4990
ELLISK. LIPPINCOTT .IND KICI~ARU L). NELSON
Fig. 3.-Dctail sketch of the styrene-palladium chloride complex. The large atoms are chlorine, the medium size atoms are palladium, and the smallest are the carbon atoms, C1, CI and CS of styrene. The intersection of the three planes lies a t 0.40 A. from carbon atom CS.
The configuration of the atoms in styrene-palla[CONTRIBUTION FROM THE
VOl. 77
diuin chloride is, therefore, entirely consistent with the theory that the palladium-olefin bond consists of interaction between one of the dsp? hybrid orbitals of palladium and the pi orbital of the double bond. The palladium orbital enters the pi bond off center and forms an angle of 74 degrees with the double bond axis. The structure of ethylene-palladium chloride indicates, however, that the palladium orbital enters the ethylene pi bond a t its center and forms an angle of DO degrees with the double bond axis. The ethylene niolecule is symmetrical about its double bond; whereas, in styrene, where one of the carbons of the double bond is attached to the large phenyl group, the bond from palladium is off center. There are several possible causes for the displacement of the palladium bond toward the end carbon of the styrene molecule. (1) There may Le an asymnietrical distribution of electrons in the pi bond of the ethylcne side-chain, the carbon atom C8 being more negative. (2) Rotation of the ethylene group into a position such that the carbon atoms are equidistant from the palladium atom rcquires, either that the ethylene side chain be twisted far out of the plane of the benzene ring, or that the molecules assume an entirely new packing arrangement in the solid. Either alternative could iiicrcase the crystal energy. IOJVACITY,IOWA
DEPARTMEKT O F CHEMISTRY, K A N S A S
STArE
COLLEGE]
The Thermodynamic Functions of Bis-cyclopentadienyl Iron, Bis-cyclopentadienylnickel and Bis-cyclopentadienylruthenium B Y ELLISR. LIPPINCOTT*AND RICIIAKD D. NELSON RECEIVED APRIL 2'3, 1'333 The thermodynamic quantities Cg, ( H o - E ; ) / T , -(F" - E : ) / T acid S"have Ixctl ca1cul:itccl from 2W.lfj to 1500'K. for bis-cyclopentadienyliron(II), nickel(I1) and ruthenium(I1) from at1 ossignirient of frequencies based or1 it~fraretlatld RanIan spectroscopic studies. The quantities A H : , AF: and AS: have been calculated for Fe(CjHi):! and .Ui(C3Hj)> tlie least stable of thesc three from 298.16 t o 1500°K. Thermodynamically Ru(CJ3j):: is tlie most stable ant1 Ni(CbHbjO conlpoutlds.
Introduction The unusual sandwich structures of a number of the transition metal dicyclopentadienyl compounds have stimulated a wide range of studies concerning their chemical and physical properties. l-g The structures of these compounds have been demonstrated conclusively and the aromatic nature of the cyclopentadienyl rings established.+-'3 However
* Department of Chemistry, University of N a r y l a n d . College P a r k , hIaryland. (1) T. J. Kealy and P. L. Pauson, Nature. 168, 1039 (1961). ( 2 ) G. Wilkinson, M. Rosenblum, h4. C . Whiting and R. B. IVoodward, T H I S J O U R N A L , 74, 2125 (1952). (3) R. B. Woodward, h'l. Rosenblum and hl. C . Whiting. rbid., 7 4 , 3-158 (1952). (4) G. Wilkinson, ibid.,7.t, 6146 (1952). ( 5 ) G. Wilkinson, i b i d . , 74, 6148 (1952). (6) E. 0. Fischer a n d R . Jira, N a l u r f o r s c l i u n g , BB, 217 (1053); BB, 327 (1953). ( 7 ) G. Wilkinson, 1'. L. Pauson, J . hI. Birmingham and F. A. Cotton. T H ~JOIJRKAL, S 7 6 , 1011 (1953). (8) G. Wilkinson, P. L. Pauson, K.A . Cotton, i b i d . . 78, 1970 ( 1 9 6 4 ) . (9) E. 0. Fischer and W. P f a b , 2. N a t i ~ r f u ~ s c h u n 7B, g . 377 (1952). (10) P. F. Eiland, D. Pepinsky. THISJ O U R N A L , 74, 4'371 ( 1 9 5 2 ) . (11) J. D. Dunitz. L. E. O r w l . . V a t u r ~ ,171, 121 (1953). (12) €I. H Jaffe, J . C h r ~ n Pitys . , 21, 156 (1053). (13) E. R . Lippincott, R . D. Nclson, ibid.. 21, 1307 (1953).
the nature of the forces binding the rings to the metal atom is only partially understood and there is no complete agreement as to how best to describc the type of bonding involved. A tabulation of the calculated thermodyrianiic functions for some representative examples of this interesting series of coinpounds is then desirable for a coinparison with their other properties. \\.'e are carrying out a detailed spectroscopic study of a number of inetal cyclopentadicnyl compounds. The results of the vibrational ~ Ni(C6H6)L analysis on Fe(C6&)2, R U ( C ~ & )and are nearly complete and permit a reliable assignment of frequencies to bc made to the fundamental inodes of vibration for these molecules. \Ye wish to present here tabulations of a number of thermodynamics functions a t 100" intervals froiii 300 to 1500°K. for these three molecules based on this assignment. At this time no heat capacity or entropy measuremeiits are available for comparison. \Then such thermal measurements are available a comparison of the experimental and calculated therinodynamic quantities will furnish information conccrning the possible existence of a barrier preventing free internal rotation of the cyclopentadi-
THERAIODYNAMIC FUNCTIONS OF BIS-CYCLOPENTADIENYL-METAL COMPOUNDS 4991
Oct. 5 , 1933
enyl rings and permit an estimate of its magnitude. Thermodynamic Functions.-The heat capacity, enthalpy] free energy and entropy functions were calculated by standard methods using the vapor phase rigid rotator, harmonic oscillator approximation. The structural data for Fe(C5H5)2were taken from the results of Eiland and Pepinsky.lo Their results give a C-C distance of 1.41 A. and FeC distance of 2.0 A. By analogy with benzene the C-H distance was assumed to be 1.08A. Using these data with a molecular symmetry of D5d the principal moments of inertia I A and I g are calculated as 355.6 X and 731.7 X g. cm.21respectively. The structural infomiation for Ru(c~,&)z and Ni(C5H5)2is nearly the same as for Fe(C5H&. -4 symmetry number of 10 was used corresponding to D5d symmetry. The vibrational contribution to the thermodynamic functions of these compounds was made using the three assignments given in Table I. The assignment for Fe(CbH5)z is based on a spectroscopic study of both Fe(ChH5)z and its deuterium derivative Fe(CbD5)z. Infrared spectra in the region from 300 to 4000 cm.-' were obtained on the vapor and solid phases as well as in solution. Raman spectra of Fe(C5H5'2 and Fe(C5D5)2were obtained in solution from 100 to 3500 cm.-'. This assignment was supported by a normal coordinate analysis for a number of the inactive frequencies. The assignment for R u ( C ~ Hwas ~)~ based on a similar spectroscopic study and normal coordinate analysis with the exception that data on Ru(C5D6)2 were not available. The assignment for Ni(C5H5)2 is somewhat more empirical than that for Fe(C5H& and Ru(C5H5)2in that no Iiaman spectrum was obtained with the result that more emphasis was placed on frequency positions and normal coordinate calculations similar to those for Fe(ChH5); and R U ( C ~ H ~ However )~. the errors due to incorrect assignment are believed to be small enough to permit a reasonably reliable calculation of thermodynamic functions. For all three molecules the inactive -41, ug frequency was assumed to correspond to a free internal rotation. The contribution to the entropy for this internal rotation was calculated from the equationI4 = (In T In ( I d 4 ) - 2 In 5 .\:,,.r.
+
+
+ 1) R/2
In ( 8 r 3 k / h 2 )
The calculated thermodynamics functions Ci, (HO - E : ) / T , - ( P o - E:)/Tand So are tabulated in Table 11. A breakdown of the total So and (2: into rotational, translational, vibrational and free internal rotational contributions a t 29S.16"K. is given in Table 111. Thermodynamic Functions of Formation for Fe(C5H&(g) and Ni(c~H&(g).-The calculated values of AH:, AF: and A$ for the formation of Fe(C5H& and Ni(C5H& from their elements in their standard states a t various temperatures are tabulated in Table IV. The heats of formation of Fe(C&Hb)*(s)and Ni(C6H5)2(s)have been given by Wilkinson1*J5as 33.8 and 62.5 kcal. per mole, re(14) G. Herzberg, "Infrared and R a m a n Spectra of Polyatomic Molecules," D. Van Nostrand C o . , iYew York. h ' . Y., 1945, p. 524. (15) F. A. C o t t o n , F . Wilkinson, THISJ O U R N A L , 74, 5764 (19623.
TABLE I ASSIGNMENTOF
FREQUENCIES5 FOR
Fe( CaH&Ni( CsH5)z AND
Ru(CsHs)z, CM. Description C v d
FeNiRu(CsHdz (CsHdz (CaHdz
1 CH stretch 3099 3100 3104 2 C H bend ( I ) 804 770 805 3 sym-ring breathing 1105 1105 1104 4 sym-ring metal stretrh 303 220 330 5 C H bend (11) 1200 1200 1200 6 Internal rotation 7 CH bend (11) 1200 1200 1200 8 CH stretch 3085 3075 3100 9 CH bend ( I ) 811 773 806 10 anti-synz-ring breathing 1108 1110 1103 11 anti-sym-ring metal stretch 478 355 456 12 CH stretch 3085 3085 3089 13 CH bend (11) 1010 1010 996 14 CH bend ( I) 800 770 804 15 sym-CC stretch 1408 1430 1412 16 sym-ring tilt 388 275 402 17 C H stretch 3075 3075 3100 18 CH bend (11) 1002 1000 1002 19 CH bend ( I ) 834 800 800 20 anti-sym-CC stretch 1411 1428 1413 21 anti-sym-ring tilt 492 355 528 22 Ring metal ring bend 150 110 165 23 CH stretch 3085 3085 3089 24 CH bend (il) 1178 1180 1193 25 CH bend ( I ) 1050 1050 1056 26 CC stretch I590 1590 1590 27 Ring distortion (11) 900 900 900 28 Ring distortion ( I ) 500 500 500 29 CH stretch 3100 3100 3100 30 CH bend (11) 1170 1170 1170 31 CH bend ( I ) 1050 1050 1050 32 CC stretch 1550 1550 1550 33 Ring deformation ( 1 1 ) 900 900 900 34 Ring deformation ( I ) 500 500 500 a The basis for this assignment will be given in a forthcoming publication. * Raman active species AI,, El,, Egg; Infrared active species A*,, and El,; inactive species AI,, A,, and Ezu. (11) and (I)indicate displacements which are parallel and perpendicular to the planes of cyclopentadienyl rings, respectively. The frequencies in species El,, El,,, EPe and Ezu are doubly degenerate and must be assigned a weight of t w o in the thermodynamic calculations.
spectively. The heat of sublimation for Fe(C5H5)z has been given by Kaplon, Kester and KatzI6 as 16.81 kcal. per mole. Following Wilkinson we have assumed the heat of sublimation of Ni(C5H5)2 to be the same as that for Fe(CbH5)2 because of the close similarity of their physical properties.8 The thermodynamic functions of iron and nickel a t various temperatures were taken from the data of Kelley.17s18 Because their data is given a t irregular temperature intervals i t was found necessary to replot their data to obtain data for our temperatures. These data along with the heats of formation, free energies of formation and entropies of graphite and hydrogen in their standard states1$were combined (16) L. Kaplan, W. J. Kester and J. J. K a t z , ibid., 7 4 , 6331 (1952). (17) K. K. Kelley, U .S. Bur. Mines Buil., 371 (1934). (18) K . K . Kelley, ibid , 3 8 3 (1935); 476 (1949). ( 1 9 ) "Tables of Selected Values of Chemical Thermilclynamic Properties," N a t l . Bur. of Standards (1947).
TABLE I1 THERMODYNAMIC FUNCTIONS FOR ve( CsHj)l(g), Ku( C&Ij)q-
(g) A N D Ni(C&)dg) Fe(C6HdZ (Ho -(FO -
TABLE IV TIIERMODTNAMIC FUNCTIONS OF FORMATIOS FOR Fe( C5Hs)r(9) AND NI(CjH&(g) cas,:, cal./ ( A H ) : , kcal./ ( A F ) ; , kcal./ T , OK.
mole
T,OK.
G
F!)/T
298.16 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
38.05 38.29 51.67 62.61 71.18 77.96 83.45 88.00 91.84 95.07 97.82 100.20 102.25 104.01
19.82 19.93 26.23 32.45 38.24 43.4t 48.10 52.29 56.08 59.47 62.54 65.35 67.92 70, 26
68.17 68.29 74.89 81.39 87.85 94.15 100.25 106.17 111.89 117.39 122.69 127.80 132.75 187.48
87.99 88.22 101.12 113.84 126.09 137.59 148.35 158.46 167.97 l76.Xfj 185.23 193.15 200.67 207.74
298.16 300 400 500 GOO 700 800 900 1000 1100 1200 1300 1400 1500
298.16 300 400 500 600 700 800 900 1000 I100 1200 1300 1400 1500
37.94 38.18 51.66 62.62 71.19 77.96 83.41 87.98 91.81 95.07 97.80 100.19 102.22 103.98
Ru(CJL)Z 19.50 19.6~i 26.03 32.28 38.08 43.33 47.97 52.19 56.01 59.39 62.47 65.29 67.86 70.20
68.42 68.53 75.08 81.53 88.00 94.19 100.29 106.24 111.98 117.44 122.74 127.79 132.78 137.53
87.98 88.20 101.11 113.81 126.08 137.52 148.26 158.43 167.99 176.83 185.21 193.09 200.65 207.73
298.16 300 400 600 600 700 800 900 1000 1100 1200 1300 1400 1500
298.16 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
39.93 40.17 53.01 63.58 71.89 78.51 83.86 88.32 92.09 95.29 97.99 100.35 102.36 104.11
Ni( CbH6)2 21.80 21.92 28.13 34.18 39.81 44.89 49.43 53.52 57.22 60.FZ 63.52 66.27 68.78 71.05
70.G1 70.74 77.91 84.81 91.60 98.11 104.41 110.46 116.34 121.92 127.31 132.50 137.42 142.30
92.41 92.66 106.04 118.99 131.41 143.01 153.84 163.98 173.56 182.44 190.83 198.77 206.20 213.35
F)/T
AND
SO Cpo
SO
CpO
So
cpo
So
FOR
Fe(CbHh)l, Ru(C&)Z Vibrational
mole
AND
Ni(CSH&
Free Total internal (cal./mole rotational degrees)
Rotational
Translational
24.44 2.98
Fe(CbHd2 41.55 16.62 4.95 29.13
5.38 0.99
87.99 38.05
24.44 2.98
Ru( CjIIj)z 42.21 15.95 4.95 29.02
5.38 0.99
87 98 37.94
24.44 2.98
Ni(CeH& 41.60 20.99 4.95 31.00
5.38
02.41 X.92
0.00
mole
50.61 50.55 48.49 46.91 45.61 44.45 43.53 42.71 41.86 41.18 10.48 40. 33 40.21 40.12
Fe( CjH& 75.97 76.18 85.29 94.93 104.83 114.96 125.24 135.60 146.07 156.63 167.25 177.90 188.53 199.26
-85.07 -85.37 -92.01 -9G.05 -98.71 - 100.74 -102.13 - 103.21 -104.21 - 104.95 - 105.(i3 - 105.82 - 105.95 - 106,09
79.rjl 79 I58 77.68 iti.12 74.91 73.80 Z3 .04 72.44 72,Ui 71.78 71.61 71.50 71.42 71.33
hri(CjH5)2 103.80 104.02 112.72 121.92 130.76 141,02 150.84 100.73 170.63 180.04 190.64 200.64 210.77 220. 72
-81. 10 -81.45 -87.59 -91.61 - 93 . IO -96.03 -97.25 - 98,lO -98.57 - 98,9(i -99.19 -99.34 -9 9 . 5 4 -99.59
S'
TABLE I11 CpO
Vol. 77
ELLISli. LIPPINCUTT A N D RICIIARD D. NELSUN
409 2
with the data in Table I1 to obtain the calculated values of AH:, AF: and A$ for the reactions of formation IOC(s) 10C(s)
+ 5Hl(g) + Fc(s) +I'e(C&h(g) + 5Hl(g) 4-X ( s ) --+ Ni(CbH&(g)
These functions are reported a t various temperatures in Table IV. The values given for the standard heats of formation a t O'K., A H $ , for Fe(CbH5)~(g)and N i ( c ~ H ~ ) ~arc ( g )57.5 and 79.6 kcal./rnole, respectively.
Discussion The thermodynamic functions C,O and So calculated for these compounds are all considerably lower than most metalorganic compounds of comparable molecular weight, the effect being particularly evident with Fe(C5H5)zand Ru(c6H6)~.This result is expected in view of the stability and high symmetry of these compounds. The calculated functions CpO, (Ho - E : ) / T , - (FO - E t ) / T and Sofor Fe(C5H6)2and R U ( C ~ H S ) ~ are nearly the same, This accidental result arises because of two opposing effects. If the cyelopentadienyl rings of R U ( C ~ & were ) ~ bound to the metal atom with the same energy as in Fe(C&h)z, the calculated thermodynamic functions for Ru(C&,)z, would be larger than those for Fe(C6H6)2because of the higher molecular weight of Ru(CgH6)). That the
Oct. 5, 1855
THERMODYNAMIC PROPERTIES OF
rings are bound with greater energy in Ru(C&,)z is seen by comparing the ring-metal-ring symmetric stretching frequencies of Fe(C6Hs)z and Ru(C6&)2 which appear a t 303 and 330 cm.-', respectively. Corresponding increases in several other frequencies occur with the result that the increase in molecular weight of RU(C&)2 is compensated completely by the higher set of frequencies resulting from increased stability. Actually the calculated So for Ru(C&)z is slightly less than that for Fe(C5H6)'L.
(CONTRIBUTION NO. 49
FROM THE
4993
2,3-DIMETllYL-~-BUTEN&
A comparison of the thermodynamic functions of Ni(CsH6)z with.Fe(C~,Hs)~ are significantly different. Since these compounds have nearly the same molecular weight, this difference is due to the much greater stability of Fe(C&,)*. Acknowledgment.-The authors wish to thank the Research Corporation for a grant-in-aid supporting this work; and Dr. J. C. Brantley of Linde Air Products Company for supplying us with a sample of bis-cyclopentadienylnickel(11). MANHATTAN, KANSAS
THERMODYNAMICS LABORATORY, PETROLEUM EXPERIMENT STATION, BUREAUO F MINES]
2,3-Dimethyl-Z-butene : Thermodynamic Properties in the Solid, Liquid and Vapor States BYD. W. SCOTT, H. L. FINKE, J. P. MCCULLOUGH, M. E. GROSS,J. F. MESSERLY, R. E. PENNINGTON AND GUYWADDINGTON RECEIVED MAY19, 1955 Various thermodynamic properties of 2,3-diniethyl-Z-butene were studied experimentally. The entropy of the liquid a 298.16"K. (&std = 64.58 cal. deg.-' mole-') was computed from calorimetric values of the heat capacity in the solid and liqilid states (11 t o 818'K.) and of the heats of transition and fusion (844 cal. mole-' at 196.82"K. and 1542 cal. mole-' at the triple point, 198.92'K., respectively). Results obtained for the heat capacity of the liquid ( Cs3td), vapor pressure ( p ) , heat of vaporization (AH,), heat capacity in the ideal gas state (C,"),and the second virial coefficient [ B = (PTi - R T ) / P ] are represented as a function of temperature by the following empirical equations: (1) Csntd = 48.178 - 0.16068T 6.2188 X 10-4T2 - 5.2083 X lo-' T3, cal. deg.-' mole-') (200-320'K.); (2) loglop (mm.) ="6.93324 - 1206.037/(t 224.400), (29-73'); (3) AH, = 10,674 - 5.713T - 0.01344T2, cal. mole-1(292-3460K.); ( 4 ) C, = 4.454 f 0.08736lT - 1.1983 X 10-6T*, cal. deg.-' mole-' (334-473'K.); and B = -2764 - 4.93 exp. (110O/T), cc. mole-1(292-4730K.). The entropy iii the ideal gas state at 298.16'K. (So= 87.16 cal. deg.-' mole-') was computed from these data. A vibrational assignment was made for 2,3-dimethyl-2-butene, and the average height of the potential barriers hindering internal rotation of the methyl groups, 680 cal. mole-', was evaluated from the calorimetric data. Values of the functions ( F " - H ; ) / T , ( H " - H ; ) / T , H" - H;, Soand C; were calculated at selected temperatures t o 1500'K.
++
As part of the continuing program of this Laboratory to measure thermodynamic properties of hydrocarbons important in petroleum technology, calorimetric studies were made on 2,3-dimethyl-2butene. These included determination of the entropy of the vapor from Third Law studies and measurement of the heat capacity of the vapor. The methods of statistical mechanics were used to compute thermodynamic functions for the ideal gas state. Experimental T h e 1951 International Atomic Weights1 and the 1951 values of the fundamental physical constants2 were used for all computations described in this paper. Measurement of temperature above 90°K. was made in terms of the International Temperature Scale of 1948,3 and International Celsius temperatures were converted to Kelvin temperatures by adding 273.16". Below 9O"K., the temperature scale was defined by a platinum resistance thermometer calibrated at the National Bureau of Standards in terms of the provisional scale established by Hoge and Brickwedde.4 All electrical and mass measurements were referred t o standards calibrated at the National Bureau of Standards. Energy measured in joules was converted t o calories by use of the definitions 1 cal. = 4.1840 abs. j . = 4.1833 int. j. Material.-The 2,3-dimethyl-Z-butene used for low temperature calorimetry and for measurement of vapor pressure (1) E. Wichers, THISJOURNAL,74, 2447 (1952). (2) F. D . Rossini, F. T. Gucker, Jr., H. L. Johnston, L. Pauling a n d G . W. Vinal, ibid., 74, 2699 (1952). (3) H. F. Stimson, J. Research Natl. BIU. S l o n d n r d s . 42, 209 (1949). (4) H. J. Hoge and F . G. Brickwedde, ibid., 22, 351 (1939).
was a n A.P.I. Research sample.6 In calorimetric melting point studies, this sample was found t o contain 0.05 mole yo liquid-soluble, solid-insoluble impurity. A second sample, of somewhat lower purity (about 99.5 mole %), supplied by A . P . I . Research Project 6, was used for the measurements of heat of vaporization and vapor heat capacity, which required a larger sample than the other studies. The samples of 2,3-dirnethyl-2-butene were received in sealed ampoules with internal break-off tips and were stored in the dark at 5". Transfers to appropriate receivers were made by vacuum distillation. At no time in the handling of the material or in the experiments were the samples in contact with gases other than helium. Such precautions were necessary to avoid oxidation of 2,3-dimethyl-2-butene by atmospheric oxygen. I n the studies of heat of vaporization and vapor heat capacity, 0.02 mole % hydroquinone was added t o the sample as inhibitor. The Heat Capacity in the Solid and Liquid States.-Tlic low temperature thermal properties of 2,3-dimethyl-Z-butene were measured in a n adiabatic cryostat similar t o t h a t described by Ruehrwein and Huffman.e The sample (about 0.44 mole) was sealed in a copper calorimeter t h a t contained horizontal perforated disks to facilitate attainment of thermal equilibrium and prevent settling of the solid phase during fusion experiments. The observed values of heat capacity a t saturation pressure ( Csatd)of the solid and liquid are presented in Table I. The temperature increments used in the experiments were small enough t h a t corrections for non-linear variation of Caatd with 2" were unnecessary (the (5) This sample, made available through t h e American Petroleum Institute Research Project 44 on t h e "Collection, Analysis, and Calculation of D a t a on Properties of Hydrocarbons," was purified by t h e American Petroleum Institute Research Project 6 on t h e "Analysis, Purification, and Properties of Hydrocarbons," from material supplied by t h e General Motors Corporation, Detroit, Michigan. (6) R. A. Ruehrwein and H. RZ. Huflman, Tms J O U R N A L . 66, 1620 (1943).
