Activation of Hydrogen by a Transition Metal Complex at Normal

Activation of Hydrogen by a Transition Metal Complex at Normal Conditions Leading to a Stable Molecular Dihydride. L. Vaska, and John W. DiLuzio. J. A...
0 downloads 3 Views 227KB Size
Feb. 20, 1962

COMMUNICATIONS TO THE EDITOR

679

ACTIVATION OF HYDROGEN BY A TRANSITION METAL COMPLEX AT NORMAL CONDITIONS LEADING TO A STABLE MOLECULAR DIHYDRIDE

Sir: The discovery of a quadricovalent iridium(1) complex, [IrCl(CO)(Ph3P)a],1has afforded access to novel chemical reactions which are believed to offer new opportunities for understanding and study of certain vital processes of catalysis. The compound responsible for this development shows a striking reactivity a t normal conditions toward a large number of molecules (e.g., hydrogen, oxygen, sulfur, organic halides, organic acids, hydrides of group VIIA, VIA elements, etc.). Yet, it displays properties which are remarkably well suited for a convenient investigation of its chemical behavior: its crystals are durable in air and stable thermally' ; its relevant reactions represent additions t o the quadricovalent species t o yield octahedral complexes of Ir(II1) (cf. equations 1, 5). These reactions are accompanied by characteristic CH,C shifts of the stretching frequency of the coordinated carbon monoxide (to higher v) in the infrared spectrum. Indeed, the (only one) CO group in the molesmall fraction of the product mixture, the amount of cule provides a singular means of detecting reacmethylethylphenylphosphine oxide formed in con- tions, following their course, and inquiring into the junction with the cyclopropane derivative was also nature of the products through the extent of Avco. small. On the basis of Denney's it is In this communication we report a reaction with conceivable that the phosphorus atom might be in- molecular hydrogen, and some related observations verted, might retain the same configuration or which contribute to the understanding of metalundergo both stereochemical transformations by hydrogen interactions in this and other systems. competitive paths. In fact, one or more of the When hydrogen (one atrn., 25') is bubbled many possible intermediates exemplified by V, VI, through a solution of [IrC1(CO)(Ph3P)2]in benzene, VI1 or VI11 should be involved in step (2) of the color of the initially yellow solution (2 X Denney's mechanism. M ) becomes lighter, followed by the appearance of As shown in the accompanying c o m m ~ n i c a t i o n , ~yellow-white ~ crystals ( 2 hr.). After allowing it t o methylethylphenylphosphine and benzylaceto- stand under hydrogen (same conditions) for one phenone arise from IV by way of the transition state day a 55% yield of the crystalline product is obIX. The fact that the phosphine is racemic is un- tained.2 Infrared spectrum (Table I), elemental doubtedly the result of the thermal racemization of analysis, molecular weight data, non-electroIytic the optically active phosphine (presumably formed character (in nitrobenzene and acetone), diamagnewith retention of configuration of the phosphorus tism, and reactions of the compound (see below) atom) a t the reaction temperature of 190-200°.16~16suggest that it is best formulated as a molecular dihydride of tervalent iridium, [IrH2C1(CO)(Ph3P)2]. CHj CsH5 (Anal. Calcd. for IrClP2C3,H320: Ir, 24.6; C1, 168 I 4.5; P, 7.9; C, 56.8; H, 4.12; 0, 2.05; mol. wt., CH,CH,-P---- -CH-,CHz 782. Found: Ir, 24.8; C1, 4.5; P, 7.9; C, 56.5; H, 4.15; 0, 2.04; mol. wt. in benzene, 772.) [IrrC1(CO)(Ph,P)2]

+ H1 250

[Ir111H~CL(CO)(Ph,P)2] (1)

Acknowledgment.-This research was supported in part by a grant from the Petroleum Research Fund administered by the American Chemical Society. Grateiul acknowledgment is made to the donors of the fund.

Treatment of [IrCl(CO) (Ph3P)2] (in benzene) with deuterium3 gives the corresponding dideuteride, [IrD2C1(CO)( P ~ Q P )showing ~], an X-ray diffraction pattern identical with that of the dihydride (according to visual examination). Two noteworthy isotope effects are evident from the infrared spectra DEPARTMENT OF CHEMISTRY WILLIAME. MCEWEX OF KANSAS ARTHURBLADB-FONT (Table I ) : YCO in the dideuteride lies markedly UNIVERSITY A. VANDERWERF higher than in the dihydride; the crystals of the LAWREXCE, KANSAS CALVIN dideuteride apparently exhibit only one VIrD while RECEIVED DECEMBER 21, 1961 (14) W. E. McEwen and A. P. Wolf, J . A m . Chrm. SOC..84, 676 (1962). (15) L. Horner, H. Winkler, A. Rapp, A. hlentrup, H . Hoffman and P. Beck, Tefrahedron Lrffers, 161 (1961). (16) D. C. Velez, J. W. Johnson, C. A. VanderWerf and W. E. McEwen, unpublished observations.

(1) L. Vaska and J. W. DiLuzio, J . A m . Chcm. Sac,, 83, 2784 (1961). (2) In CHCli the rate of the reaction with Ha is faster, but the purity of the product has been found to suffer. A partial interaction

has been observed also between the crysfals of [IrCl(CO)(PhrP)al and hydrogen (1 atm., 25'). (3) Courtesy of Dr. W. Keith HaLL

COMMUNICATIONS TO THE EDITOR

080

T.4BLE

XATURE

O F THE

METAL-HYDROGEN BOSD. Compound (P = Ph3P)

Formed by ream. with

1

ORIGIN O F HYDROGEN AND INFRARED SPECTRCM (CM.-')' W.H,D*

Solid c

[IrH2C1(CO)P?] Pt-Hh [IrDzCl(CO)Pz] Pt-Dh [IrHC12(CO)P2]' [ IrHC12P3] [IrH2C1P3lh [IrH3P3I1

Vol. 84

2190, 2110, 1571 1515, 2245 2200 2215,

2100 2060

So1n.d

-

2110

...

--e-

Solid c

S o hd

2220,f 2095

1970

...

... 2030

1983O ... 2030 ... 2045

...

... ...

1572i 1480

pyco

1995

...

... .. 2185 2180 2110, 1745

... ...

...

a f 5 cm.-1; NaCl optics. * Strong to medium-strong. Halocarbon or Nujol mull. I n C8Hs or CHCL. e Very Chemisorbed hydrogen and deuterium on y-alumina-supported platinum metal, data from strong, f Weak. g Broad. Ref. 6. ref. 4. i Medium-weak, broad. i Ref. 1. Ref. 5.

the dihydride shows two VI,H values. Structures of these compounds and the course of their formation are being studied. The dihydride is stable in air but sensitive to light (in air). I n vacuum at 25' (pumping a t l o V 5mm. for 16 hr.) the compound remains unchanged (m.p. (dec. in vac.), 315-321'). The compound reacts with hydrogen chloride (in ether) t o yield hydrogen7 and a monohydrido complex.

+

+

[IrT1*HCl2(CO)(Ph3P),] Ht

(6)

!m

h 1r"'H2Cl(CO)(Ph,P),]

[IrI1IHpC1(CO)(Ph3P),1 HC1 --f 25'

[IrTlrC U ) ( P h ~ P j 2 ~ - -[Ir''' ~ €ICI,(COj(Ph,P),]

(2)

Equivalence of H2, HCL, and C12in the primary reacThis monohydride is the same compound which was tions indicates that they allincluding hydrogen-must first obtained by the reaction of [IrC1(CO)(Ph,P)z] be considered as Lmis acids (or oxidants) in these with HC1.I We had suggested that the latter may reactions, that i s to say, accepting electrons from unibe interpreted as involving formal oxidation of Ir- oalent iridium for covalent bond formation. The (I) t o Ir(III).l This view now becomes substan- data reproduced in Table I show that the metalhydrogen bond character in [IrH2Cl(CO)(Ph,P),] tiated by this further evidence. is clearly of the same nature as in covalent iridiumWhen a suspension of (the very insoluble) [IrHClz(C0) (Ph3P)2] is refluxed in a saturated solution (111) hydrides prepared by reactions with protonic of HC1 in 1,2-dimethoxyethane for several days, acids, alcohols,5 and metallic hydrides.'j A significant observation in relating homogeneous conversion t o a trichloro complex is observed, acand heterogeneous systems emerges by comparing companied by evolution of hydrogen.? our results with the recent work by Pliskin and Eischens4 describing the first tangible evidence for [Ir111HClz(CO)(Ph3P)2] HCl 84 covalent bonding between the surface atoms of [Ir111C4(CO)(Ph3P),I HZ (3) (alumina-supported) platinum and chemisorbed ( A n a l . Calcd. for IrC13P2Ca7HaoO: Ir, 22.6; C1, hydrogen (as atoms). This comparison (Table I) i2:5; P;-7.3;-L";52;2;. FT.3.6; 9;i.9:- Found:. ir, d.emansirates a rlm-anahgy between a transition 23.1; C1, 12.5; P, 7.3; C, 51.9; H, 3.7; 0, 2.0; metal compound and a transition metal surface with respect t o their reactions with Hz and Dz. V C O , 2080 cm,-l; diamagnetic; non-conducting in acetone and nitrobenzene.) This product is ob- Another aspect of this analogy becomes evident by considering the action of halogens, hydrogen halides tained also by these reactions (cf. equation 6), hydrogen sulfide,8 and oxygens on [IrCl(CO)(Ph3P)2]. These reagents "poison" [ I ~ " I H C ~ Z ( C O ) ( P ~ ~+P )Clz . L --f ~ 250 the "active sites" of the Ir(1) complex by destroy[Ir111C13(CO)(Ph,P),1+ HC1 (4) ing its "surface" through oxidation. (8) Unpublished results. [IrlC1(CO)(PhsP)?]t CIS--+ [Ir1I1Cl3(CO)(Ph3P)2]

+

+

25'

(5)

The interrelations of the reactions described above become evident in this summary. (4) W. A. Pliskin and R. P. Eischens, 2. p h y r i k . Chem. N . F., 84, 11 (1960). (5) L. Vaska, J . A m . Chem. SOC.,83, 756 (1961). (6) R. G.Hayter, dbid., 83, 1259 (1961). (7) Determined b y gas chromatography.

L. VASKA MELLONINSTITUTE JOHN W. DILUZIO PITTSBURGH, PA. RECEIVED JANUARY 6, 1962 COORDINATION CHAIN REACTIONS

Sir : The exchange reaction between the complex ions triethylenetetramine-nickel(11))NiT+z, and ethyl-