Chemical Science

0 downloads 0 Views 443KB Size Report
Sep 12, 2013 - singlet cyclic alkylamino carbenes (CAACs).3 The latter was rationalized in ... singlet carbene reacting with CO to form a stable, albeit unusual.

Chemical Science View Article Online

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

EDGE ARTICLE

Cite this: Chem. Sci., 2013, 4, 4309

View Journal | View Issue

Coordination of CO to low-valent phosphorus centres and other related P–C bonding situations. A theoretical case study† Gerd von Frantzius,a Arturo Espinosa Ferao*b and Rainer Streubel*a The multi-faceted bonding of CO in molecular phosphorus compounds is described using calculated P–C bond strengths as a criterion. Full compliance matrices at coupled cluster level of HPCO (1a), singlet oxaphosphirane-3-ylidene HP(h2-CO)), the dimer (HPC]O)2 as well as P^CH, HP]CH2 and H2P–CH3 were calculated to obtain quantifiable data and enable comparison. The quest for CO coordination and ˚ mdyn1) reveal a activation was examined for phosphaketenes 1a–f: the P–C compliance constants (in A clear trend that shows a weakening of the P–CO bond strength from 1a to mono-ligation as in [(OC)5W {P(CO)Me}] (1c) (0.301), in H3BP(CO)Me (1b) (0.322), to bis-ligation as in [{(OC)5W}2P(CO)R] (1f) (0.488) to (H3BP)2(CO)Me (1d) (0.649). Availability of p-type electron density at phosphorus drastically strengthens the P–CO bond and weakens the C–O bond via p–back-donation, bis complexes are better described as weak CO (C/P) adducts to phosphorus. In complexes [(OC)5W{P(CO)R}] the CO activation by phosphorus equals that of CO activation through tungsten in pentacarbonyltungsten complexes. A comparative study of various CO bonding motifs in molecular compounds indicates that acyclic (2) or cyclic diphospha-urea derivatives (2–5) or isomers (6) display P–CO bond strengths (compliance constants range 0.502–0.640) well below that of the P–C bond of H2P–CH3 (0.364), thus providing insight into the bonding and the ease of CO extrusion, experimentally known for some cases. A highly unusual adduct of CO was obtained in silico through two-fold P-ligation in diphosphiren-3-ones 2a–d,

Received 19th July 2013 Accepted 19th August 2013 DOI: 10.1039/c3sc52027g www.rsc.org/chemicalscience

the parent compound of which was found to be properly described as a side-on (P]P)/(C]O) complex, in contrast to its aza-analogue 2aN. A drastic weakening of the P–CO bond strength is observed from P2CO (2a) (0.502) to the C2-symmetrical (H3BP)2CO (2b) (0.913); the latter represents an extreme case of a weakly bound CO. Furthermore, calculated 31P NMR shifts and scalar 1J(P,E) couplings were correlated with P–CO and PC–O compliance constants as a tool for experimentalists.

Introduction Based on increasing evidence, a fundamental quest has emerged in recent years: can small molecules be activated at molecular non-metal centres and to which extent?1 The problem shall be illustrated by two milestones: (1) reversible uptake of hydrogen by a phosphinoborinane to yield a phosphonium borate2 and (2) the (non-reversible) hydrogenation of singlet cyclic alkylamino carbenes (CAACs).3 The latter was rationalized in terms of activation of dihydrogen at a carbene centre and strong parallels to metal-based bond activation were drawn. The same group also provided the rst example of a a

Institut f¨ ur Anorganische Chemie der Reinischen Friedrich-Wilhelms-Universit¨ at Bonn, Gerhard-Domagk-Str.1, 53121 Bonn, Germany. E-mail: [email protected] de; Fax: +49 228 739616; Tel: +49 228 735345 Departamento de Qu´ımica Org´ anica, Universidad de Murcia, Campus de Espinardo, 30100 – Murcia, Spain. E-mail: [email protected]; Fax: +34 868 884149; Tel: +34 868 887499

b

† Dedicated to Prof. G. Frenking on the occasion of his 66th birthday.

This journal is ª The Royal Society of Chemistry 2013

singlet carbene reacting with CO to form a stable, albeit unusual bent ketene derivative.4 Since neutral terminal phosphinidene complexes, [(CO)5M(PR)],5 show electrophilic singlet carbenetype reactivity, the recently reported addition of hydrogen6 was likely to occur. On the other hand, the reaction of neutral complexes [(CO)5M(PR)] with CO at phosphorus, i.e. CO coordination to phosphorus, is still unknown.7 In contrast, the ability of transient nitrilium phosphane-ylide complexes [(CO)5M {P(N^CR0 )R}]8–10 for transylidation was noticed, as typical features of transition-metal coordination chemistry came to the fore. In the same vein, a study on the reaction of isonitriles with electrophilic phosphinidene complexes was initiated, but the transiently formed end-on phosphorus adducts [(CO)5M{P(C]NR0 )R}] showed a surprising thermal instability and decomposition yielded isobutene and an organo(cyano)-phosphane complex.11 Unexpected was also the outcome of the reaction of [(CO)5M(PR)] with CO2 as deoxygenation leading to CO and head-to-tail dimers of transient complexes [(CO)5M-(OPR)] was observed.12

Chem. Sci., 2013, 4, 4309–4322 | 4309

View Article Online

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

Chemical Science In the area of dinuclear electrophilic phosphinidene complexes and their higher homologues [{LnM}2PnR] (LnM ¼ 16e transition-metal complex, pnictogen (Pn), Pn ¼ P, As, Sb, Bi, R ¼ e.g. alkyl, aryl etc.) it has long been known that the central Pn atom can add Lewis bases (e.g. thf, acetone, NH3, NEt3, TMEDA, pyridine)13,14 – thus representing an early landmark in the coordination-to-phosphorus chemistry.15 In this case the mostly reversible and strongly temperature dependent adduct formation is accompanied by drastic changes in colour caused by the disturbance of the M–P–M p-system. In a recent study, the P-acetonitrile adduct [{W(CO)5}2P(N^CMe)Cp*] was deduced based on 31P NMR spectroscopy.16 In a broader context and although dissociation has not been reported as yet, “adducts” of N-heterocyclic carbenes to pnictinidenes, formally described by the formula NHC/PnR (Pn ¼ P, As), should be mentioned.17 Upon borane complexation of the phosphorus centre of such an NHC–PR derivative, a P–C(carbene) bond lengthening and increase of the C–P–C bond angle brought an interesting bonding feature to the fore: a loss of p back-bonding from phosphorus to the carbene carbon atom.17b Very recently, a correlation was proposed between p back-bonding properties of carbenes in such RP adducts and 31P NMR chemical shi values, thus offering a new type of scaling system.18 The nding of NHC/PnR adduct formation, with NHCs as strong Lewis donors, has initiated a new stabilization strategy that led to a series of interesting NHC adducts of P2,19 Si2,20 and B2,21 or other low coordinated main-group element species.22 In this work the following fundamental questions are addressed: how can CO, as a relatively weak Lewis donor, be bound to a low-coordinate phosphorus centre (i) and which are the consequences in terms of small molecule activation? Secondly, can CO be kept in bridging bonding motifs if one or two phosphorus centres are directly bound in acyclic and/or cyclic scaffolds (ii), and how strong are such P–C bonds? Therefore, the bonding situation of the “probe-ligand” CO attached end-on to a low-coordinate phosphorus centre in various environments (1a–f) (Fig. 1) was theoretically investigated and compliance constants calculated in order to quantify P–C and C–O bond strength changes. Since a set of reference compounds was required for the method of compliance constants used in this work, compounds 2–6 (Fig. 2) (derivatives of 2–5 are known experimentally) as well as H3C–PH2, H2C]PH, HC^P, the end-on CO adduct of the

Fig. 1 CO Donor adducts 1a–f of neutral phosphinidenes and complexes thereof.

4310 | Chem. Sci., 2013, 4, 4309–4322

Edge Article

Fig. 2

Calculated reference compounds 2–6 featuring a P–CO structural motif.

cationic phosphinidene complex [Cp*(CO)3W(PNiPr2)]+ (7) ([Cp*(CO)3W(P(CO)NiPr2)]+ (8)) and the phosphino carbonyl complexes [FeCp*CO2{C(O)PH2}] (9) and [FeCp*CO2{C(O) P(BH3)H2}] (10) were also included in the calculations. The latter examples are also used to connect to the second part of this study dealing with disubstituted CO derivatives such as diphospha-urea (-like) derivatives 2–4 and various three- and four-membered P-heterocycles 3, 5, 6 were calculated, while emphasizing on the role of the lone pair at phosphorus.

Results and discussion Computational methods All optimizations and frequency calculations were performed using GAUSSIAN03,23 ADF-2007.0124 and ORCA.25 The standard DFT method with GAUSSIAN03 in this work is B3LYP26 together with the 6-311G(d,p) basis set eventually combined with an effective core potential (ECP) description of molybdenum and tungsten using LanL2DZ by Hay and Wadt (for short: B3LYP/6311**, LanL2DZ)27 or with the Ahlrich's triple-z valence def2TZVP28 basis set and the SD(28,MWD)(Mo) or SD(60,MWB) (W) ECP29 (B3LYP/def2TZVPecp). Also the Truhlar's functionals M05-2X, M06 and M06-2X30 were checked. Some important benchmark studies were also included using geometries obtained with B3LYP and the latest (third generation) Grimme's semi-empirical atom-pairwise London dispersion correction31 (B3LYP-D3/def2-TZVPecp) and energies at the remarkably accurate double-hybrid-meta-GGA functional PWBP95 plus dispersion correction (PWBP95-D3)32 and the largest def2TZVPP (ECP for Mo and W) basis set. Additionally reference compounds were computed at a coupled cluster level33 including single, double34 and non-iterative triple35 excitations on Dunning's correlation consistent polarized valence triple zeta basis36 (CCSD-T(fc)/cc-pVTZ). Stationary points have been characterized by analytical second derivatives (the Hessian) with respect to redundant Cartesian coordinates. Transformation into non-redundant internal force constants was done using algorithms by Pulay and Fogarasi (INTC/FCTINT).37 Inversion of the Hessian into the compliance matrix was accomplished by standard methods.38 Unless otherwise stated compliance constants are given at the B3LYP/6-311G(d,p) level, with ECP at Mo or W if necessary. 31P NMR chemical shis and nuclear spin–spin couplings were computed using ADFs ZORA GIAO-DFT included in the CPL39 module (see standard method

This journal is ª The Royal Society of Chemistry 2013

View Article Online

Edge Article

Chemical Science

Table 1 P–C bond lengths (d), harmonic stretching frequencies (~n) and compliance constants (C) of reference compounds at CCSD-T(fc)/cc-pVTZ and at B3LYP/6311g(d,p) levels

[W(CO)5L]

Uncomplexed ~n

a

b

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

d

c

C

L

Sym

Exp.

CCSD-T

B3LYP

Exp.

CCSD-T

B3LYP

CCSD-T

B3LYP

Sym

~nb B3LYP

Cc B3LYP

P^CH

CNv

1.5540

1.5393

127862 127763

1285

1342

0.1012

0.1106

C4v

1382

0.0965

HP]CH2 H2P–CH3

Cs Cs

1.54257 1.544258 1.5398(2)59 1.6661(3)60 1.86361

1.6810 1.8659

1.6697 1.8728

85064 67561

978 683

1003 666

0.1839 0.3639

0.1749 0.3899

Cs Cs

1023 703

0.1688 0.3547

a

˚ b Unscaled, in cm1. c In A ˚ mdyn1. In A.

˚ mdyn1) of P^CH, HP]CH2 and H2P–CH3 at Fig. 3 Compliance constants (A DFT and at coupled cluster levels.

above) where absolute isotropic shieldings were converted to common chemical shis via d(compound, calc.) ¼ s(PH3) – s(compound, calc.) – 266.1 (d PH3) ppm.40 The compliance matrix-method Compliance constants and interaction displacement coordinates were originally introduced into vibrational theory to have Table 2

uniquely dened and physically meaningful quantities at hand which are more likely to be transferrable among molecules than are force constants.41 Later on DFT compliance matrices were used by Grunenberg et al.42 to investigate the Ga–Ga bonding situation in [H–Ga–Ga–H]Na2/[H2Ga]GaH2]Na2, to assess the inter-residue forces of Watson–Crick base pairs and to look at the P–P bond in small polyphosphorus compounds, where shorter bonds were not necessarily found to be stronger than longer ones.43 By a similar method Schaefer et al. estimated the metal–metal bond character for the homoleptic transitionmetal carbonyls Fe2(CO)n and Co2(CO)n.44 For a short comparison of the use of force- and compliance constants in Cartesian and internal coordinates in the theory of molecular vibrations the article by Watson is recommended.45 The physical model behind force constants and thus compliance constants is a spring model: if the molecule at equilibrium geometry is distorted by a vibrational movement the various internal coordinates (modelled by springs) interact according to the molecular force eld. In the harmonic approximation to vibrational theory the molecular hypersurface describing a vibrational movement is locally approximated by a quadratic form (Hik) in the displacements of the internal

C–O and M–C bond lengths and compliance constants

Method

dC–Oa

CC–Ob

CO

Experiment CCSD-T(fc)/aug-cc-pVTZ CCSD-T(fc)/cc-pVTZ MP2(fc)/6-311g(d,p) M06/def2-TZVP B3LYP/6-311g(d,p)

1.128369 1.1311142 1.1311 1.1387 1.1232 1.1270

0.0526 0.0523 0.0524 0.0543 0.0495 0.0502

BH3CO

Experiment CCSD-T(fc)/cc-pVTZ MP2(fc)/6-311g(d,p) M06/def2-TZVP B3LYP/6-311g(d,p)

1.135(10)70 1.1332 1.1374 1.1254 1.1306

W(CO)6

Experiment M06/def2-TZVPc B3LYP/6-311g(d,p)c

1.14872 1.1370 1.1419

a

dM–Ca

CM–Cb

0.0530(7)71 0.0527 0.0537 0.0503 0.0518

1.534(10) 1.5558 1.5557 1.5212 1.5240

0.31(8) 0.3998 0.3845 0.3388 0.3281

0.05928(6)73 0.0558 0.0570

2.05872 2.0811 2.0695

0.450(4)73 0.5023 0.4453

˚ b In A ˚ mdyn1. c With the corresponding ECPs. In A.

This journal is ª The Royal Society of Chemistry 2013

Chem. Sci., 2013, 4, 4309–4322 | 4311

View Article Online

Chemical Science Table 3

Harmonic frequenciesa and rotational constantsb of HPCO (at DFT/6-311g(d,p) and at CCSD-T(fc)/cc-pVTZ levels)

Experimental80 B3LYP81 M06 CCSD-T(fc)

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

a

Edge Article

In cm1.

b

Sym

n(P–H)

n(PC–O)

n(HPC)

n(P–CO)

A0

B0

C0

Cs Cs Cs

2370 2400 2391

1998 2082 2121 2042

907 882 882 867

707 708 703

252.30687 252.68166 254.29077

5.53396 5.55576 5.48795

5.41518 5.43623 5.37201

In GHz.

coordinates (bonds, angles, dihedrals and linear combinations thereof) from their equilibrium values 2V ¼ Dxt(Hik)Dx, where (Hik) is the matrix of second derivatives (Hessian) at a stationary point, the diagonal elements Hkk of which are the force constants.46 A full discussion of the relation of force constants to bond strengths can be found in the report by Cremer et al.47 Equivalently the change in potential energy during a vibration is described by a quadratic form in the forces (force displacements Df to be precise) instilled in the coordinates upon distortion from equilibrium geometry 2V ¼ D f t(Cik)Df, where (Cik) ¼ (Hik)1 is the compliance matrix (inverse Hessian) at a stationary point, the diagonal elements Ckk of which are the ˚ mdyn1] for bond stretchings and compliance constants (in [A 1 in [rad mdyn ] for angle bendings), while the off-diagonal elements are associated with the couplings of the coordinates. They can be related to the mean amplitude of the kth interatomic distance, an observable obtained from electron diffraction.48 While in the spring model force constants Hkk describe the stiffness (resistance against distortion) the Ckk are associated with the compliance of coordinate k. The columns (rows) of the compliance matrix can be interpreted as follows. Suppose the molecule is being distorted in the sense of a partial optimization: aer relaxation there remains a unit force acting on coordinate k while the forces on all other coordinates vanish. In the harmonic approximation this situation is equivalent to multiplying (Cik) with a vector of unit force (in [mdyn]) yielding a column of (Cik), the entries of which can be interpreted as the distortion of the coordinates from their equilibrium values.49 This also holds for a transition state (TS), in case of which compliance constants can be negative. Indeed a TS is characterized by one negative force constant, i.e. one negative eigenvalue of the Hessian, indicating a negative curvature of the potential (a lack of restoring force) in the direction of the coordinate associated with it.50 A criticism on the use of compliance constants was reported by Baker and Pulay51 and Table 4

immediately replied by Grunenberg with an overview on the use of compliance constants to characterize bond strengths.52 Furthermore, compliance constants or their derived reciprocal values, the relaxed force constants, have been recently used to characterize a wide variety of interatomic interactions ranging from very strong42a,b,53 to very weak ones,42c,54 as well as for the evaluation of endocyclic strained bonds.12,55

Compliance constants of P–C reference compounds (P^CH, HP]CH2, H2P–CH3) Compliance constants as a measure of bond strength are only useful if referenced against a set of model compounds. In the present case the comparison between all (kinds of) P–C contacts of adducts 1–6 with the P–C compliances of phosphaalkyne P^CH, phosphaalkene HP]CH2 and methylphosphane H2P– CH3 is made, taking the P–C bonding situations of these compounds for granted. Ab initio and DFT force constants of phosphorus–carbon singly and multiply bonded species were given by Ohno.56 A comparison of experimental data and results from our coupled cluster and DFT calculations are collected in Table 1. There is a good agreement between experimental and calculated values, in general, with the exception of a large deviation for nP]C values. The reasonable agreement between the DFT methods used and the coupled cluster results is shown in Fig. 3. While for the triple P–C bond the compliance is slightly underestimated by the DFT calculations, this trend is reversed for the singly bonded methylphosphine, and the P]C bond shows a behaviour in between. This holds true also for the basic standard (B3LYP/6-311g(d,p), ECP at W if necessary) DFT level, whose larger deviation for the single P–C bond only improves scarcely on changing to a larger basis set. It is worth mentioning that the highly reputed Truhlar's M06-2X functional only performs better than B3LYP for the single P–C bond, but not for the

Compliance matrixa of HP(h2-CO) at coupled cluster and at DFT level

CCSD(T)(fc)/cc-pVTZ

P–CO CO H–P PO a

B3LYP/6-311g(d,p)

P–CO

CO

H–P

PO

P–CO

CO

H–P

PO

0.6353 0.0788 0.0067 0.2837

0.0788 0.1389 0.0035 0.1070

0.0067 0.0035 0.3061 0.0084

0.2837 0.1070 0.0084 0.6859

0.8296 0.0070 0.1108 0.4482

0.0070 0.1314 0.0028 0.1430

0.1108 0.0028 0.3110 0.0087

0.4482 0.1430 0.0087 0.9140

˚ mdyn1. Units of the matrix elements are A

4312 | Chem. Sci., 2013, 4, 4309–4322

This journal is ª The Royal Society of Chemistry 2013

View Article Online

Edge Article Table 5

Chemical Science

Compliance matrixa of C2h (HPCO)2 at coupled cluster and at DFT level

CCSD(T)(fc)/cc-pVTZ

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

P–CO CO H–P P–CO0 a

B3LYP/6-311g(d,p)

P–CO

CO

H–P

P–CO0

P–CO

CO

H–P

P–CO0

0.5033 0.0259 0.0091 0.0394

0.0259 0.0757 0.0004 0.0259

0.0091 0.0004 0.2983 0.0091

0.0394 0.0259 0.0091 0.5033

0.6099 0.0321 0.0069 0.0411

0.0321 0.0762 0.0004 0.0321

0.0069 0.0004 0.3097 0.0069

0.0411 0.0321 0.0069 0.6099

˚ mdyn1. Units of the matrix elements are A

double and triple bonds. What is important to note here is that the distribution of P–C bond strengths is being reliably reproduced by the DFT calculations. All further compliance constants will be compared with these organophosphorus model compounds. CO bond strength: the case of CO and H3BCO An important indicator of electron ow caused by bond formation is the change in the C–O force constant of free and bound CO; for a discussion of the usual s-bonding and p-backbonding formalism the reader is referred to Goldman and Krogh-Jespersen65 and the recent review of metal carbonyl clusters by Wade.66 Compliance constant calculations of C-protonated CO indicated a considerable enforcement of the ˚ mdyn1),67 while analysis of the full force C–O bond (0.046 A eld of borane carbonyl showed an almost “neutral” effect of complexation on CO: bond enforcement due to the B–C s-bond formation is counterbalanced by a hyperconjugative effect of the borane group; a recent DFT energy decomposition analysis of H3BCO was given by Erhardt and Frenking.68 Goldman and Krogh-Jespersen, however, found no correlation between the CO force constant and the extent of s-bonding. The calculations presented here (Table 2) indicate a minor weakening of CO, which is interpretable in terms of BH-hyperconjugation. Hence, the working hypothesis here is that s-bonding of CO has almost no effect on the C–O bond strength. Consequently all changes in the C–O compliance during s-coordination of CO are affected by the coordination center. As a rule of thumb free CO shows a C–O compliance of ˚ mdyn1, while metal coordination weakens C–O by 11% 0.05 A ˚ mdyn1. to 0.06 A

Table 6

5c–W‡ a

Phosphaketenes72 R–P]C]O (1) release CO either upon irradiation73 (R ¼ tBu, Ar) or, as more recently established, during Pd-catalyzed formation of the corresponding diphosphene R–P]P–R (R ¼ Ar);74 for the dissociation of HPCO into free phosphinidene 3HP and carbon monoxide only 85 kJ mol1 are needed.75 In solution the phosphaketene is in equilibrium with its dimer, the C2h-symmetrical diphosphetanedione (R–PCO)2 (5a); e.g. Ar–P]C]O. Since H–P]C]O is a potential candidate for being radio-detectable in space76 the singlet HPCO hypersurface has been intensively studied: phosphaketene is more stable than its cyclic oxaphosphiran-3-ylidene isomer by 51.8 kcal mol1.76b Dimerization of phosphaketene to yield diphosphetanedione was found to be only slightly exothermic (0.66 kcal mol1).77 MO analysis of the PCO group (and other main group mono-carbonyls) and phosphaketene were published by Bridgeman78 and Hegarty75 respectively. Reaction of (2,4,6-tBu3C6H2)PCO with [WCl2(CO)(PMePh2)4] yielded a tungsten complex [WCl2(PMePh2)2](^PAr)] which was characterized as a linear terminal phosphinidene complex (WPC bond angle 168.2(2) ; d(31P) ¼ 193, 1JPW ¼ 649 Hz).79 Here, the full compliance matrix of H–P]C]O, its cyclic isomer H–P(m-CO) and three isomeric dimers (H–PCO)2 (C2h, Cs,C2) are calculated using coupled cluster and DFT methods to compare results from single-Slater determinant methods such as B3YLP and M05-2X with higher level coupled cluster calculation (Table 3). Remarkably, compliance constants of the phosphorus– carbonyl contact in phosphaketene (Table 4) are almost identical at coupled cluster and at DFT level. Referenced against

˚ mdyn1] of TSs for reactions (1) and (2) Computed (M05-2X/6-311G(d,p), LanL2DZ) compliance constants [A

11c–Cr‡ 11c–W‡ 5c–Mo‡

Phosphaketene H–P]C]O (phospha-isocyanate), its isomer H–P(h2-CO) and cyclic dimers (H–P]C]O)2

(A) (B) (A) (B)

Compound

P–CO

PC–O

P–M

(A)P–C(B)

C1-[Cr(CO)5PMe]‡ C1-[W(CO)5PMe]‡

8.303 22.647 9.222a 11.905b 0.430 0.428 0.426 0.440

0.048 0.048 0.046a 0.048b 0.059 0.107 0.058 0.103

0.521 0.446 0.291a 0.507b 1.660 1.083 1.007 0.917

5.5/4.9 5.5/4.9 3.4/3.0 3.4/3.0

C1-[Mo(CO)5(P(CO)Me)]2‡ C1-[W(CO)5(P(CO)Me)]2‡

M06-2X/6-311G(d,p), Lanl2DZ. b M06-2X/def2-TZVPecp.

This journal is ª The Royal Society of Chemistry 2013

Chem. Sci., 2013, 4, 4309–4322 | 4313

View Article Online

Chemical Science

Edge Article

Table 7 Reaction energies [kJ mol1] of selected reactions for phosphaketene complexes 1c. Values for the TS in parenthesis

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

(1)

Reaction

Method

DEZPE

DG

11–Cr + CO / 1c–Cr

B3LYP B3LYPa B3LYP-D3a M05-2Xb

M06a PWBP95-D3c

77 88 98 78 (72) 95 87 90 96 105 92 106 103 77 85 95 76 (67) 97 114

152 186 221 162 (69) 230 176 201 224 247 212 277 239 142 173 205 143 (64) 238 284

B3LYP B3LYPa B3LYP-D3a M05-2X M06a PWBP95-D3c B3LYP B3LYPa B3LYP-D3a M05-2X M06a PWBP95-D3c B3LYP B3LYPa B3LYP-D3a M05-2X M06a PWBP95-D3c

26 13 6 7 10 22 19 16 9 2 7 10 15 11 1 12 11 16

90 23 45 16 50 71 53 38 54 25 41 24 35 9 15 70 61 50

11–Mo + CO / 1c–Mo

11–W + CO / 1c–W

(2)

11–Cr + CO / MeP + Cr(CO)6

11–Mo + CO / MeP + Mo(CO)6

11–W + CO / MeP + W(CO)6

M06a PWBP95-D3c B3LYP B3LYPa B3LYP-D3a M05-2X M06a PWBP95-D3c B3LYP B3LYPa B3LYP-D3a M05-2X

(3)

1c–W/ 1c–Wi

B3LYP

0 (17)

0 (16)

(4)

2X 1c–W / 5c

B3LYP

89 (59) 129 137 107 131

30 (114) 58 79 40 59

B3LYP-D3a M05-2X M06a PWBP95-D3c

a def2-TZVP basis set. b BSSE estimation: 6–10 kJ mol1. c def2-TZVPP basis set and using the geometries obtained at B3LYP-D3/def2-TZVP.

model compounds H2C]PH and CH3–PH2, the P–CO bond strength of HPCO is to be described as in between a single and a double bond. The P–CO coordinate is strongly coupled (negative sign) to the HPC bending coordinate. In case of singletHP(h2-CO), an unusual carbene belonging to the class of oxaphosphiranes, the multireference character of the molecule

4314 | Chem. Sci., 2013, 4, 4309–4322

Fig. 4

Energy profile for the P/CO bond dissociation in 11c–W.

Fig. 5 Computed (M06-2X/def2-TZVPecp) structure for the initial van der Waals complex in the approach of CO to the P atom in 11–W, showing NCI (noncovalent interactions) RDG (reduced density gradient) isosurfaces (0.35isoval) coloured over the range 0.1 < sin(l2)r < 0.1 au: blue denotes strong attraction, green stands for moderate interaction, and red (not present) would indicate strong repulsion.83

˚ mdyn1] of 1c (M ¼ Cr, Table 8 Metal dependence of compliance constants [A Mo, W)a

Compound

P–CO

PC–O

P–M

1c–Cr

C1-[Cr(CO)5(P(CO)Me)]

1c–Mo 1c–W 1c–Cr

C1-[Mo(CO)5(P(CO)Me)] C1-[W(CO)5(P(CO)Me)] Cs-[Cr(CO)5(P(CO)Me)]

1c–Mo 1c–W

Cs-[Mo(CO)5(P(CO)Me)] Cs-[Mo(CO)5(P(CO)Me)]

0.291 0.295b 0.286 0.301 0.209 0.209b 0.210 0.209

0.059 0.059b 0.059 0.059 0.062 0.062b 0.062 0.062

1.722 1.750b 1.668 1.355 1.356 1.329b 1.328 1.055

a

B3LYP/6-311G(d,p), Lanl2DZ. b B3LYP/def2-TZVP.

inevitably shows the limit of the DFT approach (Table 4). While the coupled cluster compliances indicate a weak P–CO contact B3LYP gives the correct trend but a strongly deviating compliance constant; the same trend holds for the P–O contact. By comparison (Table 5), the dimer C2h (HPCO)2 shows slightly stronger bonds, but which are still signicantly weaker than in the reference P–C single bond in CH3–PH2 (Table 1). This journal is ª The Royal Society of Chemistry 2013

View Article Online

Edge Article

Chemical Science

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

˚ mdyn1] of phosphaketene 1a and Table 9 DFT compliance constants [A a complexes

1a 1b 1b 1c–W 1c–W 1d 1e 1f a

Compound

P–CO

PC–O

PB

P–M

Cs-P(CO)Me C1-BH3P(CO)Me Cs-[BH3P(CO)Me] C1-[W(CO)5(P(CO)Me)] Cs-[W(CO)5(P(CO)Me)] Cs-(BH3)2P(CO)Me C1-[W(CO)5(BH3)P(CO)Me)] C1-[{W(CO)5}2P(CO)Me]

0.223 0.322 0.202 0.301 0.209 0.649 0.589 0.488

0.062 0.058 0.061 0.059 0.062 0.053 0.056 0.057

— 1.494 1.749 — — 0.941 1.027 —

— — — 1.355 1.055 — 0.988 1.169

Fig. 7

˚ mdyn1] of 1a–f. Correlation of d(31P) [ppm] and compliance constants [A

B3LYP/6-311G(d,p), Lanl2DZ.

Fig. 8 CO end-on adduct 8 of the cationic aminophosphinidene complex [Cp*(CO)3W(PNiPr2)]+ (7).

Fig. 6 C1-TS of dimerization of phosphaketene complexes [W(CO)5P(CO)Me] (1c); unit A (left), unit B (right).

Table 10 DFT computed d(31P) chemical shiftsa and 1J(P,M)/1J(P,B) coupling constantsb of 1a–f

1a 1b 1b 1c–Cr 1c–Mo 1c–W 1c–Cr 1c–Mo 1c–W 1d 1e 1f a

a

d(31P)

1

Cs-P(CO)Me C1-BH3P(CO)Me Cs-[BH3P(CO)Me] C1-[Cr(CO)5(P(CO)Me)] C1-[Mo(CO)5(P(CO)Me)] C1-[W(CO)5(P(CO)Me)] Cs-[Cr(CO)5(P(CO)Me)] Cs-[Mo(CO)5(P(CO)Me)] Cs-[W(CO)5(P(CO)Me)] Cs-(BH3)2P(CO)Me C1-[W(CO)5(BH3)P(CO)Me)] C1-[{W(CO)5}2P(CO)Me]

244 149 186 135 173 187 193 214 236 8 34 100

— — — — 47 71 — 101 184 — 120 102

J(P,M)

1

— 33 62 — — — — — — 18 20 —

DFT compliance constantsa of complexes 7 and 8

Compound

P–CO

PC–O

P–M

P–N

[Cp*CO3W(PNiPr2)]+ [Cp*CO3W(P(CO)NiPr2)]+

— 1.575

— 0.058

0.718 1.088

0.200 0.278

˚ mdyn1. A

This journal is ª The Royal Society of Chemistry 2013

Hereaer, results are presented and discussed following the general idea of describing the P–C bond of a disturbed “CO unit” while going “from unusual/non-classical to common structures of P-ligands” in their transition-metal and borane complexes.

J(P,B)

In ppm. b In Hz.

Table 11

7 8

Compound

The quest for CO binding and activation

Reaction free enthalpies and transition states End-on addition of CO to neutral terminal phosphinidene complexes [M(CO)5PMe] (11; M ¼ Cr, Mo, W; Tables 6 and 7, reaction “(1)”) proceeds in all cases exergonically. The TS could only be located in case of 11c–Cr and 11c–W using Truhlars functionals M05-2X or M06-2X. For 11c–W the energy prole for the stepwise dissociation of CO was performed for a variety of functionals and basis sets (Fig. 4). The calculations reveal a rather at potential energy surface. Both M06/def2-TZVPecp and B3LYP-D3/def2-TZVPecp levels show similar dissociation energies with barrierless processes. On using the latter stepwise dissociation geometries (B3LYP-D3 functional) and computing energies at the most accurate PWBP95-D3/def2TZVPPecp level, a similar barrierless prole is obtained, but with larger dissociation energy. It should be noted, however, that the TS structure that could be obtained here can only be a rough estimate since, in case of low and at barriers, the TS might be identied with the free-energy maximum along the reaction path obtained by a variational search rather than with the PES saddle point.82 On the contrary, the M06-2X/def2-TZVPecp level appropriately describes the formation of a van der Waals complex at a

Chem. Sci., 2013, 4, 4309–4322 | 4315

View Article Online

Chemical Science

Edge Article

˚ mdyn1) and chemical shifts d(31P) data (ppm) of E–CO groups (E ¼ N, P) in Table 12 Computed harmonic frequencies (n/cm1), compliance constants (C/A [FeCp*CO2L] species at B3LYP/6-311g(d,p) level)

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

C L

n PC–O

EC–O

E–CO

E–CO/EC–O

E–CO/Fe–CC

E–CO/E–B

d(31P)

C(O)NH2 C(O)(BH3)NH2 C(O)PH2 C(O)(BH3)PH2 C(O)PtBu(SiMe3)a C(O)(BH3)PtBu(SiMe3)

1688 1753 1708 1673 1676 1673

0.093 0.085 0.088 0.089 0.091 0.093

0.212 0.446 0.700 0.808 0.759 0.839

0.024 0.037 0.050 0.060 0.051 0.065

0.040 0.051 0.047 0.049 0.077 0.047

— 0.140 — 0.000 — 0.014

— — 57 18 121 100

a

[FeCp*CO2{C(O)P(SiMe3)tBu}] (d(31P) ¼ 75.1, n((P)CO) ¼ 1620 cm1.91

Fig. 9

to afford the (most stable) triplet uncomplexed phosphinidene and the homoleptic metal carbonyl (entry 2) is markedly unfavoured and even systematically endergonic at the highest level. The barrier of inversion at phosphorus (entry 3) was calculated to be very small (16 kJ mol1) for 1c–W; hence a uxional behaviour is to be expected for all three phosphaketene complexes 1c–Cr/Mo/W. Within the phosphaketene metal complex series 1c–Cr to 1c–W compliance constants (Table 8) reveal a weakening of the P–CO bond strength in going from 1a to mono-ligation as in [(OC)5W{P(CO)Me}] (1c–W). Apparently, the absence of a lone pair at phosphorus (due to the coordinative bonding to tungsten) drastically weakens the P–CO bond and strengthens the C–O bond. This result clearly corroborates the lack of backdonating ability of phosphorus which became even more prominent by the results on the borane complex 1b (Table 9) and the dinuclear complexes 1d–f. In the complex [(OC)5W {P(CO)Me}] (1c–W) the CO activation through phosphorus, measured as weakening of the C–O bond strength, equals that of the CO activation through tungsten in pentacarbonyltungsten complexes. The dimerization of 1c–W to yield 1,3-diphosphetane-2,4dione complex 5c is exergonical by 30 (B3LYP) to 79 (M05-2X) kJ mol1 and proceeds with a barrier of 114 kJ mol1 (B3LYP) (Table 7, entry 4). The early C1-TS of dimerization (Fig. 6) corresponds with the [2p + 2p] pathway found by Bachrach.84 Its ˚ mdyn1) associated with the P–CO compliances (3.4 and 3.0 A P–C bonds formation are all positive and very large indicating a loosely bound TS. The [W(CO)5P(CO)Me] units A and B mainly differ in the PCO angle of 168 (A) and 135 (B), respectively. The CO unit of B is substantially weakened upon attack of phosphaketene complex A.

Borane adduct of phosphinocarbonyl complex [FeCp*CO2{C(O)PH2}] (9).

˚ (Fig. 5) and an early end-on addition TS P/CO distance of 3.094 A ˚ at P/CO 2.714 A. ˚ mdyn1 (highest level A P/CO compliance as large as 12 A in Table 6) indicates a loosely bound CO in the TS. The reaction free energy barrier amounts to only 30.9 kJ mol1 for the addition process from the separated reagents 11–W and CO (7.7 kJ mol1 from the 11–W$CO van der Waals complex), being of 87.6 kJ mol1 for the inverse phosphaketene dissociation process. Similar qualitative behaviour, although with lower dissociation energy, is observed on using a smaller basis set for both M06-2X and M05-2X. The energetics for a number of processes involving the formation or reactions of phosphaketene complexes 1c are collected in Table 7. The addition of carbon monoxide to 1a affording 1c (entry 1) has exergonic character varying in the order W > Mo > Cr at the highest (PWBP95-D3/def2-TZVPPecp) level. The largest discrepancies are observed in the Gibbs free energies for chromium and tungsten when using the smallest basis sets. In comparison, the alternative substitution reaction

Table 13

4a 4a0 4b

˚ mdyn1] of diphospha-ureas DFT computed chemical shifts d(31P) [ppm], coupling constants 1J(PB)/2J(PP) [Hz] and compliance constants [A

Compound

d(31P)

1

C1-(PMe2)2CO syn axial lp C2-(PMe2)2CO anti axial lp C2-(BH3PMe2)2CO anti axial lp

1, 32 16 63

— — 1

4316 | Chem. Sci., 2013, 4, 4309–4322

J(PB)

2

J(PP)

144 14 78

P–CO

PC–O

P–B

0.517 0.551 0.530

0.087 0.089 0.083

— — 0.620

This journal is ª The Royal Society of Chemistry 2013

View Article Online

Edge Article

Chemical Science

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

Table 14 DFT computed chemical shifts d(31P)a, coupling constants 1J(P,W)/1J(P,P)b and compliance constantsc of 1,3- and 1,2-diphosphetane-2,4-diones (M ¼ B, W(CO)5; n ¼ 1,2)

5a 5b 5c 6a 6b 6c a

Compound

d(31P)

1

C2h-(P(CO)Me)2 C2h-[BH3(P(CO)Me)]2 C2h-[W(CO)5(P(CO)Me)]2 C2-(P(CO)Me)2 C2-[BH3(P(CO)Me)]2 C2-[W(CO)5(P(CO)Me)]2 [W(CO)5(P(CO)Me)]2

75 155 130 47 94 86 131, 53 (PW)

— 24 152 — 27 158 153

J(P,M)

P–CO

PC–O

P–P

P–M

53 185 171 — 182 175 224

0.640 0.576 0.593 0.628 0.596 0.592

0.082 0.076 0.077 0.078 0.075 0.075

— — — 0.614 0.541 0.669

— 0.765 0.858 — 0.753 0.897

J(P,P)

˚ mdyn1. In ppm. b In Hz. c In A

31

P NMR chemical shis, 31P,183W, 31P,31P and31P,11B spin– spin coupling constants and compliance constants In the transition-metal pentacarbonyl complexes 1c (M ¼ Cr, Mo, W) of phosphaketene 1a, the phosphorus becomes more shielded in the direction 1c–Cr < 1c–Mo < 1c–W (Table 10). This trend also holds for the Cs-symmetrical TSs of inversion at phosphorus 1c–Cr/Mo/W, where phosphorus is more shielded than in 1c–Cr/Mo/W by 43 (Cr), 24 (Mo) and 26% (W). The phosphorus to metal couplings roughly double in going from 1c to TS 1c. The largest shielding of phosphorus within the scope of this investigation is calculated for methylphosphaketene (1a) at 244 ppm (Table 11), which compares quite well with the experimentally determined chemical shis of tBuPCO (d(31P) ¼ 180 ppm)85 and 2,4,6-tBu-C6H2PCO (d(31P) ¼ 207.4 ppm).86 Calculations predict that successive complexation of the P-lone pairs in MePCO (1a) by borane results in a downeld shi to 149 ppm in monoborane (BH3)P(CO)Me (1b) and to 8 ppm in bisborane phosphaketene (BH3)2P(CO)Me (1d). Planarization of phosphorus at the TS of inversion at P in 1b leads to an upeld shi to 186 ppm relative to 1b. Similarly upon borane complexation |1J(P,B)| drops from 33 Hz in 1b to 18 Hz in 1d. The largest P–B coupling however is calculated for the TS of inversion at phosphorus (1b). Note that in this investigation all 1J(P,B) couplings are found to have a negative sign, which is in contradiction with an early experimental investigation.87 Nevertheless, it will not be discussed further as this is a somewhat minor aspect within the current context. The correlation of phosphorus shielding and P–CO compliance constants is illustrated in Fig. 7. According to this, one may note that the more shielded the phosphorus the less compliant is the P–CO and the more compliant the PC–O bond. A rule of thumb for the correlation of P–CO compliances and the 31P

Table 15

3a 3b 3c a

n

NMR chemical shis can be given: the more shielded the phosphorus the smaller the P–CO compliance constant and vice versa. It is worth noting here that the quite satisfactory correlation shown by the systems under consideration spans over a wide range of P–C bond strengths, including species that display moderate chemical bonds or even weak interactions. This fact together with the aforementioned high compliances found in loosely bound TSs is in contrast to the Pulay and Baker criticism against the diagnostic character of compliance constants for weak interatomic interactions.51 All shieldings and coupling constants calculated so far are interpretable in terms of a simple “lone pair rule”: every lone pair present on phosphorus has a negative contribution to the nuclear spin–spin couplings and complexation by borane or a transition-metal pentacarbonyl removes this contribution. The effect on the coupling constant depends on its sign: a positive coupling constant is enhanced, a negative one becomes more positive (smaller in its absolute value). Some experimental results on CO adduct formation are available so far, allowing comparison with calculated data. Thus, the cationic terminal aminophosphinidene complex [Cp*(CO)3W(PNiPr2)]+ (7) displays a phosphorus resonance at 939 ppm,88 in close agreement with the computed value (941 ppm), which is upeld shied to 911 ppm on P-end-on ligation by CO affording the adduct 8 (Fig. 8). Similarly, Scheer observed an upeld shi of 955 ppm when the acetonitrile adduct [Cp* {(CO)5W}2P(NCMe)] was formed in solution.16 At the same time, the 1J(P,W) coupling enhances from 7 (7) to 131 Hz (8). Formation of P–CO removes the coplanarity of the WPNiPr-unit (:WPNC(iPr): 154 (8), 180 (2a)) and leads to a pyramidalized phosphorus – the sum of bond angles at P in 8 amounts to 345 . Both W–P and P–N bonds are signicantly ˚ (7), 2.59 A ˚ (8); P–N: 1.65 A ˚ (7), 1.68 A ˚ elongated in 8 (W–P: 2.49 A

DFT computed chemical shifts d(31P)a, coupling constants 1J(P,W)/1J(P,P)b and compliance constantsc of 1,2-diphosphiranones 3a–c (M ¼ B, W(CO)5)

Compound

d(31P)

1

C2-(PMe)2CO C2-[(BH3PMe)2(CO))] C2-[{W(CO)5}2((PMe)2(CO))]

195 65 98

— 44 155

J(P,M)

n

P–CO

PC–O

P–P

P–M

36 92 48

0.540 0.570 0.568

0.076 0.074 0.074

0.654 0.553 0.699

— 0.837 0.904

J(P,P)

˚ mdyn1. In ppm. b In Hz. c In A

This journal is ª The Royal Society of Chemistry 2013

Chem. Sci., 2013, 4, 4309–4322 | 4317

View Article Online

Chemical Science Table 16

2a 2b 2c 2c0

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

a

Edge Article

DFT computed chemical shifts d(31P)a, coupling constants 1J(P,M)/1J(P,P)b and compliance constantsc of diphosphiren-3-ones 2a–c (M ¼ B, W(CO)5)

Compound

d(31P)

1

C2v-P2CO C2-(BH3P)2CO C2-[{W(CO)5}2(P2CO)] C1-[{W(CO)5}2(P2CO)]

1238 583 465 406 (PW)

— 85 130

J(P,M)

n

P–CO

PC–O

P–P

P–M

228 217 201

0.502 0.913 0.598 0.329

0.077 0.072 0.076 0.058

0.329 0.281 0.324 0.553

— 1.271 0.773 0.676 1.079d

J(P,P)

˚ mdyn1. d PCO. In ppm. b In Hz. c In A

(8)), which is interpretable as both removal of tungsten-tophosphorus p-back-donation and nitrogen-to-phosphorus p-donation. Similarly, the W–P and P–N bond compliances are enhanced by 34% and by 28%, respectively (Table 11). The (P)C– ˚ mdyn1) corresponds more with the O compliance in 8 (0.058 A ˚ C]O double bond of HPCO ((P)C–O compliance: 0.060 A mdyn1) than with a purely s-donating CO as in H3BCO (C–O ˚ mdyn1). However, the long P–CO contact compliance: 0.052 A is associated with an extremely large compliance constant (1.58 ˚ mdyn1). Although this value might hint at a DFT error A (probably due to the multireference character of the problem) the trend is apparent: the P–CO bond in 8 has more the character of a loosely bound adduct than of a single bond, the compliance of which for a P–C reference compound such as ˚ mdyn1 (Table 1). Note that this cannot be H2P–CH3 is 0.39 A ˚ which is concluded from the P–CO bond length of 8 (1.83 A), ˚ In this case the shorter than P–C in H2P–CH3 (1.87 A). CO adduct formation is thermodynamically disfavoured by 87 kJ mol1 (DG). The P–C(O) distance in the phosphacarbamoyl89 group C(O) PR2 in complex [RuCp*CO2{C(O)P(SiMe3)tBu}]90 is known to be ˚ compared with a P–C single bond. In extremely long (1.952(3) A) the model compound [FeCp*CO2{C(O)PH2}] (9) the P–CO ˚ mdyn1) by far exceeds that of an ordinary compliance (0.7 A ˚ mdyn1, for CH3PH2W(CO)5) pointing to a single bond (0.35 A weak P–C interaction. Moreover, borane complexation of the lone pair at phosphorus (Table 12 and Fig. 9) further weakens the P–CO contact. E–CO negatively couples to EC–O and Fe–CO (E ¼ N, P): if E–CO is stretched both other coordinates will shorten (the bond strengthens).

Compliance constants of acyclic and cyclic phosphoorganic compounds possessing two phosphorus atoms bound to CO At the end of this comparative study the DFT NMR parameters as well as the compliance constants of more common phosphaorganic compounds shall be briey discussed, whereby the starting point is the case of diphospha-ureas 4a,b (Table 14). Thereaer four-membered rings (1,3- and 1,2-diphosphetane2,4-diones 5a–c and 6a–c) and three-membered rings (2a–c and 3a–c) will be discussed groupwise. The P–CO compliance of 4a exceeds that of an “ordinary P–C ˚ mdyn1) which gets larger (¼the P–C bond single bond” (0.35 A weaker) upon borane complexation of the lone pair at phosphorus (Table 13). For the sake of comparison, the reported

4318 | Chem. Sci., 2013, 4, 4309–4322

diphospha-urea (Ph3CPH)2CO displays resonances at d(31P) ¼ 16.8 ppm, |2J(PP)| ¼ 127 Hz for the conformer with syn, axial lone pairs (lp) at P, and 23.6 ppm, |2J(PP)| ¼ 23 Hz for the anti, axial lp conformer.92 In contrast, 5a and 6a experience a P–C bond strengthening upon complexation (Table 14). This could be interpreted in terms of roughly non-hybridized p-type atomic orbitals (AO) used by phosphorus in the P–C bond in the uncomplexed species 5a, in agreement with the result of the NBO (natural bond orbital) analysis pointing to 85.3% of p-character at P. Upon complexation the tetracovalency of the P atom requires sp3-hybridization, as indicated by the (just moderately) higher s-character of the involved AO at P (18.5% for 5b and 17.5% for 5c). The case of diphosphiranones 3 (Table 15) is special as they are known as ligands in complexes of the type L2M{(RP)2CO} (M ¼ Pd, Pt; R ¼ Mes).74,93 As already seen before in the cases of 5 and 6 the P–CO compliance values are quite large and only slightly decrease in their borane and tungsten complexes. The last class of compounds investigated herein is the rather unique case of the (experimentally and theoretically unknown) diphosphiren-3-one 2a and their complexes 2b,c. Stable end-on 3H-diphosphirene tungsten complexes were described by Bertrand et al.94 From Table 16 the highly unusual chemical shis of 2a–c are most remarkable and point to an unusual bonding between phosphorus and carbon atoms. The P–CO compliance of 2a falls within the range of 4–695 and increases signicantly to ˚ mdyn1 upon borane complexation of the lone pair at 0.913 A phosphorus, thus resulting in a loose side-on adduct between the bis-borane P2 complex (H3B(P2)BH3) and CO. It is worth mentioning that in 2a the energetically most important interactions (upon second-order perturbation theory analysis of Fock matrix in NBO basis) come from electron donation p(P]P) / p*(C]O) (amounting to 69.8 kcal mol1) and from s(P–P) / s*(C–O) (19.9 kcal mol1). Further support for a view of 2a as a side-on (P]P)/(C]O) complex is the high positive (natural) charge at the P2 unit (0.570 au) with the negative charge located almost exclusively at the exocyclic O atom (0.507 au). This is in sharp contrast to its aza-analogue, 3-oxodiazirene 2aN, displaying an almost electroneutral N2 unit (0.156 au) with a typically electrophilic C atom (0.647 au). The loose character of the interaction of the CO and P2 units in 2a is further evidenced not only by the long P–C bond distance featuring low values for the ˚ r(r) ¼ associated bond strength parameters (dC–P ¼ 1.813 A, 3 0.1574 e/a0 ), but also by the large distance between the C atom

This journal is ª The Royal Society of Chemistry 2013

View Article Online

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

Edge Article

Chemical Science

˚ 60.5% and the P–C bond critical point (BCP) (dC–BCP ¼ 1.097 A, the C–P bond distance). This indicates an anomalous type of ˚ 39.8% bonding with respect to the 2aN case (dC–BCP ¼ 0.549 A, the C–N bond distance) that features a stronger C–N bond95 and follows the reported tendency of BCP shiing towards the electronegative center or group (CO in this case).96 Similarly, the ring critical point (RCP) in 2a was found to be farther away from ˚ 75.5% the distance between C and the C atom (dC–RCP ¼ 1.120 A, ˚ 75.5% the the P–P midpoint) that in 2aN case (dC–RCP ¼ 1.120 A, distance between C and the N–N midpoint).

leads to a drastic P–CO bond weakening as expressed through the compliance constant, i.e. 0.913 in case of the C2-symmetrical (H3BP)2CO (2b). This marks an extreme in weak P–CO bonding interactions. The calculation of 31P NMR shis and scalar 1J(P,B) and 1J(P,B) couplings enable a correlation with P– CO and PC–O compliance constants and led to an essential rule: the more shielded the phosphorus the smaller the P–CO compliance constant and vice versa.

Conclusions

We thank the Deutsche Forschungsgemeinscha (STR 411/ 25-2), the SFB 813 “Chemistry at Spin Centers”, and the Cost action cm0802 “PhoSciNet” for nancial support.

This comparative study on the quest for CO activation by phosphorus centres being in various bonding environments provides rst time-ever quantiable results. The full compliance matrices at coupled cluster level of phosphaketene HP]C]O, its isomer HP(h2-CO) (¼ the singlet oxaphosphirane-3-ylidene) and the dimer (HP]C]O)2 as well as the reference compounds phosphaacetylene P^CH, phosphaethene HP]CH2 and methylphosphane H2P–CH3 provide the data background. Phosphaketene has a low value for the compliance ˚ mdyn1], i.e. high stiffness. corresponding to the constant [A P–CO bond (0.220), which increases in HP(h2-CO) (0.635) and in the dimer (HP]C]O)2 (0.503), thus revealing signicantly weakened P–CO single bonds in the two latter cases. The electrophilic terminal phosphinidene complexes [(CO)5MPR] (M ¼ Cr, Mo, W; R ¼ Me) add exergonically carbon monoxide (in silico) to form phosphaketene complexes [(CO)5M{P(CO)R}]. Upon ligation, the P–CO compliance constant of 1c–W increased (0.301), this effect being even more pronounced for the related BH3 complex (0.322), thus indicating a weakly Pbound CO. Remarkable is that the P–CO bond is even stronger in the TS of inversion at phosphorus (TS–1c) than in the corresponding free phosphaketene. End-on addition of CO to dinuclear phosphinidene complexes [{(CO)5W}2PR] is predicted to proceed endergonically and the compliance constant of P–CO increases even more (0.488) to result in an even weaker P–C bond; again, this becomes more pronounced in the bis-BH3 complex (0.649). In total, presence (or absence through kP-BH3 and/or kP-W(CO)5 complexation) of electron density at phosphorus available for P–C bonding drastically changes the nature of the P–CO contact in compounds 1a–f and, hence, different degrees of CO activation result: the amount of back-donation from phosphorus to CO is crucial for the activation. Notably, the degree of CO activation in phosphaketene complexes [(CO)5W {P(CO)R}] (via phosphorus) equals that in carbonyl metal complexes (via the metal). In addition, some conclusions can be drawn for small ring P-heterocycles bearing (formally) a carbonyl unit: the compliance constants of three- (2) and fourmembered rings (5, 6) having one (2, 4) or two (5, 6) P-bound keto functionalities were found to be in the range of 0.502– 0.640, thus revealing that the strengths of these P–CO bonds are well below the P–C single bond of H2P–CH3. This nding provides a rst understanding of the experimentally known ease of thermal CO extrusion from such compounds. Of particular interest is the case of diphosphirene-3-ones 2 as P-ligation here

This journal is ª The Royal Society of Chemistry 2013

Acknowledgements

Notes and references 1 P. P. Power, Nature, 2010, 463, 171–177. 2 (a) G. C. Welch and D. W. Stephan, J. Am. Chem. Soc., 2007, 129, 1880–1881; (b) G. C. Welch, R. R. San Juan, J. D. Masuda and D. W. Stephan, Science, 2006, 314, 1124– 1126. 3 G. D. Frey, V. Lavallo, B. Donnadieu, W. W. Schoeller and G. Bertrand, Science, 2007, 316, 439–441. 4 V. Lavallo, Y. Canac, B. Donnadieu, W. W. Schoeller and G. Bertrand, Angew. Chem., Int. Ed., 2006, 45, 3488–3491. 5 Reviews: (a) F. Mathey, N. H. Tran Huy and A. Marinetti, Helv. Chim. Acta, 2001, 84, 2938–2957; (b) K. Lammertsma, Top. Curr. Chem., 2003, 229, 95–119; (c) F. Mathey, Dalton Trans., 2007, 1861–1868. 6 M. P. Duffy, L. Y. Ting, L. Nicholls, Y. Li, R. Ganguly and F. Mathey, Organometallics, 2012, 31, 2936–2939. 7 For the generation of a transient end-on phosphaketene tungsten complex formed via [4 + 2] cycloreversion, see: Y. Mao, Z. Wang, R. Ganguly and F. Mathey, Organometallics, 2012, 31, 4786–4790. 8 (a) R. Streubel, H. Wilkens, A. Ostrowski, C. Neumann, F. Ruthe and P. G. Jones, Angew. Chem., Int. Ed. Engl., 1997, 36, 1492–1493; (b) R. Streubel, Coord. Chem. Rev., 2002, 227, 175–192. 9 S. Grigoleit, A. Alijah, A. B. Rozhenko, R. Streubel and W. W. Schoeller, J. Organomet. Chem., 2002, 643–644, 223– 230. 10 (a) H. Helten, C. Neumann, A. Espinosa, P. G. Jones, M. Nieger and R. Streubel, Eur. J. Inorg. Chem., 2007, 4669– 4678; (b) H. Helten, S. Fankel, O. Feier-Iova, M. Nieger, A. Espinosa and R. Streubel, Eur. J. Inorg. Chem., 2009, 3226–3237. 11 E. Ionescu, G. von Frantzius, P. G. Jones and R. Streubel, Organometallics, 2005, 24, 2237–2240. 12 C. Schulten, G. von Frantzius, A. Espinosa, G. Schnakenburg and R. Streubel, Chem. Sci., 2012, 3, 3526–3533. 13 [{LnM}2P(base)R]: H. Lang, G. Mohr, O. Scheidsteger and G. Huttner, Chem. Ber., 1985, 118, 574–596. 14 [{LnM}2Pn(thf)Cp*] (Pn ¼ P, As; M ¼ Cr, Mo, W): P. Jutzi and R. Kroos, J. Organomet. Chem., 1990, 390, 317–322.

Chem. Sci., 2013, 4, 4309–4322 | 4319

View Article Online

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

Chemical Science 15 Review: G. Huttner and K. Evertz, Acc. Chem. Res., 1986, 19, 406–413. 16 M. Schiffer and M. Scheer, Angew. Chem., Int. Ed., 2001, 40, 3413–3416. 17 (a) A. J. Arduengo III, J. C. Calabrese, A. H. Cowley, H. V. R. Dias, J. R. Goerlich, W. J. Marshall and B. Riegel, Inorg. Chem., 1997, 36, 2151–2158; (b) A. J. Arduengo III, C. J. Carmalt, J. A. C. Clyburne, A. H. Cowley and R. Pyati, Chem. Commun., 1997, 981–982; (c) Review: A. H. Cowley, J. Organomet. Chem., 2001, 617–618, 105–109. 18 O. Beck, M. Henry-Ellinger, C. D. Martin and G. Bertrand, Angew. Chem., Int. Ed., 2013, 52, 2939–2943. 19 (a) Y. Wang, Y. Xie, P. Wei, R. B. King, H. F. Schaefer III, P. v. R. Schleyer and G. H. Robinson, J. Am. Chem. Soc., 2008, 130, 14970–14971; (b) Y. Wang and G. H. Robinson, Inorg. Chem., 2011, 50, 12326–12337. 20 Y. Wang, Y. Xie, P. Wei, R. B. King, H. F. Schaefer III, P. v. R. Schleyer and G. H. Robinson, Science, 2008, 321, 1069–1071. 21 H. Braunschweig, R. D. Dewhurst, K. Hammond, J. Mies, K. Radacki and A. Vargas, Science, 2012, 336, 1420–1422. 22 For some selected examples of low-coordinate main group element adducts, see: (a) O. Back, M. A. Celik, G. Frenking, M. Melaimi, B. Donnadieu and G. Bertrand, J. Am. Chem. Soc., 2010, 132, 10262–10263; (b) A. C. Filippou, O. Chernov and G. Schnakenburg, Angew. Chem., Int. Ed., 2009, 48, 5687–5690; (c) R. S. Ghadwal, H. W. Roesky, S. Merkel, J. Henn and D. Stalke, Angew. Chem., Int. Ed., 2009, 48, 5683–5686; (d) M. Y. Abraham, Y. Wang, Y. Xie, P. Wei, H. F. Schaefer III, P. v. R. Schleyer and G. H. Robinson, Chem.–Eur. J., 2010, 16, 432–435; (e) Y. Wang, Y. Xie, M. Y. Abraham, R. J. Gilliard Jr, P. Wei, H. F. Schaefer III, P. v. R. Schleyer and G. H. Robinson, Organometallics, 2010, 29, 4778–4780; (f) For an anionic NHC, see: Y. Wang, Y. Xie, M. Y. Abraham, P. Wei, H. F. Schaefer III, P. v. R. Schleyer and G. H. Robinson, J. Am. Chem. Soc., 2010, 132, 14370–14372; (g) Y. Wang and G. H. Robinson, Dalton Trans., 2012, 41, 337–345; (h) A. P. Singh, P. P. Samuel, K. C. Mondal, H. W. Roesky, N. S. Sidhu and B. Dittrich, Organometallics, 2013, 32, 354– 357; (i) A. Higelin, S. Keller, C. G¨ ohringer, C. Jones and I. Krossing, Angew. Chem., 2013, 125, 5041–5044. 23 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery Jr, T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford,

4320 | Chem. Sci., 2013, 4, 4309–4322

Edge Article

24

25

26

27

28 29

30

31 32

33 34 35 36 37 38

39

J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, Gaussian 03, Revision B.02/B.03, Gaussian, Inc., Wallingford CT, 2004. ADF 2007.01, Theoretical Chemistry, Vrije Universiteit, Amsterdam, http://www.scm.com(a) E. J. Baerends, D. E. Ellis and P. Ros, Chem. Phys., 1973, 2, 41–51; (b) E. J. Baerends and P. Ros, Chem. Phys., 1973, 2, 52–59; (c) G. te Velde and E. J. Baerends, J. Comput. Phys., 1992, 99, 84–98. ORCA – An ab initio, DFT and semiempirical SCF-MO package, Written by F. Neese, Max Planck Institute for Bioinorganic Chemistry, D-45470 M¨ ulheim/Ruhr, 2012, Version 2.9.1, http://www.mpibac.mpg.de/bac/logins/neese/ description.phpF. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2, 73–78. (a) A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652; (b) C. T. Lee, W. T. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785–789. (a) R. Krishnan, J. S. Blinkley, R. Seeger and J. A. Pople, J. Chem. Phys., 1980, 72, 650–654; (b) P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 270–283. F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297–3305. D. Andrae, U. H¨ aussermann, M. Dolg, H. Stoll and H. Preuss, Theor. Chim. Acta, 1990, 77, 123–141. ECP parameters for Mo [SD(28, MWB)] and W [SD(60, MWB)] were obtained from the pseudopotential library of the Stuttgart/Cologne group, at http://www.theochem.uni-stuttgart.de/pseudopotentials/. M05-2X: (a) Y. Zhao and D. G. Truhlar, Org. Lett., 2006, 8, 5753–5755; (b) Y. Zhao, N. E. Schultz and D. G. Truhlar, J. Chem. Theory Comput., 2006, 2, 364–382. M06 and M06-2X: (c) Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2007, 120, 215–241. S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys., 2010, 132, 154104. (a) L. Goerigk and S. Grimme, J. Chem. Theory Comput., 2011, 7, 291–309; (b) L. Goerigk and S. Grimme, Phys. Chem. Chem. Phys., 2011, 13, 6670–6688. See the references in: M. Heckert, M. K´ allay, D. P. Tew, W. Klopper and J. Gauss, J. Chem. Phys., 2006, 125, 044108. J. Cizek, Adv. Chem. Phys., 1969, 14, 35. K. Raghavachari, G. W. Trucks, J. A. Pople and M. HeadGordon, Chem. Phys. Lett., 1989, 157, 479–483. D. E. Woon and T. H. Dunning, J. Chem. Phys., 1989, 90, 1007–1023. G. Fogarasi, X. Zhou, P. W. Taylor and P. Pulay, J. Am. Chem. Soc., 1992, 114, 8191–8201. W. H. Press, S. A. Teukolsky, W. T. Vetterling and B. P. Flannery, Numerical Recipes in Fortran: The Art of Scientic Computing, Cambridge University Press, Cambridge, UK, 1994. (a) G. Schreckenbach and T. Ziegler, J. Phys. Chem., 1995, 99, 606–611; (b) G. Schreckenbach and T. Ziegler, Int. J. Quantum Chem., 1997, 61, 899–918.

This journal is ª The Royal Society of Chemistry 2013

View Article Online

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

Edge Article 40 C. van Wuellen, Phys. Chem. Chem. Phys., 2000, 2, 2137– 2144. 41 (a) J. C. Decius, J. Chem. Phys., 1963, 38, 241–248; (b) L. H. Jones and R. R. Ryan, J. Chem. Phys., 1970, 52, 2003– 2004. 42 (a) J. Grunenberg and N. Goldberg, J. Am. Chem. Soc., 2000, 122, 6045–6047; (b) J. Grunenberg, R. Streubel, G. von Frantzius and W. Marten, J. Chem. Phys., 2003, 119, 165– 169; (c) J. Grunenberg, J. Am. Chem. Soc., 2004, 126, 16310– 16311. 43 (a) S. Katsyuba, R. Schmutzler and J. Grunenberg, Dalton Trans., 2005, 1701–1706; (b) K. Brandhorst and J. Grunenberg, Chem. Soc. Rev., 2008, 37, 1558–1567. 44 Y. Xie and H. F. Schaefer III, J. Phys. Chem., 2003, 217, 189– 203. 45 J. K. G. Watson, J. Mol. Struct., 2004, 695–696, 71–75. 46 E. B. Wilson, J. Chem. Phys., 1939, 7, 1047–1052. 47 D. Cremer, A. Wu, A. Larsson and E. Kraka, J. Mol. Model., 2000, 6, 396–412. 48 J. C. Decius, J. Chem. Phys., 1953, 21, 1121–1122. 49 L. H. Jones and B. I. Swanson, Acc. Chem. Res., 1976, 9, 128– 134. 50 (a) I. H. Williams, Chem. Phys. Lett., 1982, 88, 462–466; (b) I. H. Williams, Chem. Soc. Rev., 1993, 22, 277–283. 51 (a) J. Baker, J. Chem. Phys., 2006, 125, 014103; (b) J. Baker and P. Pulay, J. Am. Chem. Soc., 2006, 128, 11324–11325. 52 K. Brandhorst and J. Grunenberg, ChemPhysChem, 2007, 8, 1151–1156. 53 (a) J. Grunenberg, Angew. Chem., Int. Ed., 2001, 40, 4027– 4029; (b) P. R. Schreiner, H. P. Reisenauer, J. Romanski and G. Mloston, Angew. Chem., Int. Ed., 2009, 48, 8133– 8136; (c) S. Shaik, D. Danovich, W. Wu, P. Su, H. S. Rzepa and P. C. Hiberty, Nat. Chem., 2012, 4, 195–200. 54 (a) J. Grunenberg, G. von Frantzius, R. Streubel and K. Brandhorst, Organometallics, 2006, 25, 118–121; (b) A. G. Ponomareva, Y. P. Yurenko, R. O. Zhurakivsky, T. van Mourik and D. M. Hovorun, Phys. Chem. Chem. Phys., 2012, 14, 6787–6795. 55 (a) H. Helten, G. von Frantzius, G. Schnakenburg, J. Daniels and R. Streubel, Eur. J. Inorg. Chem., 2009, 2062–2065; (b) A. Espinosa and R. Streubel, Chem.–Eur. J., 2011, 17, 3166– 3178. 56 E. Kurita, Y. Tomonaga, S. Matsumoto, K. Ohno and H. Matsuura, THEOCHEM, 2003, 639, 53–67. 57 J. K. Tyler, J. Chem. Phys., 1964, 40, 1170–1171. 58 P. Dr´ ean, J. Demaison, L. Poteau and J. M. Denis, THEOCHEM, 2001, 571, 59–70. 59 G. Strey and I. M. Mills, Mol. Phys., 1973, 26, 129–138. 60 Semi-experimental, taken from: L. Margul` es, J. Demaison, P. B. Sreeja and J.-C. Guillemin, J. Mol. Spectrosc., 2006, 238, 234–240. 61 See the reference in: H.-W. Kim and D. Zeroka, J. Mol. Spectrosc., 2006, 238, 234–240. 62 J. W. C. Johns, H. F. Shurvell and J. K. Tyler, Can. J. Phys., 1969, 47, 893–920. 63 V. F. Kalasinsky and S. Pechsiri, J. Raman Spectrosc., 1985, 16, 190–192.

This journal is ª The Royal Society of Chemistry 2013

Chemical Science 64 B. Pellerin, P. Guenot and J.-M. Denis, Tetrahedron Lett., 1987, 28, 5811–5814. 65 A. S. Goldman and K. Krogh-Jespersen, J. Am. Chem. Soc., 1996, 118, 12159–12166. 66 A. K. Hughes and K. Wade, Coord. Chem. Rev., 2000, 197, 191–229. 67 A. C. Venkatachar, R. C. Taylor and R. L. Kuczkowski, J. Mol. Struct., 1977, 38, 17–23. 68 S. Erhardt and G. Frenking, Chem.–Eur. J., 2006, 12, 4620– 4629. 69 K. P. Huber and G. Herzberg, Constants of Diatomic Molecules, Van Nostrand Reinhold, New York, 1979. 70 A. C. Venkatachar, R. C. Taylor and R. L. Kuczkowski, J. Mol. Struct., 1977, 38, 17–23. 71 L. L. Jones, R. C. Taylor and R. T. Paine, J. Chem. Phys., 1979, 70, 749–757. 72 Reviews: (a) J. Escudi´ e, H. Ranaivonjatovo and L. Rigon, Chem. Rev., 2000, 100, 3639–3696; (b) R. Appel and F. Knoll, Adv. Inorg. Chem., 1989, 33, 259–361. 73 See references in: ref. 72b. 74 M.-A. David, D. K. Wicht, D. S. Glueck, G. P. A. Yap, L. M. Liable-Sands and A. L. Rheingold, Organometallics, 1997, 16, 4768–4770. 75 M. T. Nguyen, A. F. Hegarty, M. A. McGinn and P. Ruelle, J. Chem. Soc., Perkin Trans. 2, 1985, 1991–1997. 76 (a) C. Dimur, F. Pauzat, Y. Ellinger and G. Berthier, Spectrochim. Acta, Part A, 2001, 57, 859–873; (b) H.-G. Fu, H.-T. Yu, Y.-J. Chi, Z.-S. Li, X.-R. Huang and C.-C. Sun, Chem. Phys. Lett., 2002, 361, 62–70. 77 U. Salzner and S. M. Bachrach, J. Am. Chem. Soc., 1994, 116, 6850–6855. 78 A. J. Bridgeman, Inorg. Chim. Acta, 2001, 321, 27–41. 79 A. H. Cowley, B. Pellerin, J. L. Atwood and S. G. Bott, J. Am. Chem. Soc., 1990, 112, 6734–6735. 80 Ar-matrix: Z. Mielke and L. Andrews, Chem. Phys. Lett., 1991, 181, 355–360. 81 Compare the calculation: C. Dimur, F. Pauzat, Y. Ellinger and G. Berthier, Spectrochim. Acta, Part A, 2001, 57, 859– 873. 82 I. Williams, Chem. Soc. Rev., 1993, 22, 277–283. 83 NCIplot, version 1.1, Department of Chemistry, Duke University, USA, 2011, http://www.chem.duke.edu/~yang/ Soware/soware NCI.html (a) E. R. Johnson, S. Keinan, P. Mori-S´ anchez, J. Contreras Garc´ıa, A. J. Cohen and W. Yang, J. Am. Chem. Soc., 2010, 132, 6498–6506; (b) J. Contreras Garc´ıa, E. R. Johnson, S. Keinan, R. Chaudret, J.-P. Piquemal, D. N. Beratan and W. Yang, J. Chem. Theory Comput., 2011, 7, 625–632. Here, l2 stands for the second highest eigenvalue of the electron density Hessian matrix. 84 U. Salzner and S. M. Bachrach, J. Am. Chem. Soc., 1994, 116, 6850–6855. 85 R. Appel and W. Paulen, Tetrahedron Lett., 1983, 24, 2639– 2642. 86 R. Appel and W. Paulen, Angew. Chem., Int. Ed. Engl., 1983, 22, 785–786. 87 H. C. E. McFarlane, W. McFarlane and D. S. Rycro, J. Chem. Soc., Faraday Trans. 2, 1972, 68, 1300–1305.

Chem. Sci., 2013, 4, 4309–4322 | 4321

View Article Online

Published on 12 September 2013. Downloaded by Universidad de Murcia on 01/10/2013 08:41:54.

Chemical Science 88 B. Sterenberg, K. A. Udachin and A. J. Carty, Organometallics, 2001, 20, 2657–2659. 89 W. Petz and F. Weller, Z. Naturforsch., B: J. Chem. Sci., 2008, 51, 715–721. 90 L. Weber, K. Reizig and R. Boese, Organometallics, 1985, 4, 2097–2101. 91 L. Weber, K. Reizig and R. Boese, Chem. Ber., 1985, 118, 1193–1203. 92 V. Plack, J. R. Goerlich and R. Schmutzler, Z. Anorg. Allg. Chem., 1999, 625, 919–922. 93 M.-A. David, D. S. Glueck, G. P. A. Yap and A. L. Rheingold, Organometallics, 1995, 14, 4040–4042.

4322 | Chem. Sci., 2013, 4, 4309–4322

Edge Article 94 (a) D. Bourrissou and G. Bertrand, Top. Curr. Chem., 2002, 625, 919–922; (b) A transient 3H-diphosphirene derivative was reported in: E. Niecke, R. Streubel and M. Nieger, Angew. Chem., Int. Ed. Engl., 1991, 30, 90–91. 95 In the case of the diaza-analogue, 2aN, the C–N bond is found to be much stiffer as shown by its lower compliance ˚ mdyn1), which is, additionally, in constant (0.189 A agreement with the lower value displayed by the N–N bond ˚ mdyn1); dC–N ¼ 1.379 A, ˚ r(r) ¼ 0.3186 e/ao3). (0.130 A 96 (a) R. J. Boyd and K. E. Edgecombe, J. Am. Chem. Soc., 1988, 110, 4182–4186; (b) S. Hati and D. Datta, J. Comput. Chem., 1992, 13, 912–918.

This journal is ª The Royal Society of Chemistry 2013