вход по аккаунту


Boron-Phosphorus Compounds and Multiple Bonding.

код для вставкиСкачать
Boron-Phosphorus Compounds and Multiple Bonding
By PhiIip P. Power*
Boron-phosphorus compounds have not been as thoroughly studied as their boron-nitrogen
counterparts. Until recently many classes of B-P compounds that had been well established in
B-N chemistry were either unknown or poorly characterized. This statement is particularly
true for compounds involving possible multiple bonding between boron and phosphorus. For
example, detailed structural information on simple monomeric phosphino-boranes, R,BPR;,
did not become available until 1986 even though the isoelectronic S i c double bonded species,
the silenes, had already been reported. However, new work has shown that it is possible to
prepare and characterize several novel types of boron-phosphorus compounds with varying
degrees of multiple B-P bonding. These include not only monomeric phosphinoboranes but
also phosphanediylborates (borylphosphides), three- and four-membered rings (diphosphadiboretanes), boron phosphorus analogues of borazine, B-P skeletal analogues of ally1 cations
and anions, butadiene and cage compounds. Structural, spectroscopic (mainly NMR) and
theoretical studies reveal some important differences between B-P and B-N compounds which
in many cases can be traced to the presence of a high inversion barrier at phosphorus that
reduces the x interaction. This usually causes compounds such as R,BPR; to associate through
0 bonding between B and P. Supporting evidence for this view comes from species that involve
phosphorus and nitrogen in competitive x bonding with a boron p orbital in which the dative
interaction between B and N is dominant and the phosphorus center remains pyramidal.
Recently published work has shown that steric and electronic factors can be used to favor n
bonding and give an approximately planar =B-P:
system. Furthermore, theoretical studies
reveal that p p x overlap in a planar B-P system is of similar efficiency to its B-N analogue.
Good examples are seen in the phosphanediyl borates, the boron-phosphorus analogues of
borazine and the n-ally1 cations, whose molecular configurations and B-P bond lengths support strong boron-phosphorus x bonding.
1. Introduction
Since the discovery of borazine by Stock et al. in 1926"'
boron-nitrogen compounds have enjoyed continued popularity.['] The widespread interest in such compounds stems,
in large measure, from their isoelectronic relationship to carbon species. Interest in compounds having boron bonded to
the heavier main group 5 elements P,As,Sb or Bi has been
much slower to develop.[31One reason for this has been the
much greater tendency of simple units such as R,BPR, to
associate through donation of the lone pair on phosphorus
to an empty p orbital on boron. This usually results in highly
stable, generally cyclic, oligomeric, o-bonded species with
low reactivity in which boron and phosphorus are invariably
four-coordinate (e.g., l).L31Other species involving fourcoordination are the donor-acceptor adducts of formula
R,BPR;, 2, which also exhibit high thermal stability.
R. .R
R,/*' R
In contrast, well-characterized compounds involving
bonding between boron and phosphorus centers that are
three-coordinate have proved much more elusive. Unlike the
four-coordinate compounds the coordination number three
(or lower) allows for the possibility of x bonding between the
boron and phosphorus p orbitals and this review is primarily
concerned with compounds that involve varying degrees of
B-P x bonding.
Early cryoscopic measurements on the compounds 3-5
(and related derivatives) showed that they were monomeric
in benzene at concentrations of about one wt.%.[4*51In
addition it was concluded, on the basis of dipole moment
and infrared measurements, that phosphinodiarylboranes
were unassociated, not because of x bonding to phosphorus,
but because the boron center undergoes x bonding with its
aryl substituents. Prior work, using isopiestic molecular
weight measurements, had indicated that Me,BPH, was
monomeric in ether but polymerized over a period of days.[6J
Furthermore, the synthesis of the interesting cyclic diphosphadiboretane species fic5]was claimed but it was only characterized on the basis of molecular weight measurements.
and a
Other work postulated the existence of the species 7[71
mixture of compounds of formula 8.[*]None were sufficient-
['I Prof. P. P. Power
Department of Chemistry
University of California
Davis, California 95616 (USA)
Angew. C h e m . h i . Ed. Engl. 29 (1990) 449-460
0 VCH VeriagsgesellschaftmbH. 0-6940 Weinheim. 1990
ly characterized to enable structures to be assigned. In addition, it was reported that molecular orbital calculations indicated that analogues of borazine with P instead of N might
exist as stable compounds.[g1In view of these encouraging
results it is surprising to note that it was not until 1986 that
the first compound having B-P multiple bond character was
structurally characterized.[lol In fact, until recently there
were little structural or spectroscopic, i.e. 'H.' 'B and ,'P
NMR, data on any compounds involving bonding between
three-coordinate boron and phosphorus. Rare exceptions
were the compounds 9," 'I the first diphosphadiboretanes to
be spectroscopically characterized, and the three-membered
ring compound lO,[''] whose X-ray crystal structure (vide
infra) was the first such report for a compound involving
bonding between three-coordinate boron and phosphorus
centers. It is of significant interest that the last five years have
seen an exciting series of papers from a number of groups
that have described the first structurally characterized phosphanediylborates (also to be formulated as borylphosphide
(RBPSiMe,),, R = NMe,, Ph
Optimized Planar
1 2-1161 and
phadiboretanes, the B-P analogues of borazines (compounds
of type 34),1201a large number of heterocycles with boronphosphorus bonds in addition to other atoms (see Sections
4.3, 5.1, 6), monomeric ph~sphinoboranes,['~*'~~
B-P skeletal analogues of the ally1 cation,"'.
anion,[". 231 butadiene,[241and cage compounds.[251These results have already yielded much new information on B-P multiple bonding and they represent the beginning of what should prove to
be a rich vein of chemistry.
A feature common to a number of these recently characterized compounds is the presence of relatively short (1.81.9 A) B-P bond^.['^^'^-^^] This has given rise to considerable speculation on the strength of the bonding arising from
the overlap of the empty boron p orbital and the phosphorus
lone pair. This aspect of the structures is also of considerable
current interest in the context of multiple bonding in the
main group elements owing to the isoelectronic relationship
of the pairs BP, CSi and AlN. It is, perhaps, because of such
analogies and the implications for other main group 3-5
species, involving pairs such as Ga-As, ALP, and In-P, that
B-P compounds are currently enjoying a revival in interest.
2. B-P Bonding
A key parameter in the discussion of bond order is bond
length. Numerous other factors such as coordination number, type of substituent, molecular configuration and the
presence of delocalization also affect the length of a bond.
Any proposal for the existence of multiple bonding due to n
overlap has to be considered in the light of these influences.
Addition of the covalent radii of phosphorus (1.1 1 A) and
boron (0.82 A) suggests that a boron-phosphorus single
bond involving three-coordinate boron and phosphorus
should be about 1.93 8, long.[261It is thus instructive to consider this prediction in the light of theoretical studies on the
hypothetical molecules H,BPH, and the related compound
Ground S t a t e
H,BNH, . Experimental results on the various substituted
derivatives of H,BPH, will be discussed in Section 4.
The theoretical data clearly illustrate the large influence
that the molecular conformation can have on the length of
the B-P bond in H,BPH, .[''I Calculations have been carried
out on four different states of this molecule (Fig. 1). In the
Optimized Orthogonal
Fig. 1 Schematic drawings of four possible configurations of H,BPH, [27].
lowest energy (or ground) state the phosphorus center is
pyramidal (although it is somewhat flattened in comparison
to PH,) and the B-P distance is 1.91 A. In the optimized
planar and orthogonal forms the B-P bond lengths are
1.83 8, and 1.98 A. It is clear from the calculations that B-P
compounds do not show the same ability to form planar
species as B-N compounds. Furthermore, this is very probably a reflection of the high inversion barrier at phosphorus
(36.7 kcal mol-' for PH3)['*I rather than inherently poor
and inefficient overlap between the boron and phosphorus p
orbitals. The inversion barrier at phosphorus is, in fact, reduced to only 8.14 kcal mo1-l in H,BPH, which indicates
that there is strong conjugation in the planar form.[z71The
energy difference between the optimized planar form and
the non-optimized orthogonal form was calculated to be
40.4 kcal mol-'. This is only 2.2 kcal mol-' less than that in
H,BNH,. In addition, the results of population analyses are in agreement with these calculations. For example,
the total and n overlap populations in the planar form of
H,BPH, are 0.99 and 0.29, respectively, in comparison to
0.70 and 0.30 for H,BNH,. During rotation from planar to
the non-optimized orthogonal form, the electron population
in the vacant p orbital on boron is reduced from 0.29 to 0.01
in H,BPH, whereas in H,BNH, it is reduced from 0.29 to
Ange:ew. Chem. Int. Ed. Engl. 29 (1990)449-460
In short, the conjugation in planar H,BPH, is predicted to
be about the same as H,BNH,. However, non-planarity at
P is favored because of the fairly high inversion barrier at the
phosphorus center.
A further important consideration for the bonding in
boron-phosphorus compounds is the polarity of the B-P
pair. The electronegativity difference between boron and
phosphorus is very small (B = 2.01; P = 2.06, cf.
N = 3.07).[291Thus, a purely o-bonded boron-phosphorus
pair would possess only a small partial charge separation. In
a multiply bonded boron-phosphorus system charge is transferred from phosphorus back to boron via the n interaction
so that the G and n bond polarities tend to cancel each other.
Thus, it is possible to argue that the B-P pair could be significantly less polarized than the corresponding B-N system.
Such differences may become important when the aromaticity or reactivity of B-P systems is considered.
the boron-phosphorus compounds has been little investigated to date.
Various synthetic pathways for the formation of B-P
bonds are outlined in Equations 1-6.
R,_,BX, + nLiPR,R,_,B(PR,),
n = 1,2 or 3 ; R and R = large group
+ LiPH R
- LiX
+ nLiX
5 R2BP(Li)R
R and R = large group
RBX, + 2 LiPHR
n = 2 or 3 ; R and R
l / n (RBPR),
+ 2 LiX + H,PR'
= large
+ LiP(SiMe,)R 5RB(X)P(SiMe,)R'
1/2 (RBPR),
+ LiPHR' --%RB(X)PHR L*L+
1/2 (RBPR),
or NaN(SiMe,),
+ 2 LiP(SiMe,),
- 2 LiX
112 (R,NBPSiMe,),
3. Synthetic Strategies
The major synthetic approaches to the title compounds
involve either the use of sterically crowding groups or electron-rich substituents to reduce or prevent association of the
boron and phosphorus centers. Since boron is considerably
smaller than phosphorus it requires less highly hindering
substituents to block access of o donors to the empty boron
p orbital. It has been shown in a number of laboratories
that two mesityl substituents on boron are generally sufficient to restrict its coordination number to three.[301In a
different context, the Mes,B group has found use in the
stabilization of car bani on^,[^'* 311 and the reactions of these
ions display many of the characteristics of ylide~.[~']
Presumably, a similar restriction in access to the boron center
could be obtained with most substituents that are similar in
size or configuration to the mesityl group. It should be emphasized that the kinetic stabilization of the boron center is
primarily due to steric effects, and any reduction in the Lewis
acidity of the boron p orbital by interaction with the ring n
electrons of the mesityl substituents must necessarily be
small or non-existent because of the high dihedral angle between the mesityl rings and the boron plane.[311
An alternative method of preventing association of boronphosphorus compounds involves the use of n-bonding substituents on boron. The most commonly used group to
date has been the 2,2,6,6-tetramethylpiperidino ligand
(tmp).['3,'4' 1 7 * 1 8 1This species forms a n bond through a
side-on overlap of the nitrogen lone pair in a p orbital with
the empty p orbital on boron. This effectively reduces the
Lewis acid character of boron so that it is no longer very
susceptible to attack by other donors. Although tmp is a
large group, its effectiveness is primarily a result of the n
bonding to boron. The multiple nature of the boron-nitrogen bond in the tmp derivatives implies that when phosphorus substituents are included in the molecule they are, in
effect, competing with nitrogen for the boron p orbital. The
evidence to date shows that nitrogen is more effective than
phosphorus in this
Presumably, this is due
mainly to the lower barrier to inversion at nitrogen. The use
of other electron donating groups, such as
or -SR,
to reduce boron Lewis acidity is also possible. This aspect of
Angen. Chem. In1 Ed Engl. 29 (1990) 449-460
The simplest route involves the reaction between boron
halide derivatives and alkali metal salts of the phosphorus
precursors. This is currently the method of choice in the
synthesis of phosphinoboranes (Eq.
the B-P ally1 analogues (Eq. 1),lz2, 231 phosphanediylborates (Eq. 2)['01 and
the boron-phosphorus analogues of benzene and cyclobutadiene (Eq. 3).['01 In the case of the latter ring compounds the
reaction involves the elimination of a primary phosphane.
Synthetic pathways involving either the elimination of
Me,SiCI, from adjacent = PSiMe, and = BCI moieties
or HCl from =P-H and =BCl
groups in the presence of various bases (Eq. 5)[13. 14* 7. '*I or
disproportionation reactions with the elimination of
P(SiMe,), (Eq. 6)[") have also proved extremely effective.
4. Structural Data
4.1. Phosphinoboranes
Detailed structural information on a range of three-coordinate boron-phosphorus compounds did not become available until recently. In addition to the structural report on a
three-membered ring in 1981,["I the first X-ray crystal structures of several types of boron-phosphorus compounds
only appeared in 1986.[". 1 3 , 1 7 , 1 8 s 2 1 1 The simplest compounds are derivatives of the parent phosphinoborane
These are the species 11[14*171
(Fig. 2)
and 12[211(Fig. 3 ) which are accessible by addition of lithium
phosphide to the substituted boron halide [cf. Eq. (l)].Both
compounds are monomeric and feature a planar three-coordinate boron center and phosphorus geometries that have
different degrees of pyramidicity. Three further monomeric
phosphinoborane structures have also been determined[33s341and the important structural data for the five
compounds are presented in Table 1. The most notable feature of the data is the good correlation between the pyra451
Table 1. Important structural parameters of phosphinoboranes.
Bond length
Sum of angles
at P
Fig. 2. A thermal ellipsoidal plot of (tmp)ClBP(H)Mes 11 1171. The B-P distance is 1.948(3) A and the sum of the angles at P is 307.1".
H/ B-p\4H
midicity at phosphorus and the B-P bond lengths. There is
also remarkably good agreement with the theoretical predictions. When planarity is forced upon phosphorus by using
bulky mesityl s ~ b s t i t u e n t s [ ~
a ~bond
length as short as
1.839(8) 8, is observed (vs. 1.83 8, predicted for the planar
form of H,BPH,). At the other extreme, a long boron-phos-
[c] 1.91
[a] @ = out-of-plane angle at P, see formula A. [b] 1-Ad = 1-adamantyl.
[c] Calculated values.
11.4 kcal molfor H,BPH,).
(slightly higher than the value calculated
4.2. Phosphanediylborates (or Borylphosphides)
Fig. 3. A thermal ellipsoidal plot of Mes,BPh, 12 1211. The B-P distance is
1.859(3) 8, and the sum of the angles at P is 339.4".
phorus distance of 1.948(3) A is seen when there is a nitrogen
substituent on boron. The structure in Figure 2 neatly illustrates the dominance of the nitrogen lone pair interaction
with the boron p orbital. Not only is the B-P distance long,
the angles at phosphorus are typical of trivalent phosphanes
and the B-N distance is 1.380(3) 8, which is indicative of a
strong dative x interaction. The structure of 15[341perhaps
most closely resembles the ground state of H,PBH,
(ELP = 1.91 8,) and again the agreement between theory and
experiment is good. The inversion barriers in 12 and 15,
measured by variable temperature 'H NMR, are near 10 and
The phosphinoboranes above are of further interest because when there is a hydrogen substituent on phosphorus it
can easily be removed with nBuLi to give lithium salts having
the phosphanediylborate (or borylphosphide) anion
[R,BPRIe. The structures of the salts 16a-c and
17-20 have been determined.["] 20 features a naked
[PMesBMesJe ion. The structure of 16b['0"1is illustrated in
Figure 4. When [12]crown-4 is added to 16a, b, d to remove
the Li@ion,salts like 20 with anions of the type illustrated in
Figure 5 are obtained. In contrast to the neutral phosphinoborane precursors, the deprotonated phosphorus centers are
planar at phosphorus. Furthermore, the dihedral angles between the boron and phosphorus planes are small (usually
10'). The R P distances are short and vary from 1.810(4)
to 1.836(2) 8,. These features indicate considerable multiple
bond character in the B-P bond. The planarity at P is probably due to an enhanced 71 interaction owing to a greater
electron density at the phosphorus atom. Recent, more
sophisticated calculations [351 on the molecules Li[H,BPH]
and [H,BPHIe are in close agreement with experimental
findings and indicate essential double bond character in
these planar molecules and also in the planar configuration
Angew. Chem. Inr. Ed. Engl. 29 (1990) 449-460
fBuLi, THF
[ tmp-B=P-R]
I fBuLi
diate was reported to be detectable by mass spectroscopy of
the dimer which apparently dissociates at high temperat u r e , ~ i 3 . 141
Fig. 4. Structure of [Li(Et,O),][Mes,BPMes] 16b [loa]. The B-P distance is
1.823(7) A.
4.3. B-P Ring and Cage Compounds
The first diphosphadiboretanes to be characterized by
NMR and mass spectroscopy data and elemental analysis
were reported in 1977.["] These were the compounds 9,
which were synthesized by the thermolysis of 25 at 150 "C
[see Eq. (6)].[111The synthesis and X-ray structure of the
related three-membered ring compound 10 were reported in
1981.[121This was the first X-ray crystal structure report of
a compound with bonding between three-coordinate boron
and phosphorus centers (Fig. 6). The B-P distances average
Fig. 5. Structure of the anion [Mes,BPMesIe of 20, obtained by treating the
lithium salt in Figure 4 with [12]crown-4. The B-P distance is 1.835(13) A.
of H,BPH, . For example, calculations of the bond order in
[H,BPH]@ using two different
371 indicate
B-P bond orders of 1.83 and 2.01, respectively. Treatment of
the chloroborylphosphane 21 with tBuLi also gives the
anion 23.
a, R = cyclo-C,H,, ; b, R = Mes; c,
Trip = 2,4,6-iPr,C6H,
1-Ad; d, R
Fig. 6. A thermal ellipsoidal plot of the diphosphaborirane Et,NB(PfBu), 10
1.893(2) 8, and the B-N distance is 1.382(3) A. Although the
B-P bond is somewhat shortened relative to a single bond it
is notable that the B-N bond is consistent with considerable
[Li(Et,O),][Mes(H)PB(Trip)PMes] 19
R = NMe,, Ph
[Li([l2]crown-4),] [Mes,BPMes]
An interesting feature of this reaction is the concomitant
formation of the ring compound 24. This has led to the B = P
species 22 being postulated as inte~mediate."~]
This intermeAngew. Chern. I n t . Ed. Engl. 29 (1990) 449-460
A 1/2 (RBPSiMe,),
B-N bonding. The first X-ray structures of diphosphadiboretanes, involving the compounds 24," 31 26," 71 (Fig. 7),
and 27a,['*] were reported in 1986. They were synthesized
approach those found in four-coordinate B-P compounds.
This trend suggests that it should be possible to induce dissociation of the ring to give the interesting monomeric species
R,NBPR' by increasing the steric hindrance further. However, it is notable that the barrier for dimerization of the hypothetical H,N-B=PMe to (H,N-BPMe), has been calculated
(by MNDO 111) to be only 5 kcal mol-l.[lshJ
It has also proved possible to obtain diphosphadiboretane
compounds without amine substituents on boron.[20b. 341
The absence of competition for the boron p orbital allows its
tendency for interaction with the phosphorus lone pair to be
more accurately gauged. Alkyl- or aryl-substituted diphosphadiboretanes can be readily synthesized by treatment of an
organodihaloborane with two equiv. of LiPHR [see Eq. (3)].
Two examples are (ThexylBPMes),[20h1 and (1 -AdPB3 3 3
(MesBP( 1-Ad)),
32, Thexyl
Mes),[33.341(Fig. 8). The B,P, ring is planar and the phosphorus centers are pyramidal. The observed "B NMR
Fig. 7. Structure of the diphosphadiboretane (tmpBPMes), 26 [17]. The B-P
bond length is 1.916(3)A.
27 a
by reaction of tmpBC1, with LiP(SiMe,)R (+24), with one
equiv. of a 1 :1 mixture of LiPHMes and Li,PMes (-26) or
with LiPHCEt, (+27a) followed by elimination of Me,SiCl
or HC1 in the second step [see Eq. (4)]. These featured a
planar B,P, core. The phosphorus centers are pyramidal and
trigonal planar geometries were observed at both boron and
nitrogen with B-N distances of 1.38 A-1.431 A. The B-P
distances are 1.916(3) A, 1.96(2) A and 1.925 A (av.), and the
sum of the angles at P are 329.0' and 339.9'. Accordingly,
there is little evidence for dative P -+ B x bonding. It is also
notable that the exocyclic x bonding in these compounds is
stronger than that in the diazadiboretane 28 (B-N,, =
1.465(3) A) where the ring nitrogens compete for the boron
p orbital.[38] However, the B-N distance in 27a,[18J
1.431(2) A, is unusually long, although it remains consistent
with significant B-N x bonding. These data also support the
calculations (see Section 5.2)lZ7,351 which indicate a reluctance of the phosphorus center to participate in delocalization. Several other diphosphadiboretanes involving amino
substituents on boron have been reported, and three different synthetic methods have been described.['8bJThree of the
products, the compounds 29-31, have been structurally
and reveal features that are broadly similar to those that had already been r e p ~ r t e d . ~17.181
' ~ . Lo w
Fig. 8. Structure of the diphosphadiboretane (MesBP(1-Ad)), The B-P
distance averages 1.900(2) A. The sum of the angles at phosphorus is 328.9"
[33, 341
chemical shifts are consistent with three coordinate boron
whereas the 31P NMR 6 values (ca. -70) are well upfield.
The latter values are, in fact, more than A6 = 100 upfield of
those (see Table 2 below) observed in the delocalized sixTable 2. 'H and 31PNMR data (6 values) of selected B-P compounds
B-P bonds were observed in the compounds 30 and 31
[1.947(4)- 1.969(12) A]. This shows that increased steric
congestion increases the B-P bond length, as already observed in 24. I1 3, 1 7 ] It is notable that these B-P bond lengths
(MesBPcyclo-C,H, , ) 3
- 153.9
[a] Protons of the methyl groups of the mesityl substituents.
Angew. Chem. Int. Ed. Engl. 29 (1990) 449-460
membered ring (RBPR'), compounds which emphasizes the
very different bonding in the ring systems. The sums of the
angles at phosphorus are 331.3' and 328.9", respectively, and
the B-P bond lengths are 1.897(6) 8, and 1.899(2) 8,.[20b,331
These bonds are somewhat shorter than those usually observed in B,P, rings with -NR, groups on boron. A shortening is certainly predicted in the absence of competition from
the amino groups on boron. However, in view of the large
substituents on B and P a more planar phosphorus configuration might be expected. These results can be explained by
regarding the species in Figure 8 as a substituted cyclobutadiene analogue. In this case delocalization of the phosphorus
lone pairs onto the borons offers little energy advantage
since two of these electrons would occupy non-bonding x-energy levels. The energy gain (if any) upon delocalization is
probably insufficient to overcome the inversion barrier at
distances which average 1.955(3) 8, and 2.06614) 8, as the
boron atoms are substituted by exocyclic NR, groups. Even
longer B-P distances 1.97(1) 8, were observed in the more
encumbered ring system 36r401.Interestingly, the nitrogen
B .
a, E =
P; b, E = As
centers in all these rings are essentially planar whereas the P
(or As) centers are very pyramidal.
Details of other rings systems involving boron-phosphorus bonds have also appeared. Three-membered BP, rings
with various amino substituents on boron have been known
since 1981.[12]These were synthesized by the reaction of the
bis(alka1i metal) salt of a diphosphane with the aminoboron
dibalide. More recent resultsr411have extended the range of
these compounds to three-membered ring species involving
various combinations of B,N,Si and P and their reactions
with a variety of reagents. Fused three-membered boronphosphorus ring systems 37 can be prepared by the photolysis of the four-membered ring precursor 27.I4'] The structure
R 33b
Fig. 9. Schematic drawings of an amino-substituted diphosphadiboretane 26
(cf. Fig. 7) and two resonance forms a and b of an organo-substituted diphosphadiboretane 33 (cf. Fig. 8).
phosphorus. The bonding in the compounds in Figures 7 and
8 is illustrated schematically in Figure 9. In the amino-substituted compound 26 there is no delocalization and B-P single
bonding and B-N 71: bonding is observed. Although structure
33a most closely represents the bonding of the molecule in
Figure 8, it may be that structure 33b also makes a small
contribution owing to the shorter B-P bonds and the flattened geometry at phosphorus.
Other interesting four-membered ring systems involving
BP,N or AsB,N rings have also been reported.[391Examples
are 35 a and 35 b (Fig. 10). These involve long B-P and B-As
+ R,N-B\
a, NR,
= tmp;
b, NR,
of the product 37 a is illustrated in Figure 11
The B-P
distances average 1.906(6) b; and the P-P distance, 2.349(2)
A, is extremely long indicating, perhaps, considerable ring
Fig. 11. A thermal ellipsoidal plot of the bicyclic system 37a featuring two
three-membered BP, rings [42]. Some details are given in the text.
Fig. 10. Structure of the azaphosphadiboretane 35 a with B-P distances averaging 1.955(3) A. The nitrogen center is planar, the sum of the angles at phosphorus is 280.7" 1391.
Angen. Chem. Int. Ed. EngI. 29 (19901 449-460
strain. The B-N distances averaging 1.383(8) b; are indicative of an undiminished B-N x interaction. Another interesting ring system is the five-membered phosphadiborole derivative 38.[431Here the sum of the angles at phosphorus is
295.9" and the B-P distance is 1.949(2) A. The inversion
barrier at phosphorus was estimated at 10-15 kcal mol-'
on the basis of variable temperature 'H NMR spectroscopy.
Other notable B-P ring systems have also been reported recently. However, they involve either pentavalent [441 or fourcoordinate1451phosphorus and thus do not fall within the
scope of this review.
+ iPr,NBCI
pounds with endu-B-P bonds also involve boron bonding to
an exocyclic -NR, group which, for the most part, effectively prevents delocalization or dative n bonding between phosphorus and boron. In the two cases that involved no amino
substituents on boron it was argued that delocalization did
not occur because of the energetically unfavorable anti-aromatic cyclobutadiene structure of the rnolecule.Izob*
3 3 , 341
Conversely, a six-membered ring with alternate three-coordinate B and P atoms should show extensive delocalization
since three of the n-molecular orbitals would be bonding in
this case. It was found that, if the substituents on boron and
phosphorus in reaction (7) were of suitable size, good yields
+ 2LiPHR'+
+ H,PR' + 2LiX
p NiPr,
of these six-membered rings of type 34 could be obtained.["'
One of the products is illustrated in Figure 13.
Reaction of the phosphinoborane 39 with the borane 40 at
160"C affords the bicycle 41 (Fig. 12), the first example of a
cage compound involving three<oordinate boron and phosc9
Fig. 13. Structure of the borazlne analogue (MesBPcyclo-C,H,,), 3 4 b 1201.
The B-P bonds are all equal and about 1.84 A long.
Fig. 12. Structure of the first cage compound 41 involving three-coordinate
boron and phosphorus [25].
It has a trigonal bipyramidal B3P, framework.
The B-P distance 1.969(8)A is slightly longer than those
found in the four-membered ring compounds. The B-N distance 1.34(2) 8, is consistent with a high degree of B-N overlap which precludes B-P multiple bonding.
5. Possible Aromaticity in Delocalized B-P
Rings and Acyclic Systems
5.1. Benzene or Borazine Analogues
The foregoing discussion has shown that, with a few notable exceptions, the majority of currently known ring com456
The most important features of the rings are: a) the B3P3
array and the six @so-carbonatoms of the substituent groups
are coplanar; b) all the B-P distances are essentially equal;
c) the B-P bonds are short and about 1.84 in length. These
structural data clearly suggest considerable delocalization of
the phosphorus lone pairs over the six-membered ring system.
The structural differences between these systems and the
organo-substituted four-membered rings [20b, 3 3 ,341 discussed in Section 4.3 are striking. In the latter compounds
the phosphorus centers are pyramidal and the B-P bonds are
AngeK. Chem. Inr. Ed. Engl. 29 (1990) 449-460
about 1.9 A long. Important spectroscopic differences were
also observed. For example, the data in Table 2 indicate that
in the 31PNMR spectra there are large chemical shift differences between the four- and six-membered rings. The
‘H NMR data also show interesting differences. In acyclic
boron and phosphorus compounds and the four-membered
ring species the differences in chemical shift between the
ortho- and para-methyl protons tend to average ca. 0.2 ppm.
In contrast, larger differences of 0.42-0.75 ppm were observed for the benzene-like compounds. One possible explanation is that a ring current induced in the B,P, n system as
depicted in Figure 14 may account for the observed differences. With regard to the possible aromatic character of the
Fig. 15. Structure of the allyl-cation analogue PPh(BMes,), 42 [22]. The
B,PC(ipso), array is planar and the B-P bond is 1.871(2) 8, and the BPB angle
is 131.0(2)”.
Fig. 14. Schematic drawing of the possible origin of the anomalously large
difference in chemical shift between the orrho- and para-methyl protons of the
mesityl groups.
B3P3array it should be borne in mind that the B-P bonds are
not very polar. Thus, the low polarity in the B,P, ring should
facilitate ring current. However, proof of the aromatic character of these molecules awaits more sophisticated spectroscopic studies and theoretical calculations.
5.2. Acyclic Delocalized Systems
Further extended n-bonded systems involving B-P skeletons have also been synthesized. Several analogues of both
the x-ally1 cation and anion featuring B,P or BP, arrays with
a variety of substituents have been reported.[”* 231 Examples
are 42 (Fig. 15) and 43, which were synthesized by simple salt
elimination procedures.1221They may be regarded as ana(Mes,B),PPh
haves as a x acceptor. The planar nature of the B,PC, array
is good evidence for n bonding in the BPB moiety, which is
maximized by having all three p orbitals parallel to each
other even at the expense of considerably greater steric congestion caused by the coplanarity of the five organo groups
and the short B-P distance of 1.87 A
In 43 the boron is planar with B-P distances 1.89 A. However, the phosphorus centers in this case are pyramidal.
There is a slight asymmetry in the B-P distances which is
reflected in slightly differing degrees of pyramidicity.[’’I The
almost equal B-P distances in both compounds are consistent with the x-molecular orbital picture of the allyl group
since the extra pair of electrons is accommodated in a nonbonding orbital. This is also consistent with the slightly
longer B-P distance (1.89 vs. 1.87 A) on the basis of increased interelectronic repulsion. The preparation and structures of the compounds 44 and 45[231show that the electronic properties of the boron substituent exert considerable
influence on the degree of multiple P --t B bonding. In the
case of 44 with Br as substituent (-I effect, Fig. 16) the B-P
distances average 1.83 A, whereas in the OEt ( + I) derivative
45 the B-P distances average 1.93 A in length.
44,X = B r ; 45, X = O E t
logues of the allyl cation and anion, respectively. The allyl
cation analogue 42 features a twofold rotation axis along the
P-C bond. The CPB, center is therefore planar but the planarity also extends to the two ips0 carbon atoms on each
boron. There is also a wide BPB angle of 133”. Although
planar three-coordinate phosphorus is a common feature of
the structure of numerous phosphanediyl-metal complexes
of the type RP[M(CO),], ,I4’] the phosphorus in 42 differs
fundamentally from these because its valence shell possesses
four electron pairs. Three are used for CT bonding to one
carbon and two boron atoms. The remaining pair is available for n donation to the empty boron p orbitals. Conversely, in the bridging metal-phosphanediyls the phosphorus beAngew. Chem. Inr. Ed. Engl. 29 (1990) 449-460
Other B-P analogues of carbon compounds are also possible. For example, during reactions of 16 (a phosphido ligand
precursor with reduced bridging tendency owing to the
P + B effect) with some transition metal halides it was found
that the phosphido group could be oxidized to give the 1,2diboryldiphosphanes 46.r241These compounds constitute
46,a, R=l-Ad; b, R = M e s
a new class of diphosphanes with a number of unusual features (Fig. 17). The most notable are the short P-P bond
Fig. 16. Structure of the allyl-anion analogue BrB(PMes,), 44 [23].The B-P
distances of ca. 1.83 8, are quite short owing to the -I effect of the halogen.
Fig. 17. A thermal ellipsoidal plot of the butadiene analogue [P(I-Ad)BMes,],
46a 1241. The P-P single bond of ca. 2.1 A is extremely short.
distances, ca. 2.11 A, and the planar phosphorus and boron
centers. Clearly, an analogy could be drawn between these
compounds and butadiene. However, the twist angle between the two phosphorus planes is ca. 70" so that there is
little conjugative interaction between the two B-P x systems.
The short P-P bonds may be due to the changes in hybridization at phosphorus which now has -sp2 hybridized orbitals.
Overlap of these sp2 hybrids may result in shorter P-P bonds
than those in normal diphosphanes, where the P-P distances
are about 2.22-2.23 A. In the latter, the hybridization of the
bonding orbitals is less than sp3 (between sp3 and pure p)
since the angles at phosphorus are generally less than tetrahedral values. Interestingly, the "mixed" diphosphane, 47,
with one normal P-bound phosphorus atom and one sp2-hybridized B-bound phosphorus atom has a P-P bond length
which is close to that predicted by interpoof 2.173(2)
lation of the 2.11 and 2.22-2.23 A P-P values. These structural data imply that about half of the contraction observed
in P-P double bonds (ca. 2.02 8, long) is due to the rehybridization, the remainder being due to p p x overlap. This
is in contrast to doubly bonded carbon compounds where
only 25- 30 % of the shortening can be traced to rehybridization.
6. Reactivity Studies
Since many of the compounds described in the preceding
sections have only been recently discovered the investigation
of their chemistry is in the preliminary stage. In many cases
the boron-phosphorus compounds are sensitive to air and
moisture. However, the quasi-aromatic compounds described in Section 5.1 react only slowly (several days in "wet"
THF) with water to give a primary phosphane and various
boron-oxygen compounds.r201The ring systems also react
rapidly with Br, to give RPBr, and RBBr, products. The
four-membered 1,3-diphosphadiboretaneshave been shown
to undergo an interesting photolysis reaction to give fused
ring systems as in 37 a (Fig. 1
The related four-membered 1,3,2,4-azaphosphadiboretaneskeleton has also been
Addition of methyl iodide led to formation
of the methylphosphonium salt 48, which was characterized
by 'H, "B, I3C and 31PNMR spectra and by an elemental analysis.[401The phosphorus lone pair may also form
a bond to W(CO), . With selenium the P(tBu) moiety is displaced from the ring and a cyclic four-membered ring involving NB,Se is produced instead. The three-membered ring
system 49 also undergoes a variety of reactions, e.g. with
M(CO),, S, Se or MeLr4'l These facile reactions clearly
N- P-iPr
= S, Se
+ PMeiPrI
demonstrate the ready availability of the phosphorus lone
pair in 49 and its almost negligible involvement with the
boron p orbital in this case.
Compounds involving bonding between four-coordinate
boron and phosphorus are generally very stable. However,
when coordinative unsaturation is introduced rearrangement and elimination reactions may occur. For example, the
attempted synthesis of Br,BPMes, led instead to the novel
heterocycle 50, which presumably arose from the activation
Angew. Chem. Int. Ed. Engi. 29 (1990) 449-460
of the C-H bond on one of the ortho-CH, groups of a mesityl
s u b s t i t ~ e n t . Activation
of C-H has also been observed
during the attempted synthesis of the phosphinoborane 51 ;
in this case the dimer 52 was isolated.[471The proposed reaction sequence which led to this unexpected compound is
believed to involve a double elimination (retro-hydroboration) of isobutene followed by an anti-Markovnikov hydroboration and dimerization to give the product 52.
tBu, tBu
tBu -,P-B.-H
- 2 C,H,
t- [iBu(H)B-PtBu,]
7. Outlook
Several decades after the first phosphinoboranes and borane-phosphane adducts were reported, the first compounds
involving multiple boron-phosphorus bonding are being investigated. It is ironic to note that exotic compounds involving isoelectronic carbon-silicon double bonded systems were
isolated and structurally characterized some years before
their simple boron-phosphorus counterparts the phosphinoboranes. Similarly, the synthesis of the boron-phosphorus
counterparts of borazines involved nothing more complicated than the addition of two equivalents of LiPHR to
R’BHal,. The initial findings suggest that there are many
exciting discoveries to be made. The synthesis of stable B-P
analogues of the iminoboranes, RBPR’, remains an obvious
goal. The coordination chemistry of the compounds already
synthesized and the synthesis of a large variety of boronphosphorus cage compounds are areas of considerable potential. A striking feature of the work already published is
the very large effect that the boron substituents can have on
B-P multiple bonding and in most cases where nitrogen is
bonded to boron little or no multiple bonding to phosphorus
is observed. On the other hand, when there is an electronwithdrawing substituent on boron enhanced multiple bonding to phosphorus is seen. It should therefore be possible to
control boron-phosphorus multiple bonding very accurately
with the appropriate substituent.
Note added in proof (April 3, 1990): Species involving
two-coordinate boron and boron-phosphorus multiple
bonds have now appeared. The Zintl anions [BP,]3e and
[BAs,13@were crystallized as their K@ salts from the reaction of the elements at 1000- 1100 K. The bond lengths are
B-P = 1.767 A and B-As = 1.868 A. These short distances
are in line with earlier predictions[”’ and with the two-coordination and sp hybridization at
The author is grateful to his many coworkers named in the
bibliography who carried out many of the experiments described with skill and enthusiasm. The financial assistance of
Angew. Chem. Int. Ed. Engl. 29 (1990) 449-460
the National Science Foundation and the A . P . Sloan Foundation are also gratefully acknowledged.
Received: September 12, 1989 [A 758 IE]
German version: Angew. Chem. 102 (1990) 527
[l] A. Stock, E. Pohland, Ber. Dtsch. Chem. Ges. 59 (1926) 2215.
(21 K. Niedenzu, J. W. Dawson: Boron-Nitrogen Compounds, Springer, Berlin
[3] J. R. Wasson in Gmelin Handbuch der Anorganrschen Chemie, Erganzungswerk zur 8. Aufl. Bd. 19, Borverbindungen, Tetl 3 , Springer, Berlin 1975,
p. 93.
[4] G. E. Coates, J. G. Livingstone, J. Chem. SOC.1961, 1000.
[S] G. E. Coates, J. G. Livingstone, J. Chem. Sor. 1961, 5053.
[6] A. B. Burg, R. I. Wagner, J Am. Chem. SOC.75 (1953) 3872; A. B. Burg in
J. F. Liehman, A. Greenberg. R. E. Williams (Eds.): Advances in Boron and
the Boranes, VCH Verlagsgesellschaft, Weinheim, 1988, p. 1.
[7] A. D. Tevebaugh, Inorg. Chem. 3 (1964) 302.
[8] W. D. English, Chem. Abstr. 57 (1962) 11239. US Pat. 3035095.
191 D A. Brown, C. G. McCormack, J Chem. SOC.1964, 5385.
[lo] a) R. A. Bartlett, X. Feng, P. P. Power,J. Am. Chem. SOC.108(1986)6817;
b) R. A. Bartlett, H. V. R. Dias, X. Feng, P. P. Power, ibid. I f f (1989)
[ l l ] G. Fritz, W. Holderich, Z . Anorg. Allg. Chem. 431 (1977) 61.
[12] a ) M . Baudler, A. Marx, 2. Anorg. Allg. Chem. 474 (1981) 18; b);
M. Fehir, R. Frolich, K.-F. Tebbe, tbid. 474 (1981) 31.
[I31 A. M. Arif, J. E. Boggs,A. H. Cowley, JLG. Lee, M. Pakulski, J. M. Power,
J. Am. Chem. SOC.108 (1986) 6083.
[14] C. Couret, J. Escudie, B. Garrigues, M. Lazraq, J. Satge, Can. J. Chem. 65
(1987) 1230.
[15] R. A. Bartlett, X. Feng, M. M. Olmstead, P. P. Power, Phosphorus Suljur,
20 (1987) 245.
[16] M. Baudler, M. Hintze, Z. Anorg. Altg. Chem. 522 (1985) 184.
[17] A. M. Arif, A. M. Cowley, M. Pakulski, J. M. Power, J. Chem. SOC.Chem.
Commun. 1986, 889.
1181 a) P. Kolle,H. Noth, R. T. Paine, Chem. Ber. 119 (1986). 2681; b) P. Kolle,
G. Linti, H. Noth, G. L. Wood, C. K. Narula, R. T. Paine, ibid. 121 (1988)
[I91 J. Escudie, C. Couret, M. Lazraq, B. Garrigues, Synth. React. Inorg. Met..
Org. Chem. (1987) 379.
[20] a) H. V. R. Dias, P. P. Power, Angew. Chem. 99 (1987) 1320; Angew.
Chem. I n t . Ed. Engl. 26(1987) 1270; b) J. Am. Chem. SOC.11f (1989) 144.
[21] X. Feng, M. M. Olmstead, P. P. Power, Inorg. Chem. 25 (1986) 4616.
1221 R. A. Bartlett, H. V. R. Dias, P. P. Power, Inorg. Chem. 27 (1988) 3919.
[23] H. H. Karsch, G. Hanika, B. Huber, K. Mundl, S . Konig, C. Kriiger, G.
Miiller, J Chem. SOC.Chem. Commun. 1989, 373.
[24] D. C. Pestana, P. P. Power, J. Am. Chem. SOC.111 (1989) 6887.
[25] G. L. Wood, E. N. Duesler, C. K. Narula, R. T. Paine, H. Noth, J Chem.
SOC.Chem. Commun. 1987.496.
[26] The radius of boron is the subject of some controversy [loa], much of
which can be resolved by consideration of the coordination number for
boron. The value 0.82 A appears to be a reasonable one for three-coordinate boron. For example, compounds involving three-coordinate boroncarbon bonds usually have B-C bond distances of ca. 1.58-1.60 8, [26a].
If the usually accepted carbon radius (0.77 8,) is subtracted, a value of
0.81 -0.83 8, is obtained for boron. Longer distances are observed when
boron is four-coordinate. Typical B-P distances for such compounds are
about 2.0 8, or an effective radius of 0.9 8, for four-coordinate boron.
[27] 0. Gropen, J. Mol. Struct. 36 (1977) 111. Other more recent calculations
are to be found in: E. Magnusson. Tetrahedron 41 (1985) 5235; Aust. J.
Chem. 39 (1986) 735; J. A. Baban, B. P. Roberts, J. Chem. SOC.Perkin
Trans. I1 1987 487.
(281 J. M. Lehn, B. Munsch, Mol. Phys. 23 (1972) 91. However, the presence of
substituents such as silyl groups can have a dramatic effect o n the inversion
barrier e.g. 18.9 kcal mol-’ in the silylphosphane Me,SiPPh(tF’r), see
R. D. Baechler, K. Mislow, J. Am. Chem. Soc. 92 (1970) 4759 and A. Rauk,
J. D. Andose, W. G. Frick, R. Tang, K. Mislow. ibid. 93 (1971) 6507. The
studies show that there is a strong correlation between substituent electronegativity and the inversion barrier. Presumably, some of the lowering.
of the barrier in phosphinoboranes is due to the low electronegativity of
the boryl group. However, it is notable that the calculated and experimental inversion barriers in these compounds (8-11 kcal mol-’) are very
much lower than those in the silyl compounds, whereas the electronegativity of boron (2.01) IS greater than silicon (1.74) (291.
[29] A. L. Allred, E. G. Rochow, J. Inorg. Nucl. Chem. 5 (1958) 264.
[30] H. C. Brown, V. H. Dodson, J Am. Chem. Soc. 79 (1957) 2302; J. W.
Wilson, J Organomet. Chem. 186 (1980) 297; A. Pelter, B. Singaram, L.
Williams, J. W Wilson, Tetrahedron Lett. 24 (1983) 620.
[31] R. A. Bartlett, P. P. Power, Organometallics. 5 (1986) 1916; M. M. Olmstead, P. P. Power, K. J. Weese, R. J. Doedens, J Am. Chern. Soc. 109
(1987) 2541.
[32] A. Pelter, B. Singaram, J. W. Wilson, Tetrahedron Lett. 24 (1983) 635.
[33] D. C. Pestana, P. P. Power, unpublished results.
[34] H. V. R. Dias, D. C. Pestana, M. Petrie, P. P. Power, Phosphorus Sulfur, in
[35] T. L. Allen, A. C. Scheiner, H. F. Schaefer, 111, Inorg. Chem., unpublished.
[36] I. Mayer, Chem. Phys. Lett. 97(1983)270; 110(1984)440;Int.J. Quanfum
Chem. 29 (1986) 73.
[37] M. A. Natiello, J. A. Medrano, Chem. Phys. Lett. 105 (1984) 180; ibid If0
(1984) 445.
[38] P. Paetzold, Adv. Inorg. Chem. Radiochem. 31 (1987) 123.
[39] K.-H. van Bonn, P. Schreyer, P. Paetzold, R. Boese, Chem. Ber. f2ff1988)
[40] P. Kolle, H. Noth, R. T. Paine, W. Rattay, Z . Narurforsch. 543 (1988)
[41] P. Kolle, H. Noth, R. T. Paine, Chem. Ber. 122 (1989) 423.
[42] P. Kolle, G. Linti, H. Noth, K. Polborn, 1 Organomrt. Chem. 355
(1988) I .
[43] M. Driess, H. Pritzkow, W. Siebert, Angew. Chem. 99 (1987) 789; Angew.
Chem. Int. Ed. Engl. 26 (1987) 781.
[44] D. Gudat, E. Niecke, M. Nieger, P. Paetzold, Chem. Ber. 121 (1988) 565
[45] H. Noth, 2. Anorg. A&. Chem. 555 (1987) 79.
[46] M. Dness, H. Pritzkow, W. Siebert, Angew. Chem. 100 (1988) 410; Angew.
Chem. Int. Ed. Engl. 27 (1988) 399.
1471 G. Huttner, K. Evertz, Acc. Chem. Res. 19 (1986) 406.
[48] H. H. Karsch, G. Hanika, B. Huber, J. Riede, G. Muller, J. Orgonomet.
Chem. 361 (1989) C25-C29.
[491 H.-G. von Schnering, M. Somer, M. Hartweg, K. Peters, Angew. Chem.
102 (1990) 63; Angew. Chem. Int. Ed. Engl. 29 (1990) 65.
Angew. Chem. Int. Ed. Engl. 29 (1990) 449-460
Без категории
Размер файла
1 045 Кб
bonding, compounds, multiple, boron, phosphorus
Пожаловаться на содержимое документа