Phospha-alkynes: new building blocks in organic, inorganic, and organometallic chemistr J John F. Nixon

P h o s p h a - a l k y n e s , RC=_P, a re c o m p o u n d s tha t con t a in a t r ip le b o n d b e t w e e n the c a r b o n a n d

p h o s p h o r u s a t o m s . T h e y w e r e o n c e t h o u g h t , o n t heo re t i c a l g r o u n d s , to b e i n c a p a b l e o f e x i s t e n c e b u t

o v e r the pas t d e c a d e t hey h a v e b e e n s h o w n to h a v e a qu i te r e m a r k a b l e versa t i l i ty as c h e m i c a l bu i ld ing

b locks . Th i s ar t ic le r ev iews the i r i m p a c t on o rgan ic , i no rgan ic , a n d o r g a n o m e t a l l i c c h e m i s t r y a n d in

c lus te r f o r m a t i o n .

In 1961, T. E. Gier, a chemist working in the research laboratories of E. I. Du Pont de Nemours in Wilmington, Dela- ware, U.S.A., reported the synthesis of the thermally unstable compound HCEP. He presented unambiguous ex- perimental proof that the compound was the phosphorus analogue of hyd- rogen cyanide, and therefore unques- tionably contained a triple bond be- tween the carbon and phosphorus atoms [1]. This result was contrary to all theoretical expectations at that time (vide infra). Nevertheless, Gier's paper aroused little immediate activity from the chemical community and the com- pound remained a mere chemical curiosity for many years. The real im- portance of this class of compounds, called the phospha-alkynes, RCEP (R = organic group, has been fully recog- nised only in the past decade and many exciting developments have resulted.

A number of factors may be cited as contributing to the early lack of interest in phospha-alkyne chemistry. Firstly, the synthetic route to HC--P was not particularly attractive, involving as it did striking an arc between carbon elec- trodes in an atmosphere of the highly reactive gas PH3. Furthermore, the faint-hearted were no doubt discour- aged by the report that HC--P was not only spontaneously inflammable in air, but also readily polymerised to a black

John F. Nixon, B.Sc., M.Sc., Ph.D., D.Sc.

Was born in 1937 and graduated at Manchester University. After post-doctoral research at the University of Southern California and at Cam- bridge, he was appointed to a lectureship at St Andrews. He then moved to the University of Sussex, where he is now Professor of Chemis- try and Dean of the School of Chemistry and Molecular Sciences.

I Endeavour, New Series, Volume 15, No. 2, 1991. I 0160-9327/91 $3.00 + 0.00. I © 1991. Pergamon Press pie. Printed in Great Britain.

R

R = Aryl P ~ S ~ D ~ N R ~ - R=H

R R I [ R = CRy, CzH~

R R

Figure 1 Phosphorus - carbon double bonds in phosphabenzene (left) and phosphamethine cations.

pyrophoric solid even at. temperatures as low as -78°C!

Perhaps the greatest single barrier to progress, however, lay within the minds of chemists, who until comparatively recently were inhibited by their belief in the so-called 'double-bond' rule [2, 3] which stipulated that compounds con- taining multiple bonds should be con- fined to elements of the first row of the Periodic Table. This view was based on the theoretical idea that whereas 2p~- 2p~ orbital overlap, is responsible for the multiple bond formation in simple molecular compounds like 02, N2, CO, NO, CN-, etc., the corresponding bonding interactions involving 3p~ and 4p~ orbitals used by the heavier ele- ments would be too small to be of significance.

Such ideas helped to explain known structural differences between closely related elements in the Periodic Table. For example, N2 and 02 are gaseous diatomic molecules held together by three and two bonds respectively, whereas white phosphorus (P4) and elemental sulphur ($8) are both solids at room temperature, having cage and ring structures involving essentially single bonds.

Over the past 10-15 years views have had to be modified as more and more examples of compounds which contain

C=P, C--P, C=Si, P-=P,-Si=Si, and As=As multiple bonds [4, 5] have been synthesised in laboratories all over the world. This article charts the rapid de- velopments in the chemistry of one class of these compounds, the phospha- alkynes, R C - P , and describes their re- markable impact in the areas of inorga- nic, organic, and organometallic che- mistry.

During the late 1960s and early 1970s a steady stream of reports describing the syntheses of a variety of compounds containing phosphorus-carbon double bonds cast doubt on the validity of the double-bond rule, and particularly noteworthy were the phospha-benzenes and phosphamethine cyanine cations [6, 7] (figure 1).

At Sussex University, in 1976, Profes- sor Harry Kroto and I, together with our co-workers, made a number of rather unstable phospha-alkenes, CF2=PH, CH2=PH, and CH2=PCI, containing phosphorus-carbon double bonds [8(a)]. We subsequently gener- ated a series of phospha-alkynes RC--P (R = H, CH3, CF3, etc.) b y a much more straightforward method than Gier's original route, simply by pyroly- sis at ca 900°C of halogenophosphanes which readily eliminated hydrogen halide:

CH3CH 2 PC Iz --- C H 3 C - ' - : P + 2HCI

49

We quickly became convinced that a whole family of phospha-alkynes must exist when we fortuitously generated the fairly stable F C - P , even at room temperature, in a simple glass apparatus merely by removing hydrogen fluoride from CF3PH2 by treatment with sodium hydroxide pellets [8(b)]:

which has a bondlength of 152 pm ( P - C ) and is a completely stable compound. '

The report of the indefinitely stable ButC_--P by Professor Gerhard Becker's team represented a major break- through, and the paper subsequently

- H F

C F 3 P H 2 ~ CF2 PH -HF

~-- F C ~ P

Structural data readily confirmed that the phospha-alkynes were linear molecules and had extremely short phosphorus-carbon bonds (154 pm) compared to those of P= C double bonds (170 pm) and P - C single bonds (187 pm). Photo-electron spectroscopic studies revealed that the electronic ionisation energies were lower than those known for the corresponding nit- riles RC--N and, of particular import- ance, the phosphorus lone-pair elec- trons were more tightly held than those in the P - C triple bond.

Lone-pair eLectrons are more difficult, to ionise

R C ~ 'P: ~f"

ELectrons in triple bond ore easiest, to ionise

Around this time, and unknown to us, Professor Rolf Appel and his group in Bonn, Germany, were devising alter- native synthetic routes to phospha- alkynes which were based on the ready elimination of chlorosilanes from suit- able precursors [9]. They obtained C6HsC-P and Me3SiC-P, both of which showed greater thermal stability than any of the Sussex compounds, viz:

published early in 1981 [10] was a bril- liant extension of their important work on the isolation of the stable compound Me3SiP=C(OSiMe3)Bu t containing an isolated P=C double bond. Now they had found the trick necessary to cataly- se the desired elimination of disiloxane, MeaSiOSiMe3, to give the phospha- alkyne, using traces of sodium hydro- xide.

Me3Si X / O S i M e 3

P C

~ B u t

ray crystallographic study by Dr Peter Hitchcock, is shown below (figure 2).

It revealed that the phospha-alkyne was attached to the platinum in a 'side- on' fashion held by the ~-electrons of the P--C triple bond, while the phos- phorus lone-pair electrons actually pointed away from the metal atom. On coordination the phosphorus--carbon bond length is increased to a value of 167 pm and the phospha-alkyne is no longer linear. These features were ex- actly those expected on the basis of the known chemistry of alkyne metal com- plexes and showed the similarity in be- haviour of the two classes of com- pounds. Further confirmation quickly came from the synthesis of the analo- gous dinuclear cobalt carbonyl phos- pha-alkyne complex (figure 3) which also has an alkyne counterpart.

Once the phospha-alkyne is bonded side-on to the metal it turns out that it can be persuaded to interact further

Me3Si~ IVle3 S iT C--PCI2

R--

heat

NaOH P " - - - C B u t + Me3SiOSiMe3

(or base)

In late November 1980 I received a letter from Rolf containing some news which caused us great excitement at Sussex University. The middle para- graph reads as below:

' I t seems to me that the Phos- phorus-Carbon-Chemistry with multiple bonds becomes a fascina- ting field of preparative Chemistry. In the meantime we could syn- thesize the Me3Si-C_--P which has a halflife time of 50 minutes at RT. On the occasion of a visit and lec- ture at the university of G6ttingen I heard from Becker in Marburg that he prepared the t e r t . B u t . - C - P ,

B u t C - p is a remarkable compound, being a colourless liquid, boiling at 61°C without decomposition. In recent years improved methods for increasing the yields of several phospha-alkynes using the base-catalysed reaction have been developed by Professor Manfred Regitz and his group in Kaiserslautern, Ger- many, and a variety of thermally stable phospha-alkynes R C - P (R = B u t, adamantyl, isopropyl, etc.) are now readily available.

Phospha-alkynes obviously have the very interesting potential to exhibit a

RC~___-- P + 21Vle3SiCI

R = C6H5, Me3 Si )

duality of chemical behaviour. On the one hand they might react in a similar fashion to their better-known organic counterparts the alkynes, RCECR, and on the other hand the phosphorus lone- pair electrons might enable them to coordinate to a variety of metal atoms, as is the case for many phosphane (R3P) ligands. Of course, in suitable circumst- ances both types of behaviour might also be possible.

Early in 1981 my D.Phil. student James Burckett St-Laurent synthesised the very first transition metal complex [Pt(PPh3)2ButCP] containing a phos- pha-alkyne [11]. The molecular struc- ture, determined by a single-crystal X-

with other transition metals, to give compounds typified by the example shown in figure 3. Probably the most remarkable example of the duality of behaviour is shown in figure 4 which depicts the 'star' shaped compound [Pt3Pdz(PPh3)5(ButCp)3] containing a cluster of five precious metals. Cluster formation involving phospha-alkynes is proving to be an important and unex- pected feature of their chemistry, as will become clear later in this article.

It turns out that by far the most favoured bonding of a phospha-alkyne to a transition metal is via the ~r- electrons of the P - C triple bond. A few years ago, however, we were able to induce the phospha-alkyne to bind pre- ferentially to a metal using only its phosphorus lone-pair electrons by creating a very special environment at the metal centre which allowed linear molecules to approach in an 'end-on' rather than 'side-on' manner. One such compound that is fully structurally char- acterized is shown in figure 5 [12].

The predominantly alkyne behaviour of phospha-alkynes has been exploited in a large number of reactions of an 'organic' type in an extensive series of elegant studies by Manfred Regitz and his group [13]. A large number of [3 + 2J cyclo-addition reactions typified by Scheme 1 give rise to a rich assortment of heterocyclic compounds. Likewise, a selection of [4 + 2] cycio-addition reac- tions which are directly analogous to the famous Diels-Alder reaction long

50

O Cerbon @ Phosphor=

Me'coL

Co

J P (C0)3 Co~ c ~ / C o (CO)~

Bu t

Figure 3 Dinuclear and trinuclear cobalt carbonyl phospha-alkyne complexes.

Figure 2 First transition metal complex containing a phospha-alkyne. The same shading is used in subsequent figures.

( ~ I=h3P

)CH3

CH3

p~

C

P

Figure 4 Star-shaped compound containing three phospha-alkynes and a cluster of five precious metals.

Figure 5 Phospha-alkynes preferentially bound to a metal solely by phosphorus lone-pair electrons.

51

known in organic chemistry are pre- sented in Scheme 2. Further develop- ments can be expected in this area of chemistry, leading to a wide variety of novel organophosphorus compounds not available by other routes.

In the area of organotransition metal chemistry it has long been known that alkynes could be readily cyclodime- rized, cyclotrimerized, and oligome- rized by various transition metals. In 1986, Dr Mohd Jamil Maah, Dr Peter Hitchcock, and I showed that cyclo- dimerization reactions of Bu t C - P at cobalt, rhodium, and iridium centres gave rise to novel 1,3- diphosphacyclobutadiene compounds (figure 6).

Similar studies were being carried out unknown to us by a research team at the Max Planck Institut, Miilheim, Ger- many, under the direction of Professor Paul Binger, and their results and ours appeared almost simultaneously [14, 15]. These [2 + 2] cyclo-addition reactions of phospha-alkynes parallel the behaviour of alkynes which led to the first cyclobutadiene complexes in the 1960s and there are structural simi- larities between the two classes of com- pounds.

Once again the phosphorus com- pounds have additional bonding possi- bilities by virtue of their lone-pair elec- trons and this has been exploited in the synthesis of the remarkable hexarho- dium complex shown below (figure 7).

When zirconium is the metal centre however, an unexpected coupling of two B u t C - p units occurs to give a compound containing a phosphorus- phosphorus bond (figure 8), and, as discussed later, this complex has proved to be of key importance in generating novel metal-free cages containing only phosphorus and carbon.

Perhaps the most striking develop- ment in the use of phospha-alkynes as building blocks in organo-transition metal chemistry resulted from the initial report by Becker et al. of the triphos- phorus analogue of the planar cyc- lopentadienyl anion (P3f2But2) - [16].

R

,L

Me/N~ R +

1 /

H

L R'~N~y R

• - P + R R'---~.,O R

+ PO~R '

Scheme 1

R R R

1 l -X E

We wondered whether this anion might offer an entry into transition metal 'sandwhich' complexes similar to fcrrocene, [Fe(CsHs)2], and other metal- locenes extensively developed by Nobel

52

Scheme 2

R R

ButC P

oco

Figure 6 Novel 1,3-diphosphacyclobutadiene compounds. Figure 8 Zirconium complex with phosphorus-phosphorus bond.

R R

R R ~ R R R R ~ R R

M M

R R

M=Rh

Figure 7 Hexarhodium complex with lone-pair electron bonding.

Laureates Sir Geoffrey Wilkinson and E. Otto Fischer in the 1950s.

In 1987, Dr Rainer Bartsch, Dr Peter Hitchcock, and I were able to make green crystals of both [Fe(P3C2But2)2] and [Fe(P2C3But3) (P3C2But2)]. These were exceptionally thermally stable and unreactive to both air and moisture, because like ferrocene they have a spe- cially stable electronic arrangement in- volving 18 valence electrons (figure 9).

Each compound was shown to have the expected ferrocene sandwich struc- ture in which 6 or 5 carbon atoms of the two rings had been replaced by phos-

O F e

Figure 9 Two ferrocene-type compounds, seen in perspective, ~n which 6 or 5 carbon a toms of the two rings have been replaced by phosphorus.

phorus [17]. These remarkable com- pounds illustrate vividly the ease with which phosphorus and carbon atoms can be interchanged in organotransition metal compoundsand point the way for future deve!opments. Subsequently, the paramagnetic hexaphosphorus ana- logue of the 16-electron 'chromocene' [Cr(P3CzBut2)2] was also obtained, b u t attempts to make the 19-electron cobalt and 20-electron nickel analogues gave different types of products [18] shown below (figure 10), in which the prefer- red 18-electron system was maintained in each case.

As before, the phosphorus lone-pair

electrons of the ring phosphorus atoms can be used to connect these new phos- phaferrocenes to other metals, for ex- ample:

C P 0

5 3

H

) Pc l ~.. C(4)

H

Co + P(2)(

P(5)

CoG P(I)

C(2)

P(4)

~P(3)

Ni

Figure 10 Attempts to make 19-electron cobalt and 20-electron nickel analogues of 16-electron 'chromocene' led instead to structures shown above.

and there is little doubt that numerous other examples of mixed metal systems will be developed in the future, some of which might be useful catalysts.

A further series of unexpected excit- ing developments in phospha-alkyne chemistry occurred during the summer of 1989 with independent reports from the laboratories of Regitz [19], Zenneck [20], and myself [21] of unusual 'cage' structures containing carbon and phos- phorus. Regitz et al. found that when ButC--P was heated for a long period at 130°C a very small amount of a colour- less, air-stable, crystalline material melting at 241°C is formed. The new product turns out to be the tetramer of B u t C - p and its structure (figure 11) consists of a cube of carbon atoms, but with alternate corner atoms being re- placed by phosphorus. This important discovery featured on the front cover of the August 1989 issue of Angewandte Chemie. In Sussex, Rainer Bartsch and I also succeeded in combining together the P3C2But2 and P2C3But3 rings men- tioned earlier, to give an almost colour- less crystalline compound which Peter Hitchcock showed was the pentamer of B u t C - p [22] (figure 12).

P

P-----C--tBu no sotvent 150*C + =_

tBu--C~P .°L-._ /

P-'-~tBu J P- -~ tBu

P

C

P-C:1.881~ i P-C-P:94.4%C-P-C: 85.6 =

mp - 241"C (co(od,ess), can be sut~.lmed

Figure 11 Tetramer of ButC-p consisting of centre of carbon atoms with alternate corner atoms replaced by phosphorus.

Figure 12 Pentamer of ButC_--P.

A striking structural relationship exists between the [ButCP]4 and [ButCP]5 cluster molecules. The former structure is based on a cube, while the latter arises by replacing a phosphorus atom at the corner of the cube by a triangle of two phosphorus and one carbon atom. Zenneck et al. [20] also obtained the pentamer by a completely different route in which a highly reactive orga- nometallic iron compound reacts with BurCP, and other 'cage' structures were also obtained. Yet another cage struc- ture, P6CaBut4H2, this time containing six phosphorus and four carbon atoms, readily formed when the (P3C2But2) an- ion was carefully protonated and its structure is shown below (figure 13) [221.

54

P

P6C. Bu~ H z

Figure 13 Cage structure containing six phosphorus and four carbon atoms.

In all the 'cage' structures so far re- ported the P-C bond lengths lie in the range expected for single bonds, and in the P4C4 'cubane' structure the CPC bond angles are on average 4 ° smaller than the 90 ° angles found in cubane, CsHs, with the PCP angles being corres- pondingly larger. The existence of the 3-membered ring in PsC5But5 (figure ~ s o remarkable, as is its very small (ca 50 °) CPC bond angle. These polycyclic compounds are remarkably thermally robust and air-stable. Theore- tical calculations on PnC4But4 show that the charges on the P atoms are more positive than on the carbon atoms.

In very recent unpublished work, Regitz and Wettling [23] have disco- vered a clever way of greatly increasing the yield of PaC4Bu 4 by treatment of the zirconium complex, [Zr(CsHs)2 (P2C2But2)], mentioned earlier, with C2C16 to remove the metal completely. In this way the tetraphosphorus cubane compound is obtained in 70 per cent yield (figure 15).

The ready availability of P4C4But4 now enables extra features of its che- mistry to be explored, and already sul- phur atoms have been added. A very careful re-examination of the thermal pyrolysis of B u t C - p under different experimental conditions has also led to other interesting cage structures and this area of chemistry is clearly only in its infancy [23] (figures 16 and 17).

'Cage' compounds containing carbon and phosphorus can also be extended to incorporate transition metals, and one recent example is [Rh2(CsMe~) 2 P2C2But2] whose structure (figure 18) was established by X-ray crystallo- graphy [24].

These compounds provide important links between main group element che- mistry and transition metal chemistry and furthermore since each [Rh(CsMes)] fragment can be viewed as a [CH] + species, using well established

(a)

(b) P5C5

• Carbon 0 Phosphorus On each carbon there Is a t-butyL group

Figure 14 (a) Arrangement of the central atoms of PsCsButs. (b) The structure is based on a cube with one corner atom removed and replaced by a tr iangle of atoms.

tBu tBu

Cp\ CL3C_CCLs, benzene,25oC {Bu - ~ - / P

/ ~ / / -CpzZrCL z tBu - - - Cp -CL2C~-~-- CCL2

U tBu

(70%)

Figure 15 New tetraphosphorus cubane synthesis.

55

P----C--tBu no soLvent. 8h, 180°C (gLoss pressure t.ubel

p

p •

~Bu p~~p tBu

tBu tBu

benzene, 72, 180"C (gLoss pressure tube)

~HzC~--- ClVle z

tBu

~Bu

tBu

no solvent 3min, 400*C (gLoss pressure bJbe)

- t B u - - C ~ C - - tBu

J . P x ~ B u

tBu 1 l.,K.z P~ tBu P

, ( ~Bu

Figure 16

tBu

t Buolp'TL" ~, P

tBu

• , Se, C6H6, 90"C

tBu

:

30%

l~Bu ~ 7 1 / t Bu

P

S tBu

10%

Figures 16 and 17 Other examples of new cage structures.

56

theoretical concepts - it is now clear that the dirhodium complex can be for- mally considered as the ' inorganic ' ana- logue of the as yet unknown dication of cubane, [C8H8] 2+, which might there- fore be expected to have a similar 'dis- torted-cube' geometry.

Such new ideas are currently being tested by further theoretical calcula- tions [22], and so phospha-alkynes con- t inue to pose many unanswered ques- tions, even 30 years after their dis- covery.

References [1] Gier, T. E. J. Amer. Chem. Soc., 83,

1769, 1961. [2] Pitzer, K. S. J. Amer. Chem. Soc., 70,

2140, 1948. [3] Mulliken, R. S. J. Amer. Chem. Soc.,

72, 4493, 1950. [4] Reviews: Nixon, J. F., Chem. Revs., 88,

1327, 1988; Regitz, M. and Binger, P. Agnew. Chem. Int. Ed. Engl., 27, 1484, 1988; Regitz, M. and Scherer, O. J. Multiple Bonds and Low Coordination in Phosphorus Chemistry, Thieme, Stuttgart, 1990. Appel, R., Knoll, F. and R/tppert, I. Agnew. Chem. Int. Ed. Engl., 20, 731, 1981; Appel, R. Pure Appl. Chem., 59, 977, 1987; Kroto, H. W. Chem. Soc. Rev., 11,435, 1982.

[5] Cowley, A. H. Polyhedron, 3, 389, 1984; Accounts of Chem. Research, 17, 386, 1984; Ashe, A. J. Topics Current Chem., 105, 125, 1982; Scherer, O. J. Angew. Chem. Int. Ed. Engl., 24, 924, 1985.

[6] M~irkl, G. Angew. Chem. Int. Ed. Engl., 5, 846, 1966.

[7] Dimroth, K. and Hoffmann, P. Angew. Chem. Int. Ed. Engl, 3, 384, 1964.

[8] (a) Hopkinson, M. J., Kroto, H. W., Nixon, J. F. and Simmons~ N. P. C. Chem. Commun., 513, 1976. (b) Kroto, H. W., Nixon, J. F., Simmons, N. P. C. and Westwood, N. P. C., J. Amer. Chem. Soc., 100, 446, 1978.

[9] Appel, R. and Westerhaus, A. Angew. Chem. Int. Ed. Engl., 20, 197, 1981; Tetrahedron Letters, 2159, 1981.

[10] Becker, G., Gresser, G. and Uhl, W. Z. Naturforsch. B, 36, 16, 1981.

[11] Burckett St-Laurent, J. C. T. R., Hitch- cock, P. B., Kroto, H. W. and Nixon, J. F. Chem. Commun., 1141, 1981; same authors and Meidine, M. F. J. Orga- nometal. Chem., 238, C82, 1982.

[12] Hitchcock, P. B., Maah, M. J., Nixon, J. F., Zora, J. A., Leigh, G. J. and Bakar, M. A. Angew. Chem. Int. Ed. Engl., 26, 474, 1987.

[13] Regitz, M. Chem. Revs., 90, 191, 1990, and references therein.

[14] Hitchcock, P. B., Maah, M. J. and Nixon, J. F. Chem: Commun., 737, 1986.

[15] Binger, P., Milczarek, R., Mynott, R., Regitz, M. and R6sch, W. Angew. Chem. Int. Ed. Engl., 25, 644, 1986.

[16] Becker, G., Becker, W., Knebl, R., Schmidt, H., Weeber, U. and Wes- terhausen, M. Nova. Acta Leopold., 59, 55, 1985.

[17] Bartsch, R., Hitchcock, P. B. and Nix- on, J~ R. Chem. Commun., 1146, 1987.

p "~!

[18] Bartsch, R., Hitchcock, P. B. and Nix- on, J. F. J. Organometal. Chem., 356, C1, 1988; Chem. Commun., 819, 1988; J. Organometal. Chem., 373, C17, 1989.

[19] Wettling, T., Schneider, J., Wagner, O., Kreiter, C. G. and Regitz, M. Angew. Chem. Int. Ed. Engl., 28, 1013, 1989.

[20] Hu, D., Sch~iufele, Pritzkow, H. and Zenneck, U. Angew. Chem. Int. Ed. Engl., 28, 900, 1989.

[21] Bartsch, R., Hitchcock, P. B. and Nix- on, J. F. J. Organometal. Chem., 375, C31, 1989.

[22] Bartsch, R., Hitchcock, P. B. and Nix- on, J. F. Chem. Commun., 1046, 1989.

[23] Regitz, M., pers. comm., Summer 1990. [24] Hitchcock, P. B., Nixon, J. F. and

Sillett, G. J. D. (unpublished results). [25] Mingos, D. M. P., pers. comm., Spring

1990.

Figure 18 Cage structures can also include transi t ion metals, in this case rhodium.

57