety of reactions catalyzed by electrophilic metal complexes. The ability of the antimony centers in such complexes to sustain reversible two-electron redox chemistry may provide the means to modulate their catalytic activities.

Figure 12. Structures of [28][SbF6] (left) and [30][SbF6] (right).

occurs upon mixing with [Au(PPh3)2][SbF6], mixing of Ph3PO with [28][SbF6] and [30][SbF6] causes the Ph3PO resonance to shift downfield to 30.6 and 32.9 ppm, respectively. Taken collectively, these results suggest that the otherwise poorly acidic gold atoms are activated by their proximity to the Sb(III) and Sb(V) centers in [28][SbF6] and [30][SbF6], respectively. With the Sb(V) species [30][SbF6] in hand, we next investigated its catalytic activity using the same model reaction employed for [28][SbF6] (Table 2). In sharp contrast to the activity of [28][SbF6], hydroamination of phenylacetylene with p-toluidine by [30][SbF6] affords virtually complete conversion after only 40 min of reaction time at room temperature. As many transition metal-catalyzed hydroaminations require an inert atmosphere, elevated temperatures, and extended reaction times, the ability of [30][SbF6] to catalyze this reaction in air under mild conditions is noteworthy. Although [30][SbF6]catalyzed hydroaminations cannot be extended to alkylamines, the reactions proceed with a variety of arylamines as well as phenylhydrazine (Figure 13).

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Funding

This work was supported by the National Science Foundation (CHE-0952912 and CHE-1300371), the Welch Foundation (A-1423), and Texas A&M University (Arthur E. Martell Chair of Chemistry). Notes

The authors declare no competing financial interest. Biographies J. Stuart Jones was born in 1988 in Richmond, Virginia. After receiving a B.S. degree from the College of William & Mary in 2010, he began his Ph.D. studies at Texas A&M, where he works on heterobimetallic complexes featuring main group centers. ̂ François P. Gabbai ̈ received a Maitrise from the University of Bordeaux, a Ph.D. from the University of Texas at Austin under Alan Cowley and a Habilitation from TU Munich under Hubert Schmidbaur. He moved to Texas A&M in 1998, where he now holds the Arthur E. Martell Chair of Chemistry. His research deals with the chemistry of late transition metal and p-block elements.

■

REFERENCES

(1) Green, M. L. H. A new approach to the formal classification of covalent compounds of the elements. J. Organomet. Chem. 1995, 500, 127−148. (2) Shriver, D. F. Transition metal basicity. Acc. Chem. Res. 1970, 3, 231−238. (3) Parkin, G. A simple description of the bonding in transition-metal borane complexes. Organometallics 2006, 25, 4744−4747. (4) Hill, A. F. An unambiguous electron-counting notation for metallaboratranes. Organometallics 2006, 25, 4741−4743. (5) Bouhadir, G.; Amgoune, A.; Bourissou, D. Phosphine-boranes and related ambiphilic compounds: Synthesis, structure, and coordination to transition metals. Adv. Organomet. Chem. 2010, 58, 1−107. (6) Amgoune, A.; Bourissou, D. σ-acceptor, Z-type ligands for transition metals. Chem. Commun. 2011, 47, 859−871.

Figure 13. Scope of alkyne hydroaminations catalyzed by [30][SbF6]. The reaction time, conversion percentage, and (in parentheses) isolated yield are listed under each entry.

■

CONCLUSION AND OUTLOOK Our work demonstrates that the utilization of donor buttresses to enforce antimony−transition metal interactions enables the observation of unique redox and coordination chemistry of the tethered antimony ligand. We have established that tethered Sb(III) and Sb(V) centers are capable of acting as Z-type ligands toward late transition metals. The antimony centers of J


Article

Accounts of Chemical Research (7) Braunschweig, H.; Dewhurst, R. D. Transition metals as Lewis bases: “Z-type” boron ligands and metal-to-boron dative bonding. Dalton Trans. 2011, 40, 549−558. (8) Kameo, H.; Nakazawa, H. Recent developments in the coordination chemistry of multidentate ligands featuring a boron moiety. Chem. - Asian J. 2013, 8, 1720−1734. (9) Rudd, P. A.; Liu, S.; Gagliardi, L.; Young, V. G.; Lu, C. C. Metal− alane adducts with zero-valent nickel, cobalt, and iron. J. Am. Chem. Soc. 2011, 133, 20724−20727. (10) Bouhadir, G.; Bourissou, D. Complexes of ambiphilic ligands: Reactivity and catalytic applications. Chem. Soc. Rev. 2016, 45, 1065− 1079. (11) Harman, W. H.; Peters, J. C. Reversible H2 addition across a nickel−borane unit as a promising strategy for catalysis. J. Am. Chem. Soc. 2012, 134, 5080−5082. (12) Anderson, J. S.; Rittle, J.; Peters, J. C. Catalytic conversion of nitrogen to ammonia by an iron model complex. Nature 2013, 501, 84−87. (13) Inagaki, F.; Matsumoto, C.; Okada, Y.; Maruyama, N.; Mukai, C. Air-stable cationic gold(I) catalyst featuring a Z-type ligand: Promoting enyne cyclizations. Angew. Chem., Int. Ed. 2015, 54, 818− 822. (14) Benjamin, S. L.; Levason, W.; Reid, G.; Warr, R. P. Halostibines SbMeX2 and SbMe2X: Lewis acids or Lewis bases? Organometallics 2012, 31, 1025−1034. (15) Levason, W.; Reid, G. Developments in the coordination chemistry of stibine ligands. Coord. Chem. Rev. 2006, 250, 2565−2594. (16) Burton, J. W. Product class 2: Antimony compounds. Sci. Synth. 2002, 4, 53−75. (17) Lin, T.-P.; Wade, C. R.; Pérez, L. M.; Gabbaï, F. P. A mercury → antimony interaction. Angew. Chem., Int. Ed. 2010, 49, 6357−6360. (18) Wade, C. R.; Lin, T.-P.; Nelson, R. C.; Mader, E. A.; Miller, J. T.; Gabbaï, F. P. Synthesis, structure and properties of a T-shaped 14electron stiboranyl-gold complex. J. Am. Chem. Soc. 2011, 133, 8948− 8955. (19) Lin, T.-P.; Nelson, R. C.; Wu, T.; Miller, J. T.; Gabbaï, F. P. Lewis acid enhancement by juxtaposition with an onium ion: The case of a mercury stibonium complex. Chem. Sci. 2012, 3, 1128−1136. (20) Hudnall, T. W.; Chiu, C.-W.; Gabbaï, F. P. Fluoride ion recognition by chelating and cationic boranes. Acc. Chem. Res. 2009, 42, 388−397. (21) Barnett, B. R.; Moore, C. E.; Chandrasekaran, P.; Sproules, S.; Rheingold, A. L.; DeBeer, S.; Figueroa, J. S. Metal-only Lewis pairs between group 10 metals and Tl(I) or Ag(I): Insights into the electronic consequences of Z-type ligand binding. Chem. Sci. 2015, 6, 7169−7178. (22) Ke, I.-S.; Gabbaï, F. P. σ-donor/acceptor-confused ligands: The case of a chlorostibine. Inorg. Chem. 2013, 52, 7145−7151. (23) Yang, H.; Gabbaï, F. P. Activation of a hydroamination gold catalyst by oxidation of a redox-noninnocent chlorostibine Z-ligand. J. Am. Chem. Soc. 2015, 137, 13425−3432. (24) Sircoglou, M.; Bontemps, S.; Mercy, M.; Saffon, N.; Takahashi, M.; Bouhadir, G.; Maron, L.; Bourissou, D. Transition-metal complexes featuring Z-type ligands: Agreement or discrepancy between geometry and dn configuration? Angew. Chem., Int. Ed. 2007, 46, 8583−8586. (25) Lin, T.-P.; Ke, I.-S.; Gabbaï, F. P. σ-accepting properties of a chlorobismuthine ligand. Angew. Chem., Int. Ed. 2012, 51, 4985−4988. (26) Tschersich, C.; Limberg, C.; Roggan, S.; Herwig, C.; Ernsting, N.; Kovalenko, S.; Mebs, S. Gold− and platinum−bismuth donor− acceptor interactions supported by an ambiphilic PBiP pincer ligand. Angew. Chem., Int. Ed. 2012, 51, 4989−4992. (27) Benjamin, S. L.; Reid, G. Neutral organoantimony(III) and organobismuth(III) ligands as acceptors in transition metal complexes − role of substituents and co-ligands. Coord. Chem. Rev. 2015, 297− 298, 168−180. (28) Pan, B.; Gabbaï, F. P. [Sb(C6F5)4][B(C6F5)4]: An air stable, Lewis acidic stibonium salt that activates strong element-fluorine bonds. J. Am. Chem. Soc. 2014, 136, 9564−9567.

(29) Ke, I.-S.; Myahkostupov, M.; Castellano, F. N.; Gabbaï, F. P. Stibonium ions for the fluorescence turn-on sensing of F− in drinking water at parts per million concentrations. J. Am. Chem. Soc. 2012, 134, 15309−15311. (30) Jones, J. S.; Wade, C. R.; Gabbaï, F. P. Guilty on two counts: Stepwise coordination of two fluoride anions to the antimony atom of a noninnocent stibine ligand. Organometallics 2015, 34, 2647−2654. (31) Hope, E. G.; Levason, W.; Powell, N. A. Coordination chemistry of higher oxidation states. Part 21. Platinum-195 NMR studies of platinum(II) and platinum(IV) complexes of bi- and multi-dentate phosphorus, arsenic and sulphur ligands. Inorg. Chim. Acta 1986, 115, 187−192. (32) Jones, J. S.; Wade, C. R.; Gabbaï, F. P. Redox and anion exchange chemistry of a stibine nickel complex: Writing the L, X, Z ligand alphabet with a single element. Angew. Chem., Int. Ed. 2014, 53, 8876−8879. (33) Malisch, W.; Panster, P. Antimony(V)-transition metalcompounds with covalent and ionic character. Angew. Chem., Int. Ed. Engl. 1974, 13, 670−672. (34) Wade, C. R.; Gabbaï, F. P. Two-electron redox chemistry and reversible umpolung of a gold−antimony bond. Angew. Chem., Int. Ed. 2011, 50, 7369−7372. (35) Hirai, M.; Gabbaï, F. P. Lewis acidic stiborafluorenes for the fluorescence turn-on sensing of fluoride in drinking water at ppm concentrations. Chem. Sci. 2014, 5, 1886−1893. (36) Wade, C. R.; Ke, I.-S.; Gabbaï, F. P. Sensing of aqueous fluoride anions by cationic stibine−palladium complexes. Angew. Chem., Int. Ed. 2012, 51, 478−481. (37) Ke, I.-S.; Jones, J. S.; Gabbaï, F. P. Anion-controlled switching of an X ligand into a Z ligand: Coordination non-innocence of a stiboranyl ligand. Angew. Chem., Int. Ed. 2014, 53, 2633−2637. (38) Esswein, A. J.; Nocera, D. G. Hydrogen production by molecular photocatalysis. Chem. Rev. 2007, 107, 4022−4047. (39) Carrera, E. I.; McCormick, T. M.; Kapp, M. J.; Lough, A. J.; Seferos, D. S. Thermal and photoreductive elimination from the tellurium center of π-conjugated tellurophenes. Inorg. Chem. 2013, 52, 13779−13790. (40) Carrera, E. I.; Seferos, D. S. Efficient halogen photoelimination from dibromo, dichloro and difluoro tellurophenes. Dalton Trans. 2015, 44, 2092−2096. (41) Lin, T.-P.; Gabbaï, F. P. Two-electron redox chemistry at the dinuclear core of a TePt platform: Chlorine photoreductive elimination and isolation of a TeVPtI complex. J. Am. Chem. Soc. 2012, 134, 12230−12238. (42) Yang, H.; Lin, T.-P.; Gabbaï, F. P. Telluroether to telluroxide conversion in the coordination sphere of a metal: Oxidation-induced umpolung of a Te−Au bond. Organometallics 2014, 33, 4368−4373. (43) Yang, H.; Gabbaï, F. P. Solution and solid-state photoreductive elimination of chlorine by irradiation of a [PtSb]VII complex. J. Am. Chem. Soc. 2014, 136, 10866−10869. (44) Cook, T. R.; Surendranath, Y.; Nocera, D. G. Chlorine photoelimination from a diplatinum core: Circumventing the back reaction. J. Am. Chem. Soc. 2009, 131, 28−29. (45) Huang, L.; Arndt, M.; Gooßen, K.; Heydt, H.; Gooßen, L. J. Late transition metal-catalyzed hydroamination and hydroamidation. Chem. Rev. 2015, 115, 2596−2697. (46) Wu, C.-Y.; Horibe, T.; Jacobsen, C. B.; Toste, F. D. Stable gold(III) catalysts by oxidative addition of a carbon-carbon bond. Nature 2015, 517, 449−454.

K


Coordination- and Redox-Noninnocent Behavior of Ambiphilic

Recommend Documents