This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.
Article Cite This: ACS Omega 2018, 3, 608−614
Novel Bonding Mode in Phosphine Haloboranes Qammar L. Almas and Jason K. Pearson* Department of Chemistry, University of Prince Edward Island, 550 University Avenue, Charlottetown, Prince Edward Island C1A 4P3, Canada S Supporting Information *
ABSTRACT: We have predicted, using a wide range of theoretical models, the potential energy surfaces of dative bond stretching in some phosphine haloboranes and closely associated analogues. It is shown that these dative complexes demonstrate unusual bond stretching potentials that are characterized by having multiple inflection points and are not able to be fit to any traditional Morse or Lennard-Jones-type curve. Specifically, in the case of Cl3B−PH3, this effect is so pronounced that the surface actually exhibits two distinct minima. To the best of our knowledge, this is the first example of such a unique bonding phenomenon reported for these species and is explained by the competition between the energetic cost of the required pyramidalization of the Lewis acid to form a dative bond and the stabilization from the favorable attraction between the Lewis acid and base. When the cost of pyramidalization of the Lewis acid is high relative to the strength of the interaction between the acid and base, the potential well associated with dative bonding is significantly weakened and the result is a relatively flat potential energy surface that is susceptible to the unusual characteristics described herein.
■
TZVPP34−36 and CC2/TZVPP35−37 levels of theory produce the correct D3h geometries.30 Nevertheless, there have still been several valid attempts at rationalizing B−P bonding from a theoretical perspective.26 In a recent paper, other seemingly spurious computational results in a test set of nonmetallic Lewis acid−base pairs were reported.38 On surveying the potential energy surfaces (PESs) associated with the dative bond stretching of a series of small Lewis acid−base pairs, it was found that fluorinated phosphoboranes were unique in that they exhibited a shoulder at a bond length shorter that the equilibrium bond length (see Figure 1). The PES of F3B−PH3 predicted at the reasonably reliable M06L/cc-pVQZ density functional level is shown in Figure 1, where the inflection “shoulder” is observed at r ≈ 2.3 Å (illustrated with the blue dot) and the global minimum is at r ≈ 3.1 Å. This is to be contrasted with the PESs of H3B−PH3 and F3B−NH3 predicted at the same level of theory, whose curves exhibit the usual Morse-like PES behavior for bond stretching with their minima at r ≈ 1.9 and 1.7 Å, respectively. The inset plot shows the magnified PES of F3B−PH3 and highlights the unusual behavior of the PES at r < req. This anomaly is not observed for either H3B−PH3 or F3B−NH3, suggesting that this phenomenon cannot be attributed directly to the trifluoroborane nor to the phosphine alone, but rather is the result of the combined effects of each, perhaps in combination with the inaccuracies of the theoretical model.
INTRODUCTION The elucidation of the properties of boron- and phosphorousbased adducts has been of interest to chemists for many decades.1−9 These compounds play many important roles in synthetic chemistry and are high-potential candidate materials for hydrogen storage.10−14 Furthermore, B−P compounds are particularly known as efficient refractory semiconductor materials and thus are also of interest in solid-state physics13 and have shown potential as reducing agents in biological systems.11,12 The importance of B−P complexes in a wide range of applications makes it critical to fully understand their structure and stability, which has been attempted both theoretically and experimentally dating back as far as the 1940s.1−8,15−26 Understanding the bonding in these dative complexes has been the subject of much research, and many differing explanations have been offered for the trends in the bond strengths of such Lewis acids and bases. From a theoretical perspective however, understanding dative complexes can be a challenging task as it is well known that many computational methods (density functional theory, or DFT, in particular) do not describe the properties of dative bonded complexes well.27−29 Specifically with respect to B−P dative complexes, there have been several theoretical studies employing a variety of techniques that have revealed problematic or contradictory results.4,5,15,30−32 Janesko, for example, studied substitutions on boron of B−P compounds and found that the popular B3LYP33 functional cannot reproduce accurate results for bond dissociation energies.5 It has also been reported that DFT methods produce C1 geometries of B−P−benzene and B−P− coronene molecular systems, whereas ab initio MP2/ © 2018 American Chemical Society
Received: October 10, 2017 Accepted: January 8, 2018 Published: January 19, 2018 608
DOI: 10.1021/acsomega.7b01529 ACS Omega 2018, 3, 608−614
Article
ACS Omega
MP2/6-31G*,42,43 CR-CCL/aug-cc-pVDZ//MP2/6-31G*, CR-CCL/aug-cc-pVTZ//MP2/6-31G*, and an extrapolated “complete basis set” (CBS) approach for each of the coupled cluster schemes44−48 obtained from the following extrapolation formula49 E DT =
Ecorr D
(E Dcorr D3 − E TcorrT 3) D3 − T 3
(1)
Ecorr T
where and represent correlation energies predicted at the double-ζ and triple-ζ levels, respectively, D = 2, and T = 3. EDT is added to the triple-ζ energy and thus the total benchmark energy of the system (for the CR-CCL case, for instance) is given as Figure 1. Comparison of PES scans of F3B−PH3, F3B−NH3, and H3B−PH3 along the dative bond calculated at the density functional M06L/aug-cc-pVTZ level of theory. The blue dot highlights an inflection point at r < req.
ECR ‐ CCL/CBS(D,T) = ECR ‐ CCL/aug ‐ cc ‐ pVTZ + E DT
(2)
Complete active space (CAS) 50 calculations were also performed to further probe the dependence of PES on the effects of static correlation for several selected complexes herein. CAS calculations correct for static correlation errors arising from an inadequate description of the possible configuration state functions accessible to a chemical system. In CAS, all possible Slater determinants of a selected set of active occupied and virtual orbitals are taken into account and full configuration interaction is calculated on that so-called “active space”. CASSCF(8,8)/cc-pVTZ calculations were performed after preliminary calculations of the density matrices from a variety of sets of active spaces (specifically, the orbital occupation numbers), where the (8,8) designation signifies that 8 electrons were allowed to be permuted among 8 virtual molecular orbitals. The selection of the active orbital space was decided based on several observations from these preliminary calculations: (i) occupancies of orbitals outside of the (8,8) active space did not change when larger spaces were tested, (ii) the trend of the PESs obtained was qualitatively the same regardless of the selected active space and number of basis functions in the level of theory that was employed, and (iii) the usual T1 diagnostic51 used to assess multireference character in quantum systems from our coupled cluster results was consistently well below the traditional threshold of 0.02 for single reference acceptability. Specifically, we note that CASSCF(6,6) and CASSCF(8,8) produce qualitatively identical results (see Supporting Information). Additionally, the (8,8) active space allows for the favorable pairing of occupied and virtual orbitals in the active space, as suggested by Roos et al.52 In cases when the curvature of the PES was especially flat, we have employed fine integration grids (for DFT calculations) and stringent convergence criteria to ensure the reliability of the results. Additionally, relevant minima were confirmed by performing the standard frequency analysis. Consequently, we are confident that we have thoroughly explored the space of available high-accuracy computational models suitable for this application. All DFT calculations were performed using the QChem53 software package. All CR-CCL calculations were performed using the GAMESS48,54 suite of programs, and all MP2, CCSD(T), and CAS calculations were performed with the Gaussian 09 software.55 Furthermore, the PESs of the relevant compounds of interest have been decomposed into their Lewis acid and base components by a systematic fragmentation approach. In this approach, we calculate the contribution to the total PES due to the structural rearrangement of the Lewis acid and base from
Despite what is a rather long history of theoretical and experimental investigation of Lewis acidity and basicity in a wide range of contexts,23 this unusual behavior of the F3B−PH3 dative bond PES is quite rare and has only been observed in one other similar case that we are aware of.39,40 Is this behavior an artifact of the calculations or a bona fide chemical phenomenon that has yet to be fully explored? Given the importance of these species, we endeavor to further explore the nature of the bonding within these and other halogenated phosphoboranes by employing a wide range of computational techniques, including several density functional and highly accurate ab initio wave function models, comparison with the PESs of analogous complexes, molecular orbital (MO) analysis, and energy decomposition by systematic fragmentation of the Lewis acid and base components. Our aim is to reveal and characterize some rather remarkable chemical behaviors. Atomic units are used throughout unless otherwise indicated.
■
COMPUTATIONAL METHODS To probe the nature of the PES of dative bond stretching of halogenated phosphoboranes, a relaxed scan of the boron− phosphorous bond was performed for several species (vide infra) with a wide variety of electronic structure techniques. The presence of a shoulder in the PES at r < req was confirmed by the analysis of the second derivative of the PES curve, characterizing the shoulder inflection as a point where the second derivative vanishes. For all optimizations, the Lewis acid−base bond was constrained at 1 of 31 bond distances from 1.0 to 4.0 Å (in steps of 0.1 Å), whereas all other geometrical parameters were allowed to fully relax. For DFT calculations, shown herein are results from the M06L functional41 with the aug-cc-VQZ basis (unless otherwise indicated) owing to the fact that M06L is parameterized for inorganic species and can be quickly applied to many systems. However, we have also coupled this functional to a much wider series of standard basis sets (M06L/6-31G*, M06L/6-31+G**, M06L/6-311+G*, M06L/ cc-pVTZ, M06L/aug-cc-pVTZ) to investigate the basis set dependence of the PES. Additionally, a much larger swath of functionals has been probed, all of which exhibit qualitatively identical results (see Supporting Information). Selected PESs were also predicted by ab initio wave function methods, including MP2/6-31G*, MP2/6-311G*, MP2(Full)/cc-pVTZ, CCSD(T)/cc-pVDZ//MP2/6-31G*, CCSD(T)/cc-pVTZ// 609
DOI: 10.1021/acsomega.7b01529 ACS Omega 2018, 3, 608−614
Article
ACS Omega
to produce erratic behavior in some cases and can be highly sensitive to the choice of the basis set.56 In terms of the PES behavior with respect to basis set size, we have employed up to and including the rather large augmented quadruple-ζ basis set of Dunning,57,58 aug-cc-pVQZ, with density functional models and up to analogous triple-ζ basis sets with highly correlated wave function theories. It is tempting to conjecture that the shoulder is a basis set artifact due to the fact that it appears to be diminishing with increasing basis set size (see, for example, CR-CCL/aug-cc-pVDZ versus CR-CCL/aug-cc-pVTZ). We cannot confirm this however, as the shoulder persists for all models, even when a complete basis extrapolation is employed (although the CR-CCL/CBS curve appears to be an outlier). We can only confirm that increasing the basis set generally destabilizes the complex at smaller bond lengths relative to longer ones, and after about 3 Å, the curves seem to converge. It is noteworthy that this behavior is consistently observed in various coupled cluster methodologies, including CR-CCL, which is meant to be superior to the so-called “gold standard” CCSD(T) approach for bond-breaking situations due to the triple contribution to the energy being calibrated from the difference between the exact solution and the CCSD energy. Even with the exploration of configuration space via a CAS approach, we do not observe a typical Morse-like potential shape of the PES. Energy Decomposition. To further explore the theoretical rationale for such a result, the F3B−PH3 PES was decomposed into its Lewis acid and base constituents (Eacid, Ebase) as well as the interaction energy (Eint), as described above. These are plotted in Figure 3 along with the overall PES, Etotal.
their isolated structures. To accomplish this, at each point along the PES, the electronic energies of all of the Lewis acids and bases are calculated separately by holding them frozen in the geometry they adopted in the full complex at that point. These are referred to as Eacid and Ebase, respectively, and they quantify the energetic cost of rearranging the geometries of the isolated Lewis acid or base from their individual minimum-energy structure. The interaction energy between each of the “frozen” fragments (Eint) may then also be calculated for each point as the difference between the sum of the component energies (Eacid + Ebase) and that of the original point along the PES (Etotal). E int = Etotal − E PH3 − E F3B
(3)
Decomposing the PES in this way provides considerable insight into to the nature of bonding as the bond is stretched and/or compressed and will be useful in our context.
■
RESULTS AND DISCUSSION PES of F3B−PH3. The PES of F3B−PH3 has been calculated by employing several fundamentally different computational techniques of varying theoretical rigor. Some representative results are plotted in Figure 2, and with very few exceptions, the
Figure 2. PES scans of F3B−PH3 from several levels of theory, where r represents the B−P dative bond length in angstroms (Å) from the relaxed PES scan.
results are surprisingly consistent. In some cases, the shoulder is more pronounced than in others; however, it persists in all cases. This is an interesting feature that is not present in earlier work,39,40 where such subtle features of the PES were sensitive to the choice of theoretical model. The global minimum in each case is usually found at ≈3.1 Å, which is consistent with the relative weakness of the dative bond (low dissociation limit relative to the minimum) in this case.17 Although it is nontrivial to ascertain exactly which model should be considered the theoretical “reference” in such a case, it is interesting to note that the shoulder point at r < req persists despite an extensive exploration of the space of configuration state functions and the space of basis functions. Although the MP2/cc-pVQZ model does appear to be somewhat of an outlier (see Supporting Information), we have not concerned ourselves with exploring such a result further as our primary focus is the qualitative bonding behavior in these phosphine haloboranes (which is conserved in the MP2/ccpVQZ curve). Furthermore, perturbation theory is well known
Figure 3. PESs of the dative bond in F3B−PH3, BF3, PH3, and the interaction energy between the two components calculated at the M06L/aug-cc-pVQZ level of theory.
Not surprisingly, the phosphine component (Ebase) does not experience a large fluctuation in energy at all along the entire curve, which is a reflection of the relatively small structural change of this trigonal pyramidal fragment as the dative bond is formed. The boron trifluoride fragment however (Eacid) exhibits a steep relaxation energy profile as it progresses from a highly “compressed” (and relatively unstable) trigonal pyramidal structure at small r (and small ∠FBF) to a fully relaxed trigonal planar structure at large r (∠FBF = 120°). The stabilizing component of the PES comes from the favorable interaction (Eint) between each of these fragments and this is plotted in green. Although Eint is generally negative for most of the range plotted, the rather deep well in the Eint curve 610
DOI: 10.1021/acsomega.7b01529 ACS Omega 2018, 3, 608−614
Article
ACS Omega
It is therefore apparent that the unusual behavior observed in the PES of dative bond stretching in F3B−PH3 is the result of a confluence of several factors. The Lewis acid and base experience a long-range electrostatic attraction beginning at about 3.5 Å. The magnitude of this attraction slowly increases until the components reach about 2.5 Å separation, at which point the attraction becomes much stronger and |Eint| ultimately reaches its maximum at 1.9 Å. This is consistent with the expected Lennard-Jones-like r−6 behavior of interacting molecular species at long range. The overall potential, however, is further complicated by the necessary pyramidalization of the BF3 fragment to allow for dative bond formation. Such pyramidalization results in (i) the lowering of the BF3 LUMO with the decrease in ∠FBF (subsequently stabilizing the complex) and (ii) an associated increase in steric repulsion (destabilizing the complex). These factors are of similar magnitude, but the destabilizing steric repulsion is sufficient to negate the interaction energy of the fragments at small r. Because the electrostatic attraction is present even at large r, prior to any significant degree of pyramidalization of the BF3 fragment, the global minimum energy structure for this species occurs at r ≈ 3.2 Å and a feature resembling a shoulder is observed in the PES at smaller r, where |Eint| is larger. This behavior is in agreement with many earlier experimental investigations of F3B−PH3,17,22,23 which report a loosely formed adduct whose bond length is difficult to determine. Comparison with Analogous Systems. We can use this result as the basis for a guided exploration of chemical space to further elucidate this phenomenon. For example, if the multiple inflections in the PES of F3B−PH3 are the result of the interaction energy of the Lewis acid/base pair being similar in magnitude (albeit smaller) to the energetic cost of the required pyramidalization of the Lewis acid, then one should probe cases with modified electrostatic interactions and steric penalties associated with dative bond formation. Several obvious candidates include the aforementioned F3B− NH3 and H3B−PH3 complexes presented earlier in Figure 1. Decompositions of their potential energy surfaces are presented in Figure 5A,B, respectively, in the same manner as shown in Figure 3. As in the case of Figure 3, we note that Ebase does not significantly contribute to the features in any of the overall PESs. In the case of F3B−NH3, we have a system whose Eacid is essentially the same as that for F3B-PH3 (see Figure 5A). Although the NH3 Lewis base must get closer to the BF3 fragment for the dative bond to form, the ∠FBN angle at req is about 104°, which compares with the ∠FBP angle of F3B−PH3 at req of 105°. The interaction energy between the acid−base pair, however, is significantly stronger for NH3 than for PH3, which is consistent with the earlier theoretical work of Bessac and Frenking.26 It is well known that NH3 is a stronger Lewis base in this context and this effect outweighs the cost of geometric rearrangement of the BF3 fragment, which results in a far more “typical” PES. Alternatively, the arsenic analogue F3B−AsH3 should exhibit far weaker electrostatic interaction and, accordingly, has not been successfully synthesized.22 We can decrease the energetic cost of pyramidilization of the Lewis acid by considering BH3 in the case of H3B−PH3. Figure 5B illustrates the intuitive result that the BH3 fragment does not exhibit a severe pyramidilization penalty (Eacid) in the relevant distance range and, as such, the overall PES resembles the Eint curve quite closely. Again, the result is that Eint is able to dominate Eacid and produce a typical PES.
demonstrates a strong stabilizing interaction between the F3B− and −PH3 fragments near ≈1.9 Å, coinciding with the shoulder inflection in the PES. At long range, the electrostatic attraction of the fragments is small but significant enough to bind them, owing to the fact that Ebase and Eacid are negligible and the result is a global minimum predicted at r ≈ 3.2 Å. In other words, the Lewis acid and base experience an electrostatic attraction prior to any geometrical rearrangement. Molecular Orbital Analysis. Analysis of the molecular orbitals (MOs) by visualizing and comparing them at key bond distances provides deeper insight into these components of the PES of the complex. Figure 4 illustrates the frontier MOs of the
Figure 4. Frontier MO energy-level diagrams of the boron trifluoride fragment of F3B−PH3, calculated at the geometries of several bond lengths, r. Data are obtained from a M06L/6-311+G* calculation, and images are plotted at an isovalue of 0.07 au.
F3B− fragment of F3B−PH3 in geometries that match its structure at bond lengths of 1.3, 1.9, 2.5, and 3.1 Å from the full dative complex. Generally, the pyramidalization of the BF3 component induces a concomitant rise in the energy of almost all of the MOs, with the exception of the lowest unoccupied molecular orbital (LUMO). This orbital is primarily composed of side-on overlap of equivalent p3 becomes more pyramidal. In the range of 1.9 ≤ r ≤ 3.1 Å, this effect lowers the LUMO energy much more significantly than the increase in energy of the occupied orbitals. The highest occupied molecular orbital (HOMO)−LUMO gap is therefore shrinking significantly with decreasing ∠FBF and this is the driving force of the favorable interaction component of the PES. The occupied orbitals of BF3 are increasing in energy as ∠FBF decreases, which destabilizes the molecule (Eacid) but simultaneously allows the molecule to be more susceptible to dative bonding by an incoming lone pair. Dative bond formation can be thought of as the sp3 → p σ donation of electron density from the Lewis base to the acid and thus the LUMO of the Lewis acid is a significant factor in determining the strength of the interaction.25 Although there are many ways to probe this from a theoretical standpoint, inspection of the simple Mulliken populations in our case indicates that the phosphine fragment becomes increasingly positive as it approaches the haloborane. With all other things being equal, the lower the LUMO level, the more favorable dative bond formation will be and the more stabilizing Eint will be. This stabilization of the BF3 LUMO increases indefinitely with the pyramidalization of the BF3 fragment, but it is ultimately outweighed by the unfavorable change in each of the other MOs in the system as ∠FBF decreases. 611
DOI: 10.1021/acsomega.7b01529 ACS Omega 2018, 3, 608−614
Article
ACS Omega
length of this complex16 and previous geometry optimizations performed at the density functional BP86/TZ2p level.26 The second minimum is observed at r ≈ 3.5 Å and is confirmed by several independent, fully relaxed geometry optimizations at the density functional and ab initio levels of theory (see Supporting Information). To the best of our knowledge, this is the first example of a bimodal dissociation curve for a B−P dative bond in a single small molecule. Although similar behavior has been reported for CH3CN−BF3,40 it seems that our distinct minima are separated by much higher energetic barriers and consequently less sensitive to the choice of theoretical model. This rather remarkable feature is the result of a stronger close-range interaction between BCl3 and PH3, relative to BF3 due to the smaller HOMO−LUMO gap in the pyramidal chlorinated analogue (0.211 versus 0.336 au for BF3). Although decreasing ∠ClBCl comes at a similar energetic cost to decreasing ∠FBF, the LUMO for pyramidal boron trichloride is significantly lower than that of pyramidal boron trifluoride. This stabilizes the dative bond and is in close agreement with earlier comparisons of these Lewis acids.25 Because this enhanced stabilization only occurs at a close range (i.e., r ≈ 1.9 Å), the long-range behavior of the PES is relatively unaffected and the minimum at r ≈ 3.5 Å persists and is comparable to the minimum observed for the F3B−PH3 analogue. At this separation, the Lewis acid and base are electrostatically bound without the acid having to distort its trigonal planar geometry. The fully optimized structure of Cl3B−PH3 at r ≈ 3.5 Å is shown in Figure 6. Between 2.5 and 3.0 Å, the energetic cost of
Figure 6. Optimized geometry of the BCl3−PH3 dative complex at the MP2/aug-cc-pVTZ level of theory.
Figure 5. PES scans of the dative bonds in (A) F3B−NH3, (B) H3B− PH3, and (C) Cl3B−PH3 along with their decomposed components and the interaction energy between the fragments at various geometries, calculated at the M06L/cc-pVQZ level of theory. The inset plot of (C) illustrates Etotal for BCl3PH3 magnified.
pyramidalization for BCl3 increases faster than the electrostatic attraction of the fragments and a maxima is observed separating each distinct minimum.
■
Our search for simple substituents on boron that would induce a higher cost of pyramidilization led us to consider the chlorinated analogue BCl3, owing to the larger size of the chlorine atoms, relative to fluorine. In their thorough review, Staubitz et al. reported that the cost of distorting BCl3 from trigonal planar to pyramidal was higher than that of BF3;23 however, this is in stark contrast to both Hirao et al.24 and Frenking.25 Nonetheless, we present our results in Figure 5C for the Cl3B−PH3 complex and note a particularly remarkable result. We would first like to point out that the Eacid is nearly identical for each of BCl3 and BF3 paired with PH3, in perfect agreement with earlier reports.24,25 What is particularly noteworthy however is that the overall PES of the Cl3B−PH3 complex actually exhibits two distinct minima, apparently harkening back to the classic notion of bond stretch isomerism.59 The first is observed at req ≈ 2.0 Å, in close agreement with experimental estimations of the dative bond
CONCLUSIONS We have shown that the potential energy surface for dative bond stretching in a series of haloboranes demonstrates unusual behavior and is not able to be fit to any traditional Morse or Lennard-Jones-type potential. Specifically, in the case of F3B−PH3, the PES exhibits a shoulder at a dative bond length shorter than the predicted equilibrium value. Additionally, the chlorinated analogue (Cl3B−PH3) exhibits similar behavior but the effects are amplified and the PES is observed to have two distinct minima. This is explained by the competition between the energetic cost of the required pyramidalization of the Lewis acid to form a dative bond and the stabilization from the interaction energy between the Lewis acid and base (while being held frozen at their respective geometries from the complex). When the cost of pyramidalization of the Lewis acid is high relative to the electrostatic 612
DOI: 10.1021/acsomega.7b01529 ACS Omega 2018, 3, 608−614
Article
ACS Omega
(3) Allen, T. L.; Fink, W. H. Theoretical studies of boranamine and its conjugate base. Comparison of the boron-nitrogen and boronphosphorus .pi. bond energies. Inorg. Chem. 1993, 32, 4230−4234. (4) Anane, H.; Jarid, A.; Boutalib, A.; Nebot-Gil, I.; Tomás, F. Ab initio molecular orbital study of the substituent effect on phosphineborane complexes. Chem. Phys. Lett. 1998, 296, 277−282. (5) Janesko, B. G. Using Nonempirical Semilocal Density Functionals and Empirical Dispersion Corrections to Model Dative Bonding in Substituted Boranes. J. Chem. Theory Comput. 2010, 6, 1825−1833. (6) Jensen, J. O. Vibrational frequencies and structural determination of phosphirane. J. Mol. Struct.: THEOCHEM 2004, 686, 165−172. (7) Plumley, J. A.; Evanseck, J. D. Covalent and Ionic Nature of the Dative Bond and Account of Accurate Ammonia Borane Binding Enthalpies. J. Phys. Chem. A 2007, 111, 13472−13483. (8) Plumley, J. A.; Evanseck, J. D. Periodic Trends and Index of Boron Lewis Acidity. J. Phys. Chem. A 2009, 113, 5985−5992. (9) Plumley, J. A.; Evanseck, J. D. Hybrid Meta-Generalized Gradient Functional Modeling of Boron-Nitrogen Coordinate Covalent Bonds. J. Chem. Theory Comput. 2008, 4, 1249−1253. (10) Huang, Z.; Autrey, T. Boron-nitrogen-hydrogen (BNH) compounds: recent developments in hydrogen storage, applications in hydrogenation and catalysis, and new syntheses. Energy Environ. Sci. 2012, 5, 9257−9268. (11) Seidler, E. A.; Lieven, C. J.; Thompson, A. F.; Levin, L. A. Effectiveness of Novel Borane-Phosphine Complexes In Inhibiting Cell Death Depends on the Source of Superoxide Production Induced by Blockade of Mitochondrial Electron Transport. ACS Chem. Neurosci. 2010, 1, 95−103. (12) Schlieve, C. R.; Tam, A.; Nilsson, B. L.; Lieven, C. J.; Raines, R. T.; Levin, L. A. Synthesis and characterization of a novel class of reducing agents that are highly neuroprotective for retinal ganglion cells. Exp. Eye Res. 2006, 83, 1252−1259. (13) Shi, R.; Shao, J.; Wang, C.; Zhu, X.; Lu, X. Search for structures, potential energy surfaces, and stabilities of planar BnP(n = 1−7). J. Mol. Model. 2011, 17, 1007−1016. (14) Stephens, F. H.; Baker, R. T.; Matus, M. H.; Grant, D. J.; Dixon, D. A. Acid initiation of ammonia-borane dehydrogenation for hydrogen storage. Angew. Chem., Int. Ed. 2007, 46, 746−749. (15) Umeyama, H.; Kudo, T.; Nakagawa, S. Molecular orbital study on the structure and barrier to internal rotation of phosphine borane. Chem. Pharm. Bull. 1981, 29, 287−292. (16) Odom, J. D.; Riethmiller, S.; Witt, J. D.; Durig, J. R. Spectra and structure of phosphorus-boron compounds. II. A vibrational and nuclear magnetic resonance spectral investigation of phosphinetrichloroborane. Inorg. Chem. 1973, 12, 1123−1127. (17) Durig, J. R.; Riethmiller, S.; Kalasinsky, V. F.; Odom, J. D. Spectra and structure of phosphorus-boron compounds. VI. Vibrational spectra ans structure of some phosphine-trihaloboranes. Inorg. Chem. 1974, 13, 2729−2735. (18) Sibbald, P. A. A theoretical analysis of substituent electronic effects on phosphine-borane bonds. J. Mol. Model. 2016, 22, 261. (19) Jonas, V.; Frenking, G.; Reetz, M. T. Comparative Theoretical study of Lewis acid-base complexes of BH3, BF3, BCl3, AlCl3, and SO2. J. Am. Chem. Soc. 1994, 116, 8741−8753. (20) Martin, D. R. Coordination compounds of boron trichloride. I. A review. Chem. Res 1944, 34, 461−473. (21) Booth, H.; Martin, D. Boron Trifluoride and Its Derivatives; John Wiley and Sons, Inc.: New York, NY, 1949; Chapter 4. (22) Martin, D.; Dial, R. E. Systems of boron trifluoride with phosphine, arsine, and hydrogen bromide. J. Am. Chem. Soc. 1950, 72, 852−856. (23) Staubitz, A.; Robertson, A. P. M.; Sloan, M. E.; Manners, I. Amine- and Phosphine-Borane adducts: New interest in old molecules. Chem. Rev. 2010, 110, 4023−4078. (24) Hirao, H.; Omoto, K.; Fujimoto, H. Lewis acidity of boron trihalides. J. Phys. Chem. A 1999, 103, 5807−5811. (25) Bessac, F.; Frenking, G. Why is BCl3 a stronger Lewis acid with respect to strong bases than BF3? Inorg. Chem. 2003, 42, 7990−7994.
attraction between the acid and base, the potential well associated with dative bonding is significantly affected and the result is a relatively flat PES that is susceptible to the unusual characteristics described herein. The species studied in this work have been investigated by a wide variety of computational techniques, including many DFT methods and many levels of wave function approximations with a wide range of basis sets. All calculations yielded qualitatively similar results. Although the PES behavior described herein has persisted in all of our calculations, it is important to note that atypical inflections do seem to diminish with increasing basis set size. In particular, increasing the size and flexibility of the basis set seems to destabilize the PES at small r, relative to the large r portion of the curve (e.g., see Figure 1). It is conceivable that the shoulder observed for the F3B−PH3 complex PES could be resolved once higher level theories are applied; however, the inflections are even more pronounced for the Cl3B−PH3 complex and the small r portion of the curve that we have presented already agrees well with both experimental and additional theoretical benchmarks. In any case, we expect that isolation of the weakly bound Cl3B−PH3 complex at r ≈ 3.5 Å would be a significant experimental challenge.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsomega.7b01529. Full PES data and optimized structures of the F3B−PH3 and Cl3B−PH3 complexes, respectively, from a series of computational models (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Phone: +1 902 566 0934. Fax: +1 902 566 0632. ORCID
Jason K. Pearson: 0000-0002-1001-1606 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank each of the anonymous referees for helpful suggestions to improve the quality of the current manuscript. They acknowledge the Natural Sciences and Engineering Research Council (NSERC) of Canada and the Canada Foundation for Innovation (CFI) for funding that supported this work. Computational facilities are provided by the Atlantic Computational Excellence Network (ACENET), the regional high-performance computing consortium for universities in Atlantic Canada. ACENET is funded by CFI, the Atlantic Canada Opportunities Agency (ACOA), and the provinces of Newfoundland & Labrador, Nova Scotia, and New Brunswick.
■
REFERENCES
(1) Bryan, P. S.; Kuczkowski, R. L. Microwave spectra, structures, and dipole moments of trimethylphosphine-borane and methylphosphineborane. Inorg. Chem. 1972, 11, 553−559. (2) Durig, J. R.; Shen, Z. Investigation of the structure, bonding, vibrational frequencies and stability by ab initio calculations of H3BPH3, H3B-PHF2, and H3B-PF3. J. Mol. Struct.: THEOCHEM 1997, 397, 179−190. 613
DOI: 10.1021/acsomega.7b01529 ACS Omega 2018, 3, 608−614
Article
ACS Omega (26) Bessac, F.; Frenking, G. Chemical bonding in phosphane and amine complexes of main group elements and transition metals. Inorg. Chem. 2006, 45, 6956−6964. (27) Huang, F.; Jiang, J.; Wen, M.; Wang, Z.-X. Assessing the performance of commonly used DFT functionals in studying the chemistry of frustrated Lewis pairs. J. Theor. Comput. Chem. 2014, 13, No. 1350074. (28) Holme, T. A.; Truong, T. N. A test of density functional theory for dative bonding systems. Chem. Phys. Lett. 1993, 215, 53−57. (29) Gilbert, T. M. Tests of the MP2 model and various DFT models in predicting the structures and B-N bond dissociation energies of amine-boranes (X3C)mH3-mB-N(CH3)nH3-n (X = H, F; m = 0−3; n = 0−3): poor performance of the B3LYP approach for dative B-N bonds. J. Phys. Chem. A 2004, 108, 2550−2554. (30) Pupim, C. F.; Morgon, N. H.; Lopez-Castillo, A. Spurious Phosphorus Pyramidalization Induced by Some DFT Functionals. J. Braz. Chem. Soc. 2015, 26, 1648−1655. (31) Leyssens, T.; Peeters, D.; Orpen, A. G.; Harvey, J. N. Insight into metal-phosphorus bonding from analysis of the electronic structure of redox pairs of metal-phosphine complexes. New J. Chem. 2005, 29, 1424−1430. (32) Anane, H.; El Houssame, S.; El Guerraze, A.; Jarid, A.; Boutalib, A.; Nebot-Gil, I.; Tomás, F. Ab initio molecular orbital study of the substituent effect on ammonia and phosphine-borane complexes. J. Mol. Struct.: THEOCHEM 2004, 709, 103−107. (33) Becke, A. D. A new mixing of Hartree-Fock and local densityfunctional theories. J. Chem. Phys. 1993, 98, 1372−1377. (34) Head-Gordon, M.; Pople, J. A.; Frisch, M. J. MP2 energy evaluation by direct methods. Chem. Phys. Lett. 1988, 153, 503−506. (35) Weigend, F.; Ahlrichs, R. Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy. Phys. Chem. Chem. Phys. 2005, 7, 3297−3305. (36) Weigend, F. Accurate Coulomb-fitting basis sets for H to Rn. Phys. Chem. Chem. Phys. 2006, 8, 1057−1065. (37) Christiansen, O.; Koch, H.; Jørgensen, P. The second-order approximate coupled cluster singles and doubles model CC2. Chem. Phys. Lett. 1995, 243, 409−418. (38) Almas, Q. L.; Keefe, B. L.; Profitt, T.; Pearson, J. K. Choosing an appropriate model chemistry in a big data context: Application to dative bonding. Comput. Theor. Chem. 2016, 1085, 46−55. (39) Alkorta, I.; Elguero, J.; Bene, J. E. D.; Mo, O.; Yanez, M. New insights into factors influencing B-N bonding in X:BH3−nFn and X:BH3−nCln for X = N2, HCN, LiCN, H2CNH, NF3, NH3 and n = 0− 3: The importance of deformation. Chem. Eur. J. 2010, 16, 11897− 11905. (40) Giesen, D. J.; Phillips, J. A. Structure, bonding, and vibrational frequencies of CH3CN-BF3: New insight into medium effects and the discrepancy between the experimental and theoretical geometries. J. Phys. Chem. A 2003, 107, 4009−4018. (41) Zhao, Y.; Truhlar, D. G. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. In Theoretical Chemistry Accounts; Springer, 2008; Vol. 120, pp 215−241. (42) Scuseria, G. E.; Schaefer, H. F., III Is coupled cluster singles and doubles (CCSD) more computationally intensive than quadratic configuration-interaction (QCISD)? J. Chem. Phys. 1989, 90, 3700− 3703. (43) Pople, J. A.; et al. Quadratic configuration interaction - a general technique for determining electron correlation energies. J. Chem. Phys. 1987, 87, 5968−5975. (44) Piecuch, P.; Włoch, M. Renormalized coupled-cluster methods exploiting left eigenstates of the similarity-transformed Hamiltonian. J. Chem. Phys. 2005, 123, No. 224105. (45) Włoch, M.; Gour, J. R.; Piecuch, P. Extension of the Renormalized Coupled-Cluster Methods Exploiting Left Eigenstates
of the Similarity-Transformed Hamiltonian to Open-Shell Systems:? A Benchmark Study. J. Phys. Chem. A 2007, 111, 11359−11382. (46) Stevens, W. J.; Krauss, M.; Basch, H.; Jasien, P. G. Relativistic compact effective potentials and efficient, shared-exponent basis sets for the third-, fourth-, and fifth-row atoms. Can. J. Chem. 1992, 70, 612−630. (47) Aarset, K.; Hagen, K.; Stølevik, R.; Christian, S. P. Molecular structure and conformational composition of 1-Chlorobutane, 1Bromobutane, and 1-Iodobutane as determined by gas-phase electron diffraction and ab initio calculations. Struct. Chem. 1995, 6, 197−205. (48) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; et al. General atomic and molecular electronic structure system. J. Comput. Chem. 1993, 14, 1347−1363. (49) Halkier, A.; et al. Basis-set convergence in correlated calculations on Ne, N2 and H2O. Chem. Phys. Lett. 1998, 286, 243−252. (50) Siegbahn, P. E. M.; Almlöf, J.; Heiberg, A.; Roos, B. O. The complete active space SCF (CASSCF) method in a Newton-Raphson formulation with application to the HNO molecule. J. Chem. Phys. 1981, 74, 2384−2396. (51) Lee, T. J.; Taylor, P. R. A diagnostic for determining the quality of single-reference electron correlation methods. Int. J. Quantum Chem. 1989, 36, 199−207. (52) Veryazov, V.; Malmqvist, P.; Roos, B. How to Select Active Space for Multiconfigurational Quantum Chemistry? Int. J. Quantum Chem. 2011, 111, 3329−3338. (53) Shao, Y.; et al. Advances in methods and algorithms in a modern quantum chemistry program package. Phys. Chem. Chem. Phys. 2006, 8, 3172−3191. (54) Gordon, M. S.; Schmidt, M. W. Advances in electronic structure theory: GAMESS A Decade Later. In Theory and Applications of Computational Chemistry: The First Forty Years; Dykstra, C. E., Frenking, G., Kim, K. S., Scuseria, G. E., Eds.; Elsevier: Amsterdam, 2005; pp 1167−1189. (55) Frisch, M. J., et al. Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford, CT, 2009. (56) Leininger, M.; Allen, W.; Schaefer, H. F., III; Sherrill, C. Is Moller-Plesset perturbation theory a convergent ab initio method? J. Chem. Phys. 2000, 112, 9213−9222. (57) Dunning, T. H., Jr. Gaussian basis sets for use in correlated molecular calculations. I. The atoms boron through neon and hydrogen. J. Chem. Phys. 1989, 90, 1007−1023. (58) Woon, D. E.; Dunning, T. H., Jr. Gaussian basis sets for use in correlated molecular calculations. 3. The atoms aluminum through argon. J. Chem. Phys. 1993, 98, 1358−1371. (59) Parkin, G. Bond-stretch isomerism in transition metal complexes: a reevaluation of crystallographic data. Chem. Rev. 1993, 93, 887−911.
614
DOI: 10.1021/acsomega.7b01529 ACS Omega 2018, 3, 608−614
