Phosphine and Selenoether peri-Substituted Acenaphthenes and Their Transition-Metal Complexes: Structural and NMR Investigations

A series of peri-substituted acenaphthene-based phosphine selenoether bidentate ligands Acenap(iPr2P)(SeAr) (L1–L4, Acenap = acenaphthene-5,6-diyl, Ar = Ph, mesityl, 2,4,6-trisopropylphenyl and supermesityl) were prepared. The rigid acenaphthene framework induces a forced overlap of the phosphine and selenoether lone pairs, resulting in a large magnitude of through-space 4JPSe coupling, ranging from 452 to 545 Hz. These rigid ligands L1–L4 were used to prepare a series of selected late d-block metals, mercury, and borane complexes, which were characterized, including by multinuclear NMR and single-crystal X-ray diffraction. The Lewis acidic motifs (BH3, Mo(CO)4, Ag+, PdCl2, PtCl2, and HgCl2) bridge the two donor atoms (P and Se) in all but one case in the solid-state structures. Where the bridging motif contained NMR-active nuclei (11B, 107Ag, 109Ag, 195Pt, and 199Hg), JPM and JSeM couplings are observed directly, in addition to the altered JPSe in the respective NMR spectra. The solution NMR data are correlated with single-crystal diffraction data, and in the case of mercury(II) complexes, they are also correlated with the solid-state NMR data and coupling deformation density calculations. The latter indicate that the through-space interaction dominates in free L1, while in the L1HgCl2 complex, the main coupling pathway is via the metal atom and not through the carbon framework of the acenaphthene ring system.


■ INTRODUCTION
In contrast to anionic chalcogenolates RE − , neutral chalcogenoethers RER′ have been traditionally seen as weak donors, showing a reasonable affinity to bind to soft d-block metal centers only, 1 although a small number of complexes with pblock metals and metalloids have also been reported. 2The weakly bonding nature of chalcogenoethers has been utilized in the construction of hybrid hemilabile ligands in which the soft sulfur, selenium, or tellurium donor atom can bind weakly, and reversibly, to the metal centers.Concordantly, another stronger donor atom, such as phosphorus or nitrogen, anchors the metal fragment to the hybrid ligand. 1 Hemilabile ligands have been used in catalysis, 3−5 supramolecular chemistry, 6 and sensing applications. 7A large number of hybrid ligand types with selenoether functionality have been developed (see Figure 1).These include N,Se 1,2-ferrocenyl ligands A1, 8 N,Se orthosubstituted benzyl backbone ligands A2, A3, 9 and P,Se ethylene bridge ligands, such as A4, 10 as some key examples from the literature.
In addition to the motifs above, rigid peri-substituted naphthalene or acenaphthene backbones have been used as suitable scaffolds for hybrid ligands, with the two donor atoms in the peri-positions preorganized to act as a chelating ligand forming a six-membered C 3 PME metallacycle upon coordina-tion to the metal fragment.A few series of peri-substituted naphthalene-based molecules with potential hybrid hemilabile ligand characteristics have been reported in the literature alongside their respective metal complexes in some cases.These include phosphine chalcogenoethers with P,S, P,Se, and P,Te peri-atom combinations, as shown in Figure 2.
−13 Only one of these molecules, Nap(PPh 2 )(SPh), has been utilized as a ligand.The Cu(I) complexes (B2) formed Cu−(μX) 2 −Cu bridged dimers (X = halogen), 11 while the Pt(II) and Ru(I) complexes B3 were mononuclear. 14he peri-substituted systems with phosphine and telluroether groups have been studied more extensively and include simple hybrid ligands B4 15 (Figure 2) as well as geminally dinaphthyl substituted species B5 16 and ditelluride B6. 17 One of the B4 molecules, Acenap(PiPr 2 )(TeMes), was used as a bidentate ligand toward Pt(II) and Au(I) fragments in complexes B7. 15 The dinaphthyl ligand B5 acted as bidentate or tridentate ligand in mononuclear complexes, κ 2 P,P′,κTe (B8) and κP,κTe (B9).18 Most relevantly to this paper, only a single structural report on phosphine selenoether peri-substituted species has been found in the CSD.Compound B10, Nap(PMes 2 )(SePh) (Figure 2), has been synthesized with a view of stabilizing a two-center three-electron bonding motif upon single electron oxidation of B10. 13 The arsenic analogue of B10 (with the PMes 2 group replaced by an AsMes 2 group) has also been investigated.13 Unfortunately, no 77 Se NMR parameters have been reported for B10 or its arsenic analogue.
The phenyl selenoether B11 has been synthesized and characterized, including 77 Se NMR data, and displayed a remarkable 31 P− 77 Se coupling in solution ( 4TS J PSe 391 Hz, note TS superscript indicates through-space coupling), as observed in both the 31 P (as satellites) and 77 Se NMR (as a doublet) spectra (δ P −12.9; δ Se 439.6 ppm). 12,19In contrast to the P,S and P,Te congeners, no metal complexes of any P,Se peri-  substituted species with phosphine and selenoether functionalities have been reported.
A recent in-depth study of J PP and J PSe through-space coupling used a related species, B12, as a model compound, and highlighted the recent advances in computational methods with respect to determining the relative contributions from through-space and through-bond coupling pathways. 20 comprehensive study on heteroleptic bis(phosphine) metal complexes B13 showed the changes in the throughspace J PP on coordination to the metal fragments and correlated these with the peri-region geometry as observed by single-crystal diffraction. 21Several molecular systems B14 with peri-gap Si−H•••Se interactions displayed significant magnitudes of through-space J SeH and J SeSi couplings; 22 analysis of the bonding in selected examples by computational methods indicated the presence of a chalcogen−hydride bond. 23n this paper, we report syntheses of a series of potential hemilabile P,Se peri-substituted ligands as well as a number of their complexes.Possessing a combination of two NMR-active ( 31 P (I = 1/2, 100%) and 77 Se (I = 1/2, 7.6%)) donor atoms, and in several cases also NMR-active (I = 1/2) metals, it has been hoped that useful correlations can be made between the solution and solid-state NMR data, and those from singlecrystal diffraction, to provide additional insight into the nature of the (hemilabile) bonding in P,Se hybrid ligands.

■ RESULTS AND DISCUSSION
Synthesis.Ligands L1−L4.Phosphino-selanyl acenaphthenes L1−L4 were synthesized via stepwise attachment of phosphine and selenoether functionalities to the 5,6dibromoacenaphthene starting material.There are two possible synthetic routes to achieve the target Acenap(PR 2 )(SeR) compounds, in which either the phosphino group or the selanyl group is added to the acenaphthene scaffold first, followed by addition of the other group (Scheme 1).
In the subsequent step, the diisopropylphosphino group was expected to be attached to 2Ph, 2Mes, and 2Tripp, via lithium-halogen exchange and coupling with iPr 2 PCl, to synthesize the desired acenaphthenes L1−L3 (see Scheme 1).An analogous synthetic path has been used recently by Wang to prepare Acenap(PMes 2 )(SePh) by reacting Acenap-(Br)(SePh) with nBuLi and subsequently Mes 2 PCl. 13 However, in our hands, the reactions with iPr 2 PCl gave low yields of the desired products (Scheme 1, bottom).To address this, an alternative synthetic pathway was adopted in which the order of attaching the phosphine and arylselanyl groups was reversed (Scheme 1, top).
In this pathway, the dialkylphosphino moiety was added first to synthesize 1, Acenap(PiPr 2 )(Br) (80% yield). 25The lithium-halogen exchange reaction of 1 to give the intermediate Acenap(PiPr 2 )(Li) was followed by a Se−C coupling reaction with diaryl diselenides affording L1−L4 in yields of 72−78%, making this the preferred synthetic pathway.While the reactions leading to L1−L4 were performed under an inert atmosphere due to the air-and moisture-sensitive nature of the reagents and intermediates, the workup of L1− L4 was performed in air as these compounds showed no signs of decomposition in air at ambient temperature.L2−L4 were

Inorganic Chemistry
purified by column chromatography on silica, whereas L1 was purified via recrystallization.
The Se−C coupling reactions used to prepare L1−L4 utilize diaryl diselenides and proceed with the formation of aryl(nbutyl)selane byproducts (general formula nBuSeAr); these have been separated on a chromatography column, and their identity was confirmed by 1 H, 13 C{ 1 H}, and 77 Se NMR spectroscopy for Ar = Ph and Mes* and also by single-crystal X-ray crystallography for the latter (see the SI).The tentative mechanism of nBuSeAr formation involves the reaction of 1bromobutane with arylselenoate, as shown in Scheme 2. The aryl(n-butyl)selanes are removed efficiently by washing the solid crude product with cold hexane or on the chromatography column.
Complexes with L1−L4 Ligands.Given the presence of the lone pairs on both phosphorus and selenium atoms and their arrangement in the peri-region of the acenaphthene backbone, L1−L4 were expected to act as κP,κSe bidentate ligands.However, the selenoether group is a much weaker donor; hence, complexes with L1−L4 acting as a monodentate donor (κP only) were also seen as a viable structural alternative.
We used various transition-metal precursors, such as carbonyls and halides, as starting materials to obtain the coordination complexes shown in Scheme 3. A general complexation reaction procedure involved preparation of a solution or suspension of the ligand (L1−L4) in dichloromethane (DCM) or ethanol, to which the metal-containing precursor was added as either a solid or a solution at room temperature, and the mixture was stirred overnight.Removal of the volatiles in vacuo afforded the desired complexes.Further purification (e.g., column chromatography and hexane wash) was carried out where appropriate.All novel complexes were characterized by multinuclear NMR ( 1 H, 13 C{ 1 H} DEPTQ (with the exception of L2PtCl 2 and L3HgCl 2 ), 31 P{ 1 H}, and 77 Se{ 1 H} and where applicable also by 11 B{ 1 H} and 195 Pt{ 1 H}) NMR.All but two complexes (L2PtCl 2 and L3HgCl 2 ) were also characterized by either HRMS (high-resolution mass spectrometry) or elemental microanalysis or both.All complexes prepared in this work were found to be air-and moisture-stable.The silver complexes, [(L1) 2 Ag]SbF 6 and [(L2) 2 Ag](Al(OC(CF 3 ) 3 ) 4 ), were notably light-sensitive and decomposed as solids and in solution within a matter of hours when exposed to daylight.Attempts to grow diffraction quality crystals of L4HgCl 2 gave a small amount of crystals that were shown to have composition of [L4Hg 2 Cl 4 ][L4Hg 3 Cl 6 ], i.e., the desired compound with additional weakly coordinated HgCl 2 (see crystallographic discussion below).Based on the elemental analysis results, the bulk of the material was L4HgCl 2 , with only a smaller amount of excess HgCl 2 present.
In addition to the metal complexes above, the borane adduct of L1 was also prepared.The reaction of L1 with Me 2 S•BH 3 afforded the phosphine borane L1BH 3 in a very good yield of 73% (Scheme 3).We investigated if L1 would mimic the chemistry of the similar borane adduct Acenap(PPh 2 )-(PPh 2 (BH 3 )), which in chlorinated solvents undergoes a cyclization reaction forming boronium salt. 26However, the formation of a cyclic boronium salt (such as [L1BH 2 ]Cl shown in Scheme 4) was not observed during the reaction of L1 with excess Me 2 S•BH 3 in chloroform or dichloromethane, with L1BH 3 being the sole product of the reaction.
NMR Spectroscopy.Ligands L1−L4.Many of the component elements of species synthesized in this study have NMR-active nonquadrupolar isotopes, and as such NMR provides valuable information with regards to the interactions across the peri-gap.The 31 P{ 1 H} NMR spectra of L1−L4 display singlets within a very narrow range (δ P −6.0 to −6.5 ppm).In all cases, these singlets are equipped with 77 Se satellites, from which 4TS J PSe magnitudes ranging from 452.2 to 545.0 Hz were extracted, with the magnitude increasing slightly with the electron donating ability of the aryl group bound to selenium (Table 1).The complementary doublets were observed in the 77 Se{ 1 H} NMR spectra of L1−L4 with δ Se values ranging from 283.8 to 425.3 ppm.The large magnitudes of 4TS J PSe in L1−L4 contrast strongly to the 3 J SeP of only 11 Hz observed for the nonrigid phosphine selenoethers RSe(CH 2 ) 2 PPh 2 (R = Me, Ph). 10 These observations reinforce the notion that a significant interaction between lone pairs of P and Se atoms occurs in L1−L4, leading to a large magnitude of the through-space coupling.A detailed study on coupling pathways in related phosphine selenide Nap(PPh 2 (=Se))-(PPh 2 ), B12 (see Figure 2), was published recently. 20The Scheme 2. Tentative Mechanism of the Se−C Coupling Reaction with Diaryl Diselenides a a nBuSeAr have been identified as byproducts in this reaction.

Inorganic Chemistry
magnitude of the 31 P− 77 Se coupling is much smaller in this compound (54.0Hz) compared to those found in L1−L4, however this is concomitant with differing geometry and larger P•••Se separation observed in the latter (3.41 Å vs 3.06 to 3.14 Å in L1−L4). 27omplexes with L1−L4 Ligands.The multinuclear NMR data of metal complexes reported herein are summarized in Table 2. Upon coordination of L1 or L2 to a Mo(CO) 4 moiety, the phosphorus nuclei in the resulting complexes L1Mo(CO) 4 and L2Mo(CO) 4 are significantly deshielded (Δδ P 47.3 ppm for L1Mo(CO) 4 ) and the selenium nuclei become more shielded (Δδ Se −31.2 ppm for L1Mo(CO) 4 ) vs the free ligand.There is also a dramatic reduction in the magnitude of J PSe from 452.2 to 14.8 Hz in L1Mo(CO) 4 and An even more dramatic change in the magnitude of J PSe takes place on coordination to platinum(II) or palladium(II) centers.In L1PtCl 2 and L2PdCl 2 , no 77 Se satellites are observed in the 31 P{ 1 H} NMR spectra, indicating the J PSe magnitude of less than ca. 2 Hz.This is corroborated by Se{ 1 H} NMR spectra, which show singlets, and thus, there is no observable coupling to 31 P. Small but detectable 2 J PSe couplings (6.5 and 24.5 Hz) were observed in both 31 P{ 1 H} and 77 Se{ 1 H} spectra of L2PtCl 2 and L4PdCl 2 .
Coordination of platinum(II) or palladium(II) to L1, L2, or L4 results in a shift of δ P to high frequency (Δδ P up to 44.6 ppm), while δ Se is shifted to low frequency for the Pt(II) complexes L1PtCl 2 and L2PtCl 2 (Δδ Se −93.3 ppm for the former) and to higher frequencies for the Pd(II) complexes L2PdCl 2 and L4PdCl 2 (Δδ Se 33.2 ppm for the latter).
Using L1PtCl 2 as an example, both 31 P{ 1 H} and 77 Se{ 1 H} spectra display well-resolved satellite peaks for the 195 Pt isotopologue ( 195 Pt; I = 1/2, 34%) with 1 J PPt = 3528.5Hz and 1 J SePt = 656.6Hz (Figure 3).Both couplings are complemented in the 195 Pt{ 1 H} NMR spectrum, which displays a doublet with 77 Se satellites at δ Pt −4190.0ppm.These are reliably similar to the coupling constants observed for the (structurally verified) Pt chelate complex 3 ( 1 J PPt = 3580 Hz and 1 J SePt = 588 Hz, see Figure 4). 28ddition of one equivalent of HgCl 2 to L1 results in the formation of the complex of composition L1HgCl 2 .The 2 J PSe coupling was obtained from the solution 31 P{ 1 H} NMR spectrum, which shows 77 Se satellites of the singlet at δ P 54.0 ppm with J PSe of 86.7 Hz and the complementary doublet at δ Se 378.3 ppm in the 77 Se{ 1 H} NMR spectrum (Figure 5).In addition, the 199 Hg satellites were observed in the 31 P{ 1 H} NMR spectrum showing a remarkably large 1 J PHg of 6,611 Hz ( 199 Hg, I = 1/2, 16.9%).The 1 J PHg coupling observed in L1HgCl 2 vastly exceeds the 1 J PHg magnitudes of 3655 and 2337 Hz observed in the related HgCl 2 complex 4 (Figure 4). 21he 77 Se{ 1 H} NMR spectrum of L1HgCl 2 (solution in CDCl 3 ) also shows coupling of 77 Se to 199 Hg, with wellresolved satellites (J SeHg 721.3 Hz) (Figure 5).
To gain further insight into the nature of the Hg•••Se interaction, a 77 Se{ 1 H} SS-MAS NMR spectrum of a sample of the dimeric complex (L1HgCl 2 ) 2 was acquired (Figure 6).The magnitude of J SeHg obtained (785 Hz) showed only a marginal increase when compared with the magnitude observed for the solution of L1HgCl 2 in d-chloroform (721.3Hz).The similarity of the two J SeHg magnitudes indicates that the mercury is bound relatively loosely to the selenium atom in chloroform solution, i.e., the bonding in the solution is similar to that observed in the crystal of the dimer (L1HgCl 2 ) 2 .However, in both cases, a significant overlap of the Se and Hg orbitals still takes place to give rise to the observed high magnitudes of J SeHg .Not many examples of J SeHg couplings have been reported in the literature, 31 particularly those involving selenoethers.Those published span a large range of magnitudes.For example, a large magnitude of 1 J SeHg was reported in the Zintl anion [HgSe 2 ] 2− (2258 Hz), 32 while very weak bonding in a Hg(CN) 2 complex of a crown selenoether (with Hg•••Se distances 3.38−3.44Å obtained from singlecrystal X-ray diffraction) resulted in much smaller J SeHg 110 and 123 Hz being observed in the 77 Se CP-MAS SS NMR spectra. 33A magnitude particularly similar to that observed by us was recorded in the phosphine selenide complex Cl 2 Hg((Se =)PnBu 3 ) 2 ( 1 J SeHg 751 ± 10 Hz); however, no structural data were reported for this complex. 34he solution state 31 P{ 1 H} and 77 Se{ 1 H} NMR spectra of L2HgCl 2 show that increasing the steric bulk and the electron donating ability of the aryl group attached to the selenium atom results in an increase in the magnitudes of J PSe (186.9Hz in L2HgCl 2 cf.86.7 Hz in L1HgCl 2 ) and J SeHg (909 Hz in L2HgCl 2 cf.721 Hz in L1HgCl 2 ).The 77 Se{ 1 H} SS-MAS NMR spectrum of L2HgCl 2 corresponds well with the solution state spectra, with only a small increase of the magnitude of J SeHg in the solid state to ca. 1040 Hz (cf.909 Hz in CDCl 3 ), indicating that similar Hg−Se interactions exist in both solution and solid-state environments.A Hg−Se distance of 2.9132(5) Å was measured crystallographically in L2HgCl 2 (see the structural discussion below).The bulkier L3HgCl 2 and L4HgCl 2 display J PSe values comparable to that seen in L2HgCl 2 (Table 2).Unfortunately, a high signal-to-noise ratio in the 77 Se{ 1 H} NMR spectra of L3HgCl 2 and L4HgCl 2 precluded the observation of 199 Hg satellites and hence determination of the J SeHg for these complexes.
As indicated above, the J PSe coupling in L1 is diminished tremendously upon complexation to Hg (from 452 Hz in L1 to 87 Hz in L1HgCl 2 ).To gain additional understanding for this change, we performed density functional theory (DFT) calculations of these couplings and analyzed them with the coupling deformation density (CDD) approach. 35The heavymetal complex calculations were performed at a suitable relativistic level (unrestricted 4-component Dirac−Kohn− Sham level, see the SI for details and references).The    magnitudes being overestimated and others underestimated (see Table S3 in the SI).However, from the visualization of the corresponding CDD coupling paths (Figure 7), it is clear that in free L1, the coupling pathway is mainly through-space (note the large turquoise area between P and Se atoms in Figure 7a) and to a lesser extent along the P−C−C−C−Se bonds of the acenaphthene scaffold (as indicated by the much smaller contributions on these C atoms in Figure 7a).This finding is reminiscent of couplings involving heteroatoms that are formally nonbonded but forced in close proximity.This constraint imposed by the acenaphthene backbone can cause overlap of the lone pairs of the peri-atoms, leading to J couplings approaching or even exceeding 1 J couplings between the same nuclei when they are covalently bound (see, for example, J TeTe in Buḧl et al. 36 ).In contrast to the throughspace J SeP coupling in the free ligand, in the L1HgCl 2 complex, this coupling is propagated predominantly through the P−Hg (optimized bond length 2.525 Å) and Hg−Se bonds (optimized bond length 2.878 Å).This is shown as the major contributions on the P, Hg, and Se atoms in Figure 7b.There is rather little direct through-space contribution (optimized P•••Se distance 3.672 Å) and negligible propagation along the P−C−C−C−Se framework.
In summary, the NMR data indicate a significant interaction between the mercury and the selenoether moiety in these complexes; however, combining the single-crystal structural  data with the solution and solid-state NMR highlights that the Hg−Se interaction is rather flexible.This applies, in particular, when the Hg−Se interaction is compared to the Hg−P interaction, which appears to be much more insensitive to the local environment.
Silver has two naturally occurring isotopes, both of which have nuclear spin 1/2 with very similar natural abundancies, 107 Ag (51.84%) and 109 Ag (48.16%).Considering this and the presence of other NMR-active nuclei ( 31 P and 77 Se) in our complexes, the 31 P{ 1 H} and 77 Se{ 1 H} NMR spectra of [L1 2 Ag]SbF 6 were expected to be rather complex.The 31 P{ 1 H} NMR spectrum shows two major doublets stemming from 107 Ag and 109 Ag isotopomers with NMR-inactive Se atoms.These doublets are both centered at δ P 26.2 ppm, with 1 J 31P−107Ag of 426.2 Hz and 1 J 31P−109Ag of 494.0 Hz (Figure 8).These J PAg magnitudes are significantly larger than those observed in the bis(phosphine) analogue 5 (see Figure 4, 1 J PAg couplings iPr 2 P−Ag 249.1 and 287.5 Hz and Ph 2 P−Ag 172.9 and 200.6 Hz, respectively). 21or isotopologues with an NMR-active selenium atom ( 77 Se, I = 1/2, 7.6% abundance), the spin system becomes dramatically more complicated, and a corresponding complex satellite pattern is observed in the 31 P{ 1 H} NMR spectrum (Figure 8).This is located at the heel of the major doublets, with some parts of the signals obscured by the major doublets.By carrying out a simulation of these spin systems, we were able to reproduce the experimental 31 P{ 1 H} and 77 Se{ 1 H} NMR spectra accurately as shown in Figures 8−10, with the latter being a complex multiplet centered at δ Se 369.1 ppm.The spin simulations were also performed for the [L2 2 Ag]-[Al(OC(CF 3 ) 3 ) 4 ] complex (see Figure S6 and Figure S7 in the SI).
The approximate magnitudes of the 1 J SeAg couplings of 43 and 38 Hz were obtained from the spin simulations of the 77 Se{ 1 H} NMR spectra of the two isotopomers of [L1 2 Ag]-SbF 6 with 109 Ag and 107 Ag atoms.These magnitudes are seemingly rather small; however, a literature search indicated that the observation of J SeAg is rather unusual and seldom reported, as generally selenoether silver complexes give no observable couplings in the 77 Se{ 1 H} NMR spectra (with singlet signals only) even at low temperatures.The lack of observable one-bond Ag−Se couplings has been attributed to fast reversible ligand dissociation due to the extreme lability of the selenoether complexes.This was recorded for both aryl and alkyl selenoethers, such as in the complex [Ag-(PhSeCH 2 CH 2 SePh) 2 ]BF 4 37 and others. 38,39Observation of    107 Ag and 109 Ag centers and one 77 Se (NMR-active) atom; these form the complex satellite pattern.Two additional major isotopomers (not shown) with both Se atoms being NMR-inactive contribute to the major doublets seen in the 31 P{ 1 H} NMR spectra.The acenaphthene scaffold was simplified for clarity.Coupling constants are given in Hz.

Inorganic Chemistry
412.6 ppm with a smaller, yet broadened, line width of 56 Hz.The 11 B{ 1 H} NMR spectrum shows a broad doublet at δ B −41.7 ppm, which allows for determination of approximate 1 J BP coupling of 35 Hz.Broadening of the signals precluded measurement of J PSe and J SeB couplings.
Structural Investigations.All new compounds reported in this article (shown in Scheme 3), including ligands L1−L4 but excluding complex L3HgCl 2 were subjected to singlecrystal X-ray diffraction studies.Selected crystallographic information is presented in Table 3, Figures 11−13, with additional information in the SI.In addition, diffraction data were also collected and solved for the intermediates 2Mes and 2Tripp and the side product Mes*SeBu.The data for the latter three compounds are listed in the SI but are not discussed in the main text.
The sum of the van der Waals radii of P and Se (∑r vdW 3.85 Å) is much larger than the ideal peri-distance of ca.2.5 Å. 30,40 Despite incorporation of a large phosphine and a selenoether group in the peri-positions, the crystal structures of L1 to L4 (Figure 11) show only moderate in-plane and out-of-plane distortions.Interestingly, L4, bearing the bulkiest aryl group (Mes*), shows only slightly more pronounced out-of-plane distortions compared to the other three ligands (Table 3).The P•••Se distances in L1−L4 are rather similar in each of the ligands and range from 3.055(1) to 3.135(1) Å, i.e., 79 to 81% of ∑r vdW .On the other hand, these distances are significantly longer than ∑r covalent , for P and Se which is 2.27(7) Å. 29 Due to the constrained geometry and mutual orientation of the two substituents as seen in Figure 11, a significant degree of overlap between the phosphorus and the selenium lone pairs is expected, and this is confirmed by large magnitudes of the observed 4TS J PSe , as mentioned in the NMR Spectroscopy section.A quasi-linear P•••Se−C Aryl arrangement with angles ranging from 160.8(1) to 170.0(1)°is present in all four ligands.This indicates the presence of a weak attractive chalcogen bond-like n(P) → σ*(Se−C Ar ) orbital interaction.
The structure of L1BH 3 (Figure 11 and Table 3) shows that the phosphine borane adduct is formed, with no bonding interaction between the boron and selenium atoms (interatomic separation of 3.256(8) Å). 41 The splay angle of 18(1)°i ndicates a moderate amount of strain; in addition, there are significant out-of-plane displacements of the peri-atoms (ca.0.57 and 0.70 Å) from the mean acenaphthene plane, and the P−C•••C−Se torsion angle is 31.1(3)°.These suggest that the observed placement of the borane group within the peri-gap is due to minimized steric repulsion in such a configuration rather than an attractive interaction between selenium and boron atoms.
The crystal structures of the complexes L2Mo(CO) 4 , L2PdCl 2 , L2PtCl 2 , and [(L2) 2 Ag][Al(OC(CF 3 ) 3 ) 4 ] are shown in Figure 12, and key details are given in Table 3.The structures of the related metal complexes with L1 and L4 ligands (L1Mo(CO) 4 , L4PdCl 2 , L1PtCl 2 , and [(L1) 2 Ag]-SbF 6 ) are rather similar.Their key details are displayed in Table 3, and the relevant figures are available in the SI (Figure S10).These complexes display κP,κSe coordination of the phosphino-selenoether ligands to the metals.The P−M distances in all reported complexes indicate tight bonding of the phosphine group to metals, with very little change of the bond lengths as the aryl group bulk on the selenium is varied.The geometries on the metal atoms in this study span (distorted) tetrahedral, square planar, and octahedral.
Both L1HgCl 2 and L2HgCl 2 adopt a monomeric structure with the HgCl 2 motif spanning the two donor atoms of the P,Se-ligands (Figure 13).Attempts to crystallize L4HgCl 2 gave a small amount of crystals that were shown to consist of the desired complex, with additional weakly coordinated HgCl 2 .The coordination geometry is similar to that found in L1HgCl 2 , with the HgCl 2 motif bridging P and Se periatoms in L4; the additional weakly coordinated HgCl 2 molecules form further weakly bridging interactions through the chlorine atoms, forming a structure of [L4Hg 2 Cl 4 ]-[L4Hg 3 Cl 6 ] (Figure 13      The varying Hg−Se distances observed within the Hg complexes were partly discussed in the NMR discussion above in connection with the observed J SeHg couplings.In the complexes L1HgCl 2 , L2HgCl 2 , and [L4Hg 2 Cl 4 ][L4Hg 3 Cl 6 ], the Hg−Se distances range from 2.8083(2) to 2.9132(5) Å, while in the complex [L1HgCl 2 ] 2 •2CHCl 3 , the Hg•••Se distance is elongated to 3.2100(4) Å, indicating a significantly weaker interaction.
A comprehensive literature search revealed that flexibility within the Hg−Se distances is a natural feature in selenoether mercury(II) complexes.Selected literature examples also point to a limited correlation between the coordination number of the Hg atom and the Hg−Se bond length.Thus, the Hg•••Se distance in the five-coordinated complex 9 (2.535(1)Å) 46 is much contracted compared to that in the four-coordinate complex 10 (3.058 Å) (see Figure 14). 47The weak bonding in the latter example is presumably (at least partially) a result of steric constraints imposed by the specific geometry of this particular bis(selenoether) ligand.Even longer (albeit still subvan der Waals) Hg•••Se distances were found in 11 (3.351(4)− 3.777(4) Å), where o-phenylene mercury and tetraselenafulvalene components are both planar and form cofacial stacks, although the authors consider these as Hg•••Se "contacts" rather than bonds. 48No NMR data have been reported for 11 unfortunately to allow comparison of its J SeHg couplings with that in [L1HgCl 2 ] 2 •2CHCl 3 .
Considering the literature precedents as well as our structural and spectroscopic data, it seems appropriate to view the [L1HgCl 2 ] 2 •2CHCl 3 complex as κP-bound and to consider the long Hg•••Se contact as a weak secondary interaction rather than a standard coordination bond.It is interesting to note that despite this elongated Hg•••Se distance, the (ligand-forced) proximity of the Hg and Se atom results in large magnitudes of J SeHg in [L1HgCl 2 ] 2 •2CHCl 3 , as observed in both solid-state NMR (785 Hz) and solution (721 Hz).The ease with which both the monomeric L1HgCl 2 and dimeric [L1HgCl 2 ] 2 •2CHCl 3 forms were obtained indicates a close to equilibrium process.This somewhat resembles the halideinduced ligand conversion observed in complex A4 (Figure 1), where addition/removal of chloride led to coordination/ decoordination of the selenoether donor atom to the Pt(II) center. 10It appears our situation has a lower barrier as simple change of the solvent of crystallization induces the change.

■ CONCLUSIONS
A series of phosphino and selenoether peri-substituted species L1−L4 with varying bulk of the aryl substituent on selenium atom were synthesized.Coordination properties of these species were investigated in reactions with metal motifs Mo(0), Pt(II), Pd(II), Hg(II), and Ag(I) as well as BH 3 .In all but one case, the ligands coordinated in a κP,κSe bidentate manner.The exception was the mercury complex [L1HgCl 2 ] 2 , which shows a monodentate κP coordination in the solid state with a rather long Se•••Hg contact also present.Notwithstanding this, a notable J SeHg (≥700 Hz) was observed in the solution and solid-state 77 Se NMR spectra for this compound.
To provide further insight, we correlated NMR (solution and in some cases solid state) data with structural data as far as possible for this series of complexes.
The peri-substitution geometry with the two preorganized donor atoms (P and Se) appears to contribute to the overall fair stability of the studied complexes; all of these are air-stable and, apart from one of the mercury complexes, show no signs of (coordination/decoordination) fluxional behavior in the solution NMR spectra or variability of coordination modes in the solid state (as judged by single-crystal diffraction).
Preference for the formation of six-membered chelate rings (i.e κP,κSe coordination) is seen also in other phosphinechalcoether peri-substituted ligands, namely, in the Cu I , Pt II , and Ru II complexes of a P,S ligand (complexes B2, Figure 2) 11,14 and in the Pt II complex of a P,Te ligand, B7. 15 On the other hand, the AuBr complex B7 showed very elongated Se••• Au interaction, consistent with very weak bonding (i.e., κP coordination only). 15It appears that the P,Se and P,Te ligands are therefore weaker-binding; however, because of a relatively small number of known metal complexes in each of the P,S, P,Se and P,Te ligand series, we hesitate to postulate a clear pattern in these chalcogeno-phosphine complexes.
Large magnitudes of through-space 31 P− 77 Se couplings were observed in series L1−L4 (452−545 Hz).This is due to the forced overlap of the lone pairs in the peri-region.A similar effect is observed in P,Te systems B4, with concomitant 31 P− 125 Te couplings in a range of 1213 to 1357 Hz. 15 Lack of any NMR-active isotope of sulfur precludes the observation of P−S couplings in the P,S species.
Quasi-linear P•••Se−C ipso arrangement with the angles ranging from 160.8(1) to 170.0(1)°is present in all four ligands L1−L4, and such quasi-linear P•••Te−C ipso geometry was also observed for all tellurium ligands B4. 15 This is concomitant with a dative n(P) → σ*(Ch−C Ar ) orbital interaction (i.e., intramolecular chalcogen bond), which, in turn, is believed to contribute to aligning the relevant orbitals to allow for significant through-space couplings mentioned above.By the way of contrast, great variety of geometries with respect to the orientation of the organyl group on the sulfur atom was observed in P,S ligands, with the P•••S−C ipso angles ranging from 84 to 165°. 12,13ASSOCIATED CONTENT

Figure 4 .
Figure 4. Literature complexes related to the species reported in this paper.

Figure 10 .
Figure 10.Coupling pathways for the two isotopomers of the [L1 2 Ag]SbF 6 complex with107  Ag and109  Ag centers and one 77 Se (NMR-active) atom; these form the complex satellite pattern.Two additional major isotopomers (not shown) with both Se atoms being NMR-inactive contribute to the major doublets seen in the 31 P{ 1 H} NMR spectra.The acenaphthene scaffold was simplified for clarity.Coupling constants are given in Hz.
bottom right).Crystallization of L1HgCl 2 from a different solvent (CHCl 3 ) afforded crystals which were shown to adopt a dimeric structure [L1HgCl 2 ] 2 • Table 3. Selected Bond Distances (Ångstroms, Å) and Angles (Degrees, °) for the Ligands and Metal Complexes Compound L1 [DFT] a with the chloride ligands forming Hg−(μCl) 2 −Hg bridges (Figure 13 top left).Coordination to Hg(II) results in an increase of the P•••Se distances in all four complexes (Δ = 0.261−0.707Å); in the structurally divergent [L1HgCl 2 ] 2 •2CHCl 3 , the increase was moderate at 0.368 Å.The P−Hg distances in all four complexes are consistent with a strongly coordinated phosphine group.

Figure 11 .
Figure 11.From left to right and top to bottom: molecular structures of L1, L1BH 3 , L2, L3, and L4.Carbon-bound hydrogen atoms are omitted for clarity.Anisotropic displacement ellipsoids are plotted at the 50% probability level.

Figure 13 .
Figure 13.From left to right and top to bottom: molecular structures of [L1HgCl 2 ] 2 , L1HgCl 2 , L2HgCl 2 , and L4HgCl 2 •1.5HgCl 2 .Solvating molecules and hydrogen atoms are omitted for clarity.Anisotropic displacement ellipsoids are plotted at the 50% probability level, and peripheral groups are drawn as sticks only.

Figure 14 .
Figure 14.Literature complexes are mentioned in the discussion.Note that 8 is a chain polymer with bridging triflate anions, with each Ag atom coordinated by P, Se, and two O atoms.In 11, the shortest Hg•••Se contact is indicated by a dashed line.

Table 1 .
Selected NMR Data for Free Ligands L1−L4

Table 2 .
NMR Parameters of the Metal Complexes Reported in This Paper a Values for109Ag and107Ag isotopomers.f Low signal-to-noise ratio in 77 Se{ 1 H} NMR spectrum precluded observation of 199 Hg satellites.
a Both solution and solid-state (CP-MAS SS) NMR data are included.b2 J PP 58.2 Hz. c2 J PP 73.5 Hz. d Values for the two P atoms in the complex.e g Signal broadening precluded reading of J couplings.
SeAg in both [L12Ag]SbF 6 and [L22Ag][Al(OC(CF 3 ) 3 ) 4 ] indicates increased chelate stability of the peri-P,Se geometry ligands compared to other more flexible geometries.The 31 P{ 1 H} NMR spectrum of L1BH 3 shows a broad singlet at δ P 46.4 ppm with a line width of ca.230 Hz, while the 77 Se{ 1 H} NMR spectrum shows a broad singlet at δ Se 1 J