Organometallics 1986,5, 245-252
245
An Examination of the Unsaturation in the Clusters
Tilman Jaeger," Silvio Aime," lband Heinrich Vahrenkamp" la Institut fur Anorganische und Analytische Chemie der Universitat, 0-7800 Freiburg, Germany, and Istituto di Chimica Generale ed Inorganica della Universiti, I- 10 125 Torino, Italy Received May 29, 1985
The clusters Fe4(CO)11(PR)2(1, R = t-Bu, Ph, Tol) were investigated chemically and NMR spectro(2,3). Subsequent scopically. They easily add one ligand L (CO, P(OMe)3,t-BuNC) to form Fe4(CO)11L(PR)2 (4). Further addiCO elimination under vacuum forms the new unsaturated clusters Fe4(CO)loL(PR)2 tion-elimination cycles allow the introduction of a total of four P(OMe)3and three t-BuNC ligands. The products with two, three, and four P(OMe), and with three t-BuNC ligands can only be obtained in the unsaturated form. lH and 13CNMR spectroscopy reveals the high fluxionality of the CO and t-BuNC ligands in these compounds. 'H,13C,and 31PNMR data of the unsaturated clusters can be interpreted in terms of an "aromatic ring current" in the Fe4 system. Crystal structure determinations of Fe4(CO)ll[P(OMe)3](PTol)2[3b, triclinic, Pi,a = 1300.0 (2) pm, b = 1506.9 (3) pm, c = 944.5 (2) pm, a = 100.65 (1)O, /3 = 107.35 (1)O, y = 84.43 (1)O, V = 1.7339 (3) nm3, 2 = 21 and Fe4(CO)lo[P(OMe)3](PTol)2 [4b, orthorhombic, P212121,a = 1289.0 (8) pm, b = 1478.5 (4) pm, c = 1819.7 (3) pm, V = 3.4678 (8) nm3, 2 = 41 reveal the difficulty of locating a Fe-Fe double bond in the unsaturated systems and the steric strain in the ligand sphere of the saturated systems. Introduction Cluster unsaturation is one of the simple conditions for the investigation of the cluster-surface analogy. The number of stable unsaturated organometallic clusters is, however, relatively small; and, with the exception of HzOs3(CO)lo,few studies on the nature of the unsaturation have been made.2-4 In simplest terms, the chemical consequences of cluster unsaturation should be related to the chemistry of multiple bonds in general. But the disputable nature of metal-metal multiple bonds in clusters5 and the hidden unsaturation due to donor-acceptor metal-metal bonds6cause a more diversified picture which still lacks systematization. In the course of our cluster construction studies we found a group of tetrairon clusters Fe4(CO)11(PR)2(1)' which according to the l&electron rule lack two electrons. The unsaturation of these compounds was obvious from the ease of CO addition. The location of a Fe-Fe double bond proved to be difficult since Fe-Fe bond lengths (244-269 pm) are not significantly different and the shortest Fe-Fe bond is CO bridged. A delocalized 2a situation in the planar tetrairon ring could therefore not be excluded. In order to better understand the unsaturation in the clusters 1, we undertook a chemical and NMR spectroscopic investigation. The aim was to learn about the double bond fluctuation or possible delocalization, the interconvertibility of saturated and unsaturated cluster forms when donor ligands have been introduced, and the isomerism and bonding situation in derivatives with more than one donor ligand. Parts of this work have been published as a communication.* (1) (a) Universitiit Freiburg. (b) Universiti di Torino. (2) Cf. Johnson, B. F. G.; Lewis, J. Adu. Inorg. Chem. Radiochem. 1981, 24, 225. (3) Cf. Vahrenkamp, H.Adu. Organomet. Chem. 1983,22, 169. (4) Muetterties, E. L.; Krause, M. J. Angew. Chem. 1983, 95, 135; Angew. Chem., Int. Ed. Engl. 1983,22, 135. (5) Granozzi, G.; Benoni, R.; Tondello, E.; Casarin, M.; Aime, S.; Osella, D.Inorg. Chem. 1983, 22, 3899. (6) Mayr, A.; Ehrl, W.; Vahrenkamp, H. Chem. Ber. 1974,107, 3860. (7) Vahrenkamp, H.; Wucherer, E. J.; Wolters, D. Chem. Ber. 1983, 116, 1219. (8) Vahrenkamp, H.; Wolters, D. Organometallics 1982, 1 , 874.
0276-733318612305-0245$01.50/0
Experimental Section All experimental techniques were as previously reported? The clusters la and IC were prepared according to the published procedure.7 Reagents were obtained commercially,distilled under nitrogen before use, and applied as 0.5 M solutions in benzene. Chromatographicseparations were done on silica gel (0.063-0.200 mm) which was dried for 6 h at 180 "C vacuum. All new compounds are characterized in Table I. IR spectra were measured of cyclohexane solutions on a Perkin-Elmer 782 spectrometer. 'H NMR spectra were obtained on an Varian T60 A spectrometerusing CDC13as a solvent and Me,Si as internal standard. 13C and 31PNMR spectra were obtained on a JEOL GX 270189 spectrometer operating at 67.9 and 109.4 MHz, respectively. The 13C measurements were carried out on 13CO-enriched(ca. 10%) samples which were prepared by exchange between the natural abundance Fe4(CO)11(PR)2 samples and 13C0 (Monsanto, 90%) in n-hexanelCHCl, solutions contained in sealed vials (11 days at +30 "C). The variable-temperature 13C spectra of la and 2a were recorded also at higher field (JEOL GX-400 operating at 100.6 MHz for 13C). Improved Preparation of Fe4(CO)11(PPh)z(lb). Fez(CO)6(PHPh)27 (0.62 g, 1.24 mmol) and Fe3(C0)12(1.35 g, 2.68 mmol) in benzene (250 mL) were irradiated by a water-cooled 150-W high-pressure Hg lamp (type Hanau TQ 150 Z 3) immersed into the reaction vessel in which a circulating flow of the reactants was maintained. The reaction vessel was connected to a 500-mL flask, evacuated,and heated to 60 "C. After 3.5 h the solution was fiitered and evaporated to dryness. The residue was subjected to flash chromatography (20 X 4 cm column, excess pressure 0.10-0.15 bar) with hexane. A green band of Fe3(C0)12 and an were quickly eluted. Hexane/ orange band of Fe3(CO)9(PPh)2 benzene (101) eluted lb as a dark brown band. Filtration and recrystallization from hexanelbenzene (51) yielded 300 mg (32%) of lb. Preparation of Fe4(CO)ll[P(OMe)3](PPh)2 (3a). lb (90 mg, 0.12 mmol) in benzene (15 mL) was treated dropwise with P(OMe)3 in benzene (0.5 M, 0.24 mL, 0.12 mmol). Immediate
reaction was obvious from a color change from black to red. The solution was concentrated to 5 mL in a stream of CO, 10 mL of hexane was added,and 3a was crystallized at -30 O C ; yield 90 mg (86%).
Preparation of Fe4(CO)11[P(OMe)3](PTol)2 (3b): as above from IC(70 mg, 0.09 mmol) and P(OMe)3(0.18 mL, 0.09 mmol); yield 70 mg (86%). (9) Muller, R.;Vahrenkamp, H. Chem. Ber. 1980, 113, 3517.
0 1986 American Chemical Society
246 Organometallics, Vol. 5, No. 2, 1986
Jaeger et al.
Table I. Characterization of New ComDounds no. 3a
3b 3c" 4a
4b 4ca 6ab 6b 6c"
6d" 7a' 7d"sd 8a
color red-black red-black red-brown black black black black black black black black black black
mu. "C >50 dec >60 dec >75 dec 118 >90 dec 178 >160 dec >120 dec >155 dec 118 >130 dec >lo0 dec 122 dec
C C Z ~ H ~ S F ~(871.8) ~ O I ~ P ~ 35.82 (35.61) 37.37 (37.57) CmHZ3Fe4Ol4P3 (899.8) 40.48 (39.77) CzsHlSFe4NOllPz(830.8) 35.58 (35.48) C25H1sFe4013P3 (843.7) 37.19 (37.19) C27H23Fe4013P3 (871.8) C Z ~ H ~ S F ~ (802.8) ~ N O ~ ~ P40.39 ~ (39.45) 34.51 (34.62) Cz7HZsFe4Ol5P4 (939.8) 35.99 (36.18) C2sH32Fe4015P4 (967.9) 43.40 (43.29) C31H28Fe4N~O~P2 (857.9) 38.75 (38.83) CZsHBFe4NOl2P3 (898.8) 33.63 (33.76) C2sH37Fe4017P5 (1035.8) C31H37Fe4N014P4 (994.9) 37.42 (37.70) 32.89 (33.10) C31H4sFe401sP6 (1131.9) formula (mol wt)
anal. calcd (found) H 2.20 (1.93) 2.58 (2.43) 2.30 (1.78) 2.26 (1.96) 2.66 (2.58) 2.38 (1.92) 3.00 (2.88) 3.33 (3.30) 3.28 (2.54) 3.13 (3.03) 3.60 (3.20) 3.74 (3.84) 4.00 (3.99)
Fe 25.63 (26.41) 24.83 (24.51) 26.88 (26.54) 26.47 (26.38) 25.63 (25.76) 27.82 (27.94) 23.77 (23.82) 23.08 (21.82) 26.03 (25.81) 24.85 (25.03) 21.56 (21.68) 22.45 (22.38) 19.73 (19.59)
N analyses calcd (found): 3c, 1.68 (1.60); 4c, 1.74 (1.82);6c, 3.26 (2.90); 6d, 1.55 (1.69);7d, 1.40 (1.54). *Mol wt 940 (FD-MS). Mol wt 1036 (FD-MS). dMol wt 995 (FD-MS). a
Preparation of Fe4(CO)l,(t-BuNC)(PPh)2 (3c). lb (100 mg, 0.13 mmol) in benzene (50 mL) was treated dropwise with t-BuNC in benzene (0.5 M, 0.26 mL, 0.13 mmol). The reaction was complete within a few minutes. After filtration the solvent was removed under vacuum and the residue recrystallized from npentane/dichloromethane (52) a t -30 "C to yield 98 mg (88%) of 3c. Preparation of Fe4(CO)lo[P(OMe)3](PPh)z (4a). 3a (150 mg, 0.17 mmol) in benzene (30 mL) in a 250-mL flask was evacuated and heated to 60 "C for 1 h while the color changed from red to black. After filtration the solvent was removed under vacuum and the residue recrystallized from hexane/benzene (4:l) a t -30 "C to yield 130 mg (91%) of 4a. Preparation of Fe4(CO)lo[P(OMe)3](PTol)2 (4b). 3b (50 mg, 0.055 "01) in benzene (10 mL) was treated as above to yield 45 mg (94%) of 4b. Preparation of Fe4(CO)lo(t-BuNC)(PPh)2 (4c). 3c (90 mg, 0.10 mmol) in benzene (25 mL) was treated as above to yield 57 mg (71%) of 4c. Conversions 4 3. Solutions of 4a, 4b, and 4c (ca. 30 mg) in benzene (ca. 15 mL) were stirred under an atmosphere of CO for 1day when IR and NMR evidence indicated complete conversion t o 3a, 3b, and 3c. The isolated yield of 3a from 4a (34 mg) after crystallization from hexane/benzene (51) at -30 "C was 28 mg (80%). Preparation of Fe,(CO),,(t - B U N C ) ~ ( P P (5c). ~ ) , 4c (110 mg, 0.14 mmol) in benzene (50 mL) a t 8-10 "C was treated dropwise with t-BuNC in benzene (0.5 M, 0.28 mL, 0.14 mmol) while the color changed from black to red. Evacuation to dryness and recrystallization from hexane/toluene (51) at -70 "C yielded 19 mg (18%)of 5c. The low yield of the compound allowed only a spectroscopic identification. Preparation of Fe4(CO)9[P(OMe)3]2(PPh)2 (6a). 4a (66 mg, 0.08mmol) and P(OMe)3(0.5 M, 0.16 mL, 0.08 mmol) in benzene (25 mL) were heated t o 60 "C for 1 h. After filtration and evaporation to dryness, the residue was recrystallized from hexane/benzene (4:l) at -30 "C to yield 65 mg (88%) of 6a. Preparation of Fe4(CO)9[P(OMe)3]2(PTol)2 (6b). IC (70 mg, 0.09 mmol) and P(OMe)3 (0.5 M, 0.36 mL, 0.18 mmol) in benzene (15 mL) were heated t o 60 "C for 30 min. Filtration, evaporation to dryness, and recrystallization from hexane a t -30 "C yielded 70 mg (80%) of 6b. Preparation of Fe4(C0),(~ - B U N C ) ~ ( P P (6c). ~ ) *4c (260 mg, 0.32 mmol) in benzene (50 mL) was treated with t-BuNC in benzene (0.5 M, 0.64 mL, 0.32 mmol). The solution was stirred for 1.5 h a t room temperature and for 1.5 h a t 65 "C. After evaporation to dryness, the residue was chromatographed with hexane/benzene (5:l) over a 2 X 20 cm column. From the first black band after evaporation and recrystallization from n-pentaneltoluene (5:l) a t -30 "C 46 mg (16%) of 6c was obtained. Preparation of Fe4(CO)S[P(OMe)3]( ~ - B U N C ) ( P P(6d). ~)~ 4c (130 mg, 0.16 mmol) and P(OMe)3 (0.5 M, 0.32 mL, 0.16 "01) in benzene were stirred a t room temperature for 1 h. The color of the solution turned from black to red. Upon concentration under vacuum the solution became black again. To complete the CO elimination, the remaining solution (40 mL) was heated to
-
60 OC for 1 h. Filtration, evaporation t o dryness, and crystallization from n-pentaneldichlormethane (4:l) at -30 "C yielded 57 mg (40%) of 6d. Preparation of Fe4(CO)8[P(OMe)3]3(PPh)2 (7a). 6a (73 mg, 0.08 mmol) and P(OMe)3(0.5 M, 0.16 mL, 0.08 mmol) in benzene (40 mL) were heated a t reflux for 1.5 h. Filtration, evaporation to dryness, and crystallization from hexane/dichloromethane (4:l) yielded 66 mg (82%) of 7a. Preparation of Fe4(CO),(~ - B U N C ) ~ ( P(7c). P ~ )6c ~ (28 mg, 0.03 mmol) and t-BuNC (0.5M, 0.06 mL, 0.03 mmol) in benzene (25 mL) were heated at 60 "C for 1.5 h. After evaporation to dryness, the residue was chromatographed with hexane/benzene (51)over a 2 X 15 cm column. From the first dark brown band 15 mg (55%) of 7c remained as an oily material which could not be purified by crystallization.
Preparation of Fe4(CO)8[P(OMe)3]2(t-BuNC)(PPh)2 (7d). and P(OMe)3(0.5 M, 0.06 mL, 0.03 mmol) 6d (28 mg, 0.03"01) in benzene (40 mL) were heated a t reflux for 3.5 h. Filtration, evaporation to dryness, and crystallization from n-pentaneldichloromethane (2:l) yielded 12 mg (39%) of 7d. Preparation of Fe4(CO)7[P(OMe)3]4(PPh)2 @a). 7a (49 mg, 0.05mmol) and P(OMe)3(0.5 M, 0.10 mL, 0.05 mmol) in benzene (25 mL) were refluxed for 6 h. After filtration and evaporation to dryness, the residue was recrystallized twice from hexane/ dichloromethane (5:1), resulting in 27 mg (52%) of 8a. Crystal Structure Determinations. The quality of the crystals was checked by photographic methods. All crystal data and details of the structure solutions are given in Table 11. The positions of the iron atoms were found from Patterson maps. Two successive Fourier cycles then revealed the positions of all other non-hydrogen atoms. The structures were refined by standard least-squares techniques using unit weights. The CsH4 units of the tolyl groups were treated as rigid bodies with C-C = 139.5 pm and C-H = 108 pm, only the positions of the rigid bodies and the thermal parameters of the C atoms being varied while the H atoms were assigned a common variable-temperature factor. All atoms except the aromatic C and H atoms were refined anisotropically. Tables I11 and IV list the atomic positions and U(eq) or U(ij)values. The supplementary materid contains, in addition, the anisotropic temperature factors, all bond lengths and angles, and the F J F , lists.
Results and Discussion Reactions. The m a t u r a t i o n of the clusters 1 manifests itself i n the easy CO addition t o form t h e clusters 2.' The latter, however, are n o t significantly more stable and are quantitatively reconverted t o 1 under vacuum. The color change between black (1) a n d red (2) m a k e s i t easy to follow these reaction^.^ It was now found that PX3and RNC ligands are a d d e d the same way. F r o m reactions of lb and ICwith P(OMe)3 and t-BuNC the new s a t u r a t e d clusters 3 are obtained. H e a t i n g of these in v a c u u m converts t h e m back t o uns a t u r a t e d clusters b y elimination of CO. The resulting
Organometallics, Vol. 5, No. 2, 1986 247 R
R
R
R
20, R= t - B u b, R=Ph
la, R = t - B u b. R = P h c . R=Tol
c. R=Tol
clusters 4 can be called substitution derivatives of 1,just as the clusters 3 are substitution derivatives of 2. Reconversion of 4 to 3 occurs readily under a CO atmosphere. Again, the saturated compounds 3 are red and the unsaturated compounds 4 are black in solution. R
R
R
R
3a, R=Ph, L=P(OMe),
b, R=Tol. L=P(OMe), C.
R=Ph, L =t-BuNC
4a, R=Ph. L=P(OMe), b, R=Tol. L=P(OMe), C, R=Ph, L = t - B u N C
The next cycle of ligand additions should produce saturated clusters of the type Fe4(CO)l&L,(PR)z starting from 4. However, only in the case of the bis(isocyanide) compound 5c, which was not analytically pure, was the isolation of such a product possible. Heating of the reactants 4c and t-BuNC caused the reaction to proceed directly to the unsaturated product 6c. Likewise, a red, probably saturated intermediate could be observed but not isolated in the reaction between 4c and P(OMe)3which led to the isolation of 6d. The phosphite substituted clusters 4a and 4b did not react fast enough with further phosphite at room temperature which caused them to be directly converted to 6a and 6b upon heating. These conversions are net substitution reactions the addition/elimination nature of which is obvious. Fe4(CO),o(t-BuNC),(PPh),
5c
Fe4(CO),L2(PR),
Sa. b. c. d.
R=Ph, L=P(OMe), R=Tol. L=P(OMe), R=Ph. L=t-BuNC R:Ph. L,=t-BuNC
+
P(OMe)3
The introduction of a third and fourth donor ligand was only possible as a substitution reaction due to the more drastic reaction conditions required. The higher isocyanide derivatives of 1 seem to be progressively more unstable since 7c could only be obtained impure and a tetraisocyanide compound was not accessible. The tris(phosphite)-substituted cluster 7a is, however, moderately stable, the tetrakis(ph0sphite)-substituted cluster 8a could be obtained under conditions close to decomposition,and the mixed phosphite-isocyanide derivative 7d was accessible as well. Further reactions with these donor ligands could not be observed. Fe4(C%!,L,!PPh),
Fe4(CO),CP(OMe)314(PPh)2
7a. L=P(OMe),
b,
L =t-BuNC
C.
L3=t-BuNC
Sa
+
two tris-substituted derivatives (7a,d) were established by FD-MS. The MS method failed on the tetrakis(phosphite) derivative 8a. On the basis of the IR spectroscopic similarity within the groups of clusters and the correct intensity ratios in the IH NMR spectra (see Table V) all new clusters 3-8 can be considered as confirmed. The IR spectra of the clusters, like their colors, allow a clear distinction between the saturated and the unsaturated compounds (cf. Table V). The presence of 11 terminal ligands around the tetrairon core goes along with one CO being in a bridge position. All unsaturated clusters show this by v(C0) bands around 1800 cm-'. As the number of donor ligands increases from 2 to 8, so the general pattern of v(C0) absorptions moves to lower wavenumbers. All isocyanide-substituted compounds here show the u(CN) absorption. Its position (between 2130 and 2150 cm-') is hardly shifted with respect to the u(CN) band of free t-BuNC (2135 cm-l), as might be expected for compounds of metals in relatively low oxidation states.lOJ1 The 'H NMR spectra of 3-8 (cf. Table V) give some insight into possible ligand arrangements. There is a large number of possible isomers for all new clusters. Extrapolating from the CO positions in 1 and 2 (and taking into account the molecular structures of 3b and 4b), the additional ligands can be axial or equatorial with respect to the tetrairon ring. They can be adjacent to or distant from the CO bridge (respectively the Fe-Fe double bond), and with more than one P(OMe)3or t-BuNC ligands these can be placed in gem, vic, cis, trans, ortho, and para positions. On this background most of the 'H NMR spectra are surprisingly simple. They indicate that the clusters 3 and 4 exist as just one isomer. They allow the interpretation that the bis(isocyanide) compounds 5c and 6c either contain equivalent t-BuNC groups or are fluxional molecules. And they seem to prove the symmetrical configuration of 8a with four equivalent phosphite ligands. The crystal structure determination of 4b has located the P(OMe):, ligand nonadjacent to the CO bridge. This was taken into account in the formula drawing of all clusters 4, whereas in the absence of further information no detailed ligand arrangements can be suggested for 5-8. Since two ligands L on one iron atom are unlikely and since axial/equatorial isomerism or the position of the CO bridge seem to be without effect in the clusters 3 and 4, it is realistic to take into consideration only the possibility of ortho and para arrangements of the two L in the clusters 5 and 6. Whereas no information relating to this can be obtained from the NMR spectra of 5c and 6c (see above), 6a and 6b seem to be mixtures of isomers. In the P(OMe), region they show two signal groups marked A and B in Table V whose intensity ratio is approximately 2:1 in 6a and 3:2 in 6b. The same holds for the 31Pand 13C resonances of 6a (see below). If the signals marked A are assigned to the ortho isomer (P(OMe), on adjacent Fe atoms) and the signals marked B to the para isomer, then the relative abundances of A and B correspond roughly to the statistical chance of their formation: for the P(OMe),-substituted iron atoms in 4 there are two ortho iron atoms and one para iron atom. Just like the bis(isocyanide) derivatives 5c and 6c the mixed phosphite-isocyanide derivative 6d does not give NMR evidence for isomerism. Again a possible explanation for this is fluxionality; and since the phosphite ligands seem to be firmly
2P(OMe)3
Constitution and Isomerism of the Products. The identity and geometry of the compounds lc,' 3b, and 4b (see below) is confirmed by crystal structure analyses. In addition, the constitutions of one bis-substituted (sa) and
(10)Treichel, P.M. Adu. Organomet. Chem. 1973, 11, 21. (11) Adams, R. D.; Chodosh, D. F.J.Am. Chem. SOC.1977,99,6544. (12) Cf.Grant, S.; Newman, J.; Manning, A. R.J. Organomet. Chem. 1975,96, C11. Adams, R. D.; Cotton, F.A. J . Am. Chem. SOC.1973,95, 6589,6594.
248 Organometallics, Vol. 5, No. 2, 1986
Jaeger et al.
Table 11. Crystallographic Data for Fel(CO)le(P-p-Tol)z*P(OMe)s (4b) and Fe4(CO)II(P-p-Tol)z*P(OMe)3 (3b) 4b 3b mol formula mol wt cryst from color space group cell dimens (20 "C) Q , pm b, pm c , pm a,deg P, deg 72 deg V , nm3 Z dobsd, g cm-3 desl!d g cm-3 radiatn monochromator cryst dimens, mm 1,cm-' absorptn correctn diffractometer data collection 28 range reflectns measd obsd data, 1 > 3u(Z) structure soln programs scattering factors R(unit weights) residual electron density (106e.pm-3)
C27H23Fe4013P3
C28H23Fe4014P3
871.8 toluene/hexane (1:2) black P2A21
Pi
1289.0 (8) 1478.5 (4) 1819.7 (3) 90 90 90 3.4678 (8) 4 1.70 1.67 Mo K u graphite 0.15 X 0.15 X 0.30 17.5 none Nonius CAD 4 w-2e 1-45 2944 1782 Patterson/Fourier
1300.0 (2) 1506.9 (3) 944.5 (2) 100.65 (1) 107.35 (1) 84.43 (1) 1.7339 (3) 2 1.70 1.72 Mo K a graphite 0.20 X 0.25 X 0.50 16.8 none Nonius CAD 4 0-28 1-45 4832 3700 Patterson/Fourier
SHELX"
SHELX"
Cromer/Mannb 0.045 +0.5/-0.4
Cromer/Mannb 0.055 +1.1/-0.9
899.8 benzene/hexane (1:2) red
ODistributed by G. M. Sheldrick, University of Gottingen. West Germany. bCromer, D. T.; Mann, J. B. Acta Crystallogr., Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1968, A24, 321.
bound, this must be due to isocyanide mobility. We have not been able so far to enrich or isolate the different isomers of the clusters 5 and 6. With the assumption of one PR3 or RNC ligand per iron atom, the trisubstituted clusters 7a and 7c cannot form different isomers. This is borne out for 7a by the exact 2:l ratio of the NMR signals due to the P(OMe), ligands. For 7c there are two t-BuNC resonances of approximately 1:l intensity which cannot be interpreted nor discussed at the moment since the compound is not analytically pure. The mixed derivative 7d behaves like the bis(phosphite) derivatives 6a and 6b showing signal groups A and B with intensity ratio ca. 2:l in the P(OMe), and t-BuNC spectral regions. Again, mobility of the isocyanide ligand offers the simplest explanation for this. Fluxionality and NMR Anisotropy. To gain some insight into the possible orientation of the ligands and the location of the CO bridge and the assumed Fe-Fe double bond in the unsaturated clusters, some of the clusters were also examined by 31Pand 13C NMR spectroscopy. Table VI lists the results which for 6a and 7a confirm the isomeric situation discussed above. It became obvious immediately from the I3C NMR spectra in the CO region that the basic clusters 1 and 2 are highly fluxional. There is just one resonance which down to the lowest accessible temperatures (ca. -130 OC) and at the high field strength (400-MHz instrument) does not significantly broaden. This situation does not change significantly upon introduction of one to four P(OMeI3 ligands. It has not been possible to observe more than one 13C0resonance in 3a, 4a, 6aA, 6aB, 7a, and 8a. Although the phosphite ligands slow down the fluxional process as evidenced from the collapse of the 13C0signals of 6aA and 8a around -110 "C, no "frozen out" situation could be reached. The room-temperature 13C0signals show multiplets with J(C-P) coupling of ca. 6 Hz according to the
number of phosphite ligands present. The coupling to the p4 phosphorus atoms seems to be negligible (