- Email: [email protected]
The raman spectrum and structure of dicyclopentadienylberyllium
293
of Organometallic
Journal
@ Ekevier
Sequoia
S-A.,
Chemistry, Lausanne
-
170 (1979) 293-297 Printed in The Netherlands
THE RAMAN SPECTRUM AND STRUCTURE DICYCLOPENTADIENYLBERYLLIUM
J. LUSZTYK
Institute 00-662
and K.B.
STAROWIEYSKI
of Organic Chemistry Warsaw (Poland)
(Received
November
OF
15th,
and Technology,
Warsaw
Technical
University
(Politechnika),
1978)
summary The Raman spectra of solid and liquid dicyclopentadienylberyllium are described_ They suggest the presence of two n-bonded rings in the molecule, one pentahapto and one polyhapto coordinated to the Be atom.
Introduction In 1951 the first metallocene, CpzFe, was discovered [l], and shown to have the sandwich structure of Dsd symmetry (Fig. 1A). Dicyclopentadienylberyllium, Cp,Be, the smallest metallocene, was first synthesized by Fischer and Hoffmann in 1959 [2]. In 1964 Haaland et al. [3,4] concluded from electron diffraction studies that in the gas phase Cp,Be had a structure of Csv symmetry (Fig. lB), highly unusual for metallocenes. On the other hand, Wong et al. [ 5, 61 showed by X-ray diffraction studies that in the solid state Cp,Be had a slip sandwich structure_ Wong et al. [ 53 proposed a G-K bond system in the molecule at low temperatures (Fig. 1C). However, Drew and Haaland [ 71 interpreted Wong’s results differently, suggesting the presence of two x-bonded rings in the molecule, one pentahapto and one polyhapto coordinated to the Be atom (Fig. 1D). The structure of Cp,Be was also investigated by means of vibrational spectroscopy. Fritz et al. [S,9,10] concluded from the IR spectrum of Cp,Be in the solid state that the molecule contained two equivalent centrally o-bonded rings. After the structure of Csu symmetry has been proposed for the gas phase [ 3,4], Fritz and Sellmann [ 111 and McVicker and Morgan [12] suggested the same structure for the solid state and solutions on the basis of new IR data. McVicker and Morgan found the model (CpBe)’ Cp- to be the most consistent with their IR data for solid Cp,Be. The IR spectra mentioned above are not consistent_ Furthermore, no papers
%,
C 5v
Fig_ 1. Possible structures of Cp2Be.
concerned with the vibrational spectra of Cp,Be have appeared since the slip sandwich structure was established by X-ray diffraction [ 5,6]. Thus, there is a disagreement between X-ray and IR results. Moreover, the Raman spectrum of this compound has not been reported_ We thus decided to undertake measurements of Raman spectra of Cp,Re in order to remove the inconsistency between vibrational and X-ray data and provide a new approach to the question of the bonding in this compound_ Results and discussion The Raman spectra of solid Cp,Be at -160° C and 25” C and the spectrum of liquid Cp,Be at 65°C were recorded. They are shown in Fig. 2. The band positions are listed in Table 1. We have considered the four models presented in Fig. 1 in analysing the obtained spectra. The simplest spectrum is expected for the structure of the highest symmetry, A. The spectrum obtained is much more complicated than the spectra of compounds of D 5d symmetry, e.g. ferrocene 1141 or magnesocene [ 151. Thus we conclude that the molecule contains two nonequivalent Cp rings. The obtained spectrum contains all the bands corresponding to the normal modes of one Cp ring of local CSUsymmetry r-bonded to Be atom, see Table 1. Their positions and intensities are very similar to those observed in the spectra of monocyclopentadienylberyllium derivatives which are believed to have a CpBe moiety of CSv symmetry as CpBeCl [16,17,18], CpBeBr, CpBeCH,, CpBeCSH [ 181. The character of the second Cp ring is not so clear. For the model B we expect two sets of bands corresponding to two different Cp rings, both of local CSUsymmetry, as assumed in the analyses of the IR spectrum of Cp,Be [ll, 121. However, in our opinion the existence of a polarized band at 3000 cm-’ in the C-H stretchingregion and polarized bands in the region 13~80-1420 cm-‘, which are unexpected for a Cp ligand of CSU symmetry, excludes this model.
295
3000
Fig. 2. Raman
25’00
spectra
2000 of
1500
1000
500
0
Cp2Be.
The model with an electrostatically bonded second ring of local Dsh symmetry is rejected for similar reasons. In the case of the o-bonded ring, model C, a more complicated spectrum is expected [ 10,19,20]. The most characteristic bands should appear in the C-H stretching region 2800-3000 cm-’ (unexpected for any other models), and in the C=C stretching region, 1500-1600 cm-‘. We have not observed-such bands in our spectra and the model C has accordingly been rejected. Thus we are left with model D, with pentahapto/poZyhapto bonded Cp rings, which we assume to be most consistent with our data. The only published vibrational spectra of a polyhap to-bonded Cp ring of which we are aware are those presented by Stadelhofer et al. [ 211 for CpAlEt:! and CpGaEt,. These are similar to those obtained for Cp,Be, especially in the C-H stretching region. Our conclusions can be summarized as follows: (a) there are two different Cp rings in the molecule of solid and liquid Cp,Be, (b) one of the rings possesses C,, local symmetry, (c) the second ring deviates from the local CsUsymmetry and is probably polyhapto-bonded to the metal, and (d) the proposed bond system of Cp,Be is consistent with X-ray data obtained by Wronget al. at room tempertiture [ 6]_ We have not observed any significant differences between the spectrum of Cp,Be in the solid at -160 and 25°C and that of the liquid at 65°C. Hence in the case of the crystals at low temperature we are inclined to
296 TABLE1.RAMANSPECTRALFREQUENCIES(cm-l)OFCp~Be -Liquid Solid Solid 65OC -160°C 25OC
--__.-___ POkUiZation ratio
Proposedassignments for CPM
Approximate description
Symmetry
mode 87s 124s 136s 150s 165m 187s 320s 447"s 461s 600s 630 VW 736 vvw 762~ 780 w
313.5 430s(br)
313s 447m
0.23 0.76
U(lO)
El
~(CP-Be)= v(Cp-Be)
598m 627 w(sh)
598m 628vw
0.37 0.60
~(16)
E2
YKP-Be)" X(CCC)
75ovw
745 ww
0.80
795w
795w
0.27
V(3)
AI
P(CW
841~
839
w
0.50
D(15)
E?_
P(CH)
870~
873 m
0.75
WI
El
P(CH)
0.75 0.50
V(l4)
E2
A
U(8)
El
P(CW
789 xv 838 w(sh) 846~ 869 rv(sh) 883~ 905 ww 934"W 996w 103O"W 1058m 1075m 1089m 1118~s 1129 vs 1227 ww 1256~~ 1267 vvw 1325 VW 1356m 1390s 1420~ 144ovvw
3013s 3069m 3076m 3087s 3094s 3100rn(sh) 3110m 3116m(sh) 3122s
1
--_-__
925 vvv., 994 VW 1030vw(sh) 1055w(sh)
0.30
1083m
1082m
0.68
U(l3)
E2
P(CW
1118 s(sh) 1127 YS 1225~~1~
1116 s(sh) 1130 vs
0.15 0.12
V(2)
Al
ccc>
1249 VW
1246 VW
0.28
2v(16)
-41
1360m 1388m 1420~ 1440 vw(sh) 1506 vxw 2783 \~w 2869ww 3020~
1362m 1395m 1420111
0.75 0.28 0.19
U(l2)
E2
3002 w(br)
0.25
3070m
3074m(sh)
0.60
3102s
3100s
0.25
3123s
.---. _
0.10
3128s
..---__
0_15
_.--_~.._.
___~._.
V(CC)
.
V(7) El C5Hg imp.
V(CC)
~(6).Wl)
El.Ez
NCH)
U(l)
A1
v(CH)
.___
___
"We believe.thosetwo bandsbelongto sYmmetricstretching u(h5-Cp-Be) and v(h2/h3-Cp-Be). failedio assign them specifically. We found [183the bands corresponding to v(h5-Cp-Be)in of CpBeXasfollows: CpBeC1313 cm-' .CpBeBr246cm-'.C~BeCH3396cm-~.CpBeC~CH
but ,,,e the spectra 351cm-1.
29-i
accept the pentahapto/polyhapto bond description proposed by Drew and Haaland, rather than the (~---7cbond description proposed by Wong et al. [5]. Experimental Cp,Be was obtained by the method described by Fischer et al. [Z] and purified by vacuum transfer. The melting point of Cp,Be was 59-6O”C, its purity was also checked by mass spectrometry. The Raman spectra were recorded on a Coderg Spectrophotometer, with the 4880 A line of an Ar ion laser as exciting radiation_ Acknowledgement The authors thank Dr. A. Okniiiski and Prof. S. Pasynkiewicz for valuable discussion and the Institute of Low Temperature and Structural Investigation PAN Wroclaw for financial support of this work. References 1 T.J. Kealy and P.L. Pauson. Nature. 168 (1951) 1039. 2 E.O. Fischer and H.P. Hoffmann. Chem. Ber.. 92 (1959) 482. 3 -4. Almenningen. 0. Bastiansen and A. Haaland. J. Chem. Phps.. 40 (1964) 4 A. Haaland. Acta Chem. Stand., 22 (1968) 3030. 5 6 7 8 9
C-HC-H. D.A. H.P. H.P.
3434.
Wang, T.Y. Lee, K-J. Chao and S. Lee, Acta Cryst. B. 28 (1972) 1662. Wang. T-T. Lee. T-J. Lee, T-W. Chang and C.S. Liu. Inorg. Nucl. Chem. Drew and A. Haaland. .4cta Crsst. B, 28 (1972) 3671. Fritz, Chem. Ber.. 92 (1959) 780. Fritz and R. Schneider. Chem. Ber., 93 (1960) 1171.
Lett.,
9 (1973)
667.
10 11 12 13 14
H.P. Fritz. Advan. Organometal. Chem.. 1 (1964) 262. H.P. Fritz and D. Sellmann, J. Organometal. Chem.. 5 (1966) 501. G.B. hlcvickerand G.L. Morgan, Spectrochim. Acta A. 26 (1970) 23. V.T. Aleksanvan and B.V. Lokshin, J. Organomctal. Chem., 131 (19i7) D. Hartley and X1.1. Ware, .I. Chem. Sot. A. (1969) 138.
15
V.T. Aleksanvan. I.A. Garbwova, V.V. Gavrilenko and L.I. Zncharkin. J. Orgnnomctal. Chem.. (197i) 139. D.A. Coe. J.W. Nibler. T-Y. Cook, D. Drew and G.L. Morgan, J. Chem. Phvs., 63 (19i5) 4842. K.B. Starowieyski and J. Lusztyk, J. Organometal. Chem.. 133 (1977) 281. J. Lusztpk. Ph.D. Thesis. Warsaw Technical University. 1978. P-C. Angus and S.R. Stobart, J. Chem. Sot. Dalton Trans., (19i3) 237-Z. E. Maslor-sky. .Jr_ and K. Nakamoto. Inorg. Chem.. 8 (1969) 1108. .J. Stadelhofer. .J. Weidlein. P. Fischer and A. Hanlanci, .J. Organometal. Chem., 116 (1976) 55.
16 17 18 19 20 11
113. 129
of Organometallic
Journal
@ Ekevier
Sequoia
S-A.,
Chemistry, Lausanne
-
170 (1979) 293-297 Printed in The Netherlands
THE RAMAN SPECTRUM AND STRUCTURE DICYCLOPENTADIENYLBERYLLIUM
J. LUSZTYK
Institute 00-662
and K.B.
STAROWIEYSKI
of Organic Chemistry Warsaw (Poland)
(Received
November
OF
15th,
and Technology,
Warsaw
Technical
University
(Politechnika),
1978)
summary The Raman spectra of solid and liquid dicyclopentadienylberyllium are described_ They suggest the presence of two n-bonded rings in the molecule, one pentahapto and one polyhapto coordinated to the Be atom.
Introduction In 1951 the first metallocene, CpzFe, was discovered [l], and shown to have the sandwich structure of Dsd symmetry (Fig. 1A). Dicyclopentadienylberyllium, Cp,Be, the smallest metallocene, was first synthesized by Fischer and Hoffmann in 1959 [2]. In 1964 Haaland et al. [3,4] concluded from electron diffraction studies that in the gas phase Cp,Be had a structure of Csv symmetry (Fig. lB), highly unusual for metallocenes. On the other hand, Wong et al. [ 5, 61 showed by X-ray diffraction studies that in the solid state Cp,Be had a slip sandwich structure_ Wong et al. [ 53 proposed a G-K bond system in the molecule at low temperatures (Fig. 1C). However, Drew and Haaland [ 71 interpreted Wong’s results differently, suggesting the presence of two x-bonded rings in the molecule, one pentahapto and one polyhapto coordinated to the Be atom (Fig. 1D). The structure of Cp,Be was also investigated by means of vibrational spectroscopy. Fritz et al. [S,9,10] concluded from the IR spectrum of Cp,Be in the solid state that the molecule contained two equivalent centrally o-bonded rings. After the structure of Csu symmetry has been proposed for the gas phase [ 3,4], Fritz and Sellmann [ 111 and McVicker and Morgan [12] suggested the same structure for the solid state and solutions on the basis of new IR data. McVicker and Morgan found the model (CpBe)’ Cp- to be the most consistent with their IR data for solid Cp,Be. The IR spectra mentioned above are not consistent_ Furthermore, no papers
%,
C 5v
Fig_ 1. Possible structures of Cp2Be.
concerned with the vibrational spectra of Cp,Be have appeared since the slip sandwich structure was established by X-ray diffraction [ 5,6]. Thus, there is a disagreement between X-ray and IR results. Moreover, the Raman spectrum of this compound has not been reported_ We thus decided to undertake measurements of Raman spectra of Cp,Re in order to remove the inconsistency between vibrational and X-ray data and provide a new approach to the question of the bonding in this compound_ Results and discussion The Raman spectra of solid Cp,Be at -160° C and 25” C and the spectrum of liquid Cp,Be at 65°C were recorded. They are shown in Fig. 2. The band positions are listed in Table 1. We have considered the four models presented in Fig. 1 in analysing the obtained spectra. The simplest spectrum is expected for the structure of the highest symmetry, A. The spectrum obtained is much more complicated than the spectra of compounds of D 5d symmetry, e.g. ferrocene 1141 or magnesocene [ 151. Thus we conclude that the molecule contains two nonequivalent Cp rings. The obtained spectrum contains all the bands corresponding to the normal modes of one Cp ring of local CSUsymmetry r-bonded to Be atom, see Table 1. Their positions and intensities are very similar to those observed in the spectra of monocyclopentadienylberyllium derivatives which are believed to have a CpBe moiety of CSv symmetry as CpBeCl [16,17,18], CpBeBr, CpBeCH,, CpBeCSH [ 181. The character of the second Cp ring is not so clear. For the model B we expect two sets of bands corresponding to two different Cp rings, both of local CSUsymmetry, as assumed in the analyses of the IR spectrum of Cp,Be [ll, 121. However, in our opinion the existence of a polarized band at 3000 cm-’ in the C-H stretchingregion and polarized bands in the region 13~80-1420 cm-‘, which are unexpected for a Cp ligand of CSU symmetry, excludes this model.
295
3000
Fig. 2. Raman
25’00
spectra
2000 of
1500
1000
500
0
Cp2Be.
The model with an electrostatically bonded second ring of local Dsh symmetry is rejected for similar reasons. In the case of the o-bonded ring, model C, a more complicated spectrum is expected [ 10,19,20]. The most characteristic bands should appear in the C-H stretching region 2800-3000 cm-’ (unexpected for any other models), and in the C=C stretching region, 1500-1600 cm-‘. We have not observed-such bands in our spectra and the model C has accordingly been rejected. Thus we are left with model D, with pentahapto/poZyhapto bonded Cp rings, which we assume to be most consistent with our data. The only published vibrational spectra of a polyhap to-bonded Cp ring of which we are aware are those presented by Stadelhofer et al. [ 211 for CpAlEt:! and CpGaEt,. These are similar to those obtained for Cp,Be, especially in the C-H stretching region. Our conclusions can be summarized as follows: (a) there are two different Cp rings in the molecule of solid and liquid Cp,Be, (b) one of the rings possesses C,, local symmetry, (c) the second ring deviates from the local CsUsymmetry and is probably polyhapto-bonded to the metal, and (d) the proposed bond system of Cp,Be is consistent with X-ray data obtained by Wronget al. at room tempertiture [ 6]_ We have not observed any significant differences between the spectrum of Cp,Be in the solid at -160 and 25°C and that of the liquid at 65°C. Hence in the case of the crystals at low temperature we are inclined to
296 TABLE1.RAMANSPECTRALFREQUENCIES(cm-l)OFCp~Be -Liquid Solid Solid 65OC -160°C 25OC
--__.-___ POkUiZation ratio
Proposedassignments for CPM
Approximate description
Symmetry
mode 87s 124s 136s 150s 165m 187s 320s 447"s 461s 600s 630 VW 736 vvw 762~ 780 w
313.5 430s(br)
313s 447m
0.23 0.76
U(lO)
El
~(CP-Be)= v(Cp-Be)
598m 627 w(sh)
598m 628vw
0.37 0.60
~(16)
E2
YKP-Be)" X(CCC)
75ovw
745 ww
0.80
795w
795w
0.27
V(3)
AI
P(CW
841~
839
w
0.50
D(15)
E?_
P(CH)
870~
873 m
0.75
WI
El
P(CH)
0.75 0.50
V(l4)
E2
A
U(8)
El
P(CW
789 xv 838 w(sh) 846~ 869 rv(sh) 883~ 905 ww 934"W 996w 103O"W 1058m 1075m 1089m 1118~s 1129 vs 1227 ww 1256~~ 1267 vvw 1325 VW 1356m 1390s 1420~ 144ovvw
3013s 3069m 3076m 3087s 3094s 3100rn(sh) 3110m 3116m(sh) 3122s
1
--_-__
925 vvv., 994 VW 1030vw(sh) 1055w(sh)
0.30
1083m
1082m
0.68
U(l3)
E2
P(CW
1118 s(sh) 1127 YS 1225~~1~
1116 s(sh) 1130 vs
0.15 0.12
V(2)
Al
ccc>
1249 VW
1246 VW
0.28
2v(16)
-41
1360m 1388m 1420~ 1440 vw(sh) 1506 vxw 2783 \~w 2869ww 3020~
1362m 1395m 1420111
0.75 0.28 0.19
U(l2)
E2
3002 w(br)
0.25
3070m
3074m(sh)
0.60
3102s
3100s
0.25
3123s
.---. _
0.10
3128s
..---__
0_15
_.--_~.._.
___~._.
V(CC)
.
V(7) El C5Hg imp.
V(CC)
~(6).Wl)
El.Ez
NCH)
U(l)
A1
v(CH)
.___
___
"We believe.thosetwo bandsbelongto sYmmetricstretching u(h5-Cp-Be) and v(h2/h3-Cp-Be). failedio assign them specifically. We found [183the bands corresponding to v(h5-Cp-Be)in of CpBeXasfollows: CpBeC1313 cm-' .CpBeBr246cm-'.C~BeCH3396cm-~.CpBeC~CH
but ,,,e the spectra 351cm-1.
29-i
accept the pentahapto/polyhapto bond description proposed by Drew and Haaland, rather than the (~---7cbond description proposed by Wong et al. [5]. Experimental Cp,Be was obtained by the method described by Fischer et al. [Z] and purified by vacuum transfer. The melting point of Cp,Be was 59-6O”C, its purity was also checked by mass spectrometry. The Raman spectra were recorded on a Coderg Spectrophotometer, with the 4880 A line of an Ar ion laser as exciting radiation_ Acknowledgement The authors thank Dr. A. Okniiiski and Prof. S. Pasynkiewicz for valuable discussion and the Institute of Low Temperature and Structural Investigation PAN Wroclaw for financial support of this work. References 1 T.J. Kealy and P.L. Pauson. Nature. 168 (1951) 1039. 2 E.O. Fischer and H.P. Hoffmann. Chem. Ber.. 92 (1959) 482. 3 -4. Almenningen. 0. Bastiansen and A. Haaland. J. Chem. Phps.. 40 (1964) 4 A. Haaland. Acta Chem. Stand., 22 (1968) 3030. 5 6 7 8 9
C-HC-H. D.A. H.P. H.P.
3434.
Wang, T.Y. Lee, K-J. Chao and S. Lee, Acta Cryst. B. 28 (1972) 1662. Wang. T-T. Lee. T-J. Lee, T-W. Chang and C.S. Liu. Inorg. Nucl. Chem. Drew and A. Haaland. .4cta Crsst. B, 28 (1972) 3671. Fritz, Chem. Ber.. 92 (1959) 780. Fritz and R. Schneider. Chem. Ber., 93 (1960) 1171.
Lett.,
9 (1973)
667.
10 11 12 13 14
H.P. Fritz. Advan. Organometal. Chem.. 1 (1964) 262. H.P. Fritz and D. Sellmann, J. Organometal. Chem.. 5 (1966) 501. G.B. hlcvickerand G.L. Morgan, Spectrochim. Acta A. 26 (1970) 23. V.T. Aleksanvan and B.V. Lokshin, J. Organomctal. Chem., 131 (19i7) D. Hartley and X1.1. Ware, .I. Chem. Sot. A. (1969) 138.
15
V.T. Aleksanvan. I.A. Garbwova, V.V. Gavrilenko and L.I. Zncharkin. J. Orgnnomctal. Chem.. (197i) 139. D.A. Coe. J.W. Nibler. T-Y. Cook, D. Drew and G.L. Morgan, J. Chem. Phvs., 63 (19i5) 4842. K.B. Starowieyski and J. Lusztyk, J. Organometal. Chem.. 133 (1977) 281. J. Lusztpk. Ph.D. Thesis. Warsaw Technical University. 1978. P-C. Angus and S.R. Stobart, J. Chem. Sot. Dalton Trans., (19i3) 237-Z. E. Maslor-sky. .Jr_ and K. Nakamoto. Inorg. Chem.. 8 (1969) 1108. .J. Stadelhofer. .J. Weidlein. P. Fischer and A. Hanlanci, .J. Organometal. Chem., 116 (1976) 55.
16 17 18 19 20 11
113. 129