Dec. 3, 1960
HYDROGEN ISOTOPE EFFECT ON DEOXYMERCURATIOX
[CONTRIBUTION FROM
THE
6127
SCHOOL OF CHEMISTRY OF THE UNIVERSITY OF MINNESOTA, MINNEAPOLIS 14, MINN.]
Secondary Hydrogen Isotope Effects on Deoxymercurationl BY MAURICEM. KREEVOY AND LEROY T . DITSCH RECEIVED JUNE 3, 1960 The ratio k H / k D for acid cleavage of CH30CH2CH:HgI and its l,l,Z,Z-tetradcuterio analog is 1.06 zt 0.02. This isotope effect, combined with the known vibrational frequencies of the substrates and metal-olefin complexes, preclude the extensive participation of a mercuric-olefin complex in the transition state resonance hybrid. The small isotope effect also tends to preclude extensive carbonium ion character in the transition state. It can be accommodated by a substrate-like model of the transition state in which the methylene-oxygen bond is breaking.
I n previous work2 strong evidence has been presented that the rate-determining step in acidcatalyzed deoxymercuration is the reaction shown in eq. 1. Equation 1says nothing, however, about the degree to which the protonated starting ma-
I -C-CI
HgI
HOCHa e
1 1
I l + -C.y.yC.. ..
e
+ HOCHs
(1)
HgI
lo-' 1. mole-' sec.-l and a probable error of 0.0,3 X 10+ 1. mole-' sec.-'.j The ratio & / K D is, therefore, 1.06, with a probable error of 0.02.
Discussion The reaction coordinate for the reaction shown in eq. 1 has previously been pictured as shown in Fig. 1. The observed isotope effect will be discussed in terms of this model and eq. 2. Equation 2 is
terial, the mercuric iodide-olefin complex and perhaps other structures participate in the structure of the transition state. The position of the transition state along the reaction coordinate is also obtained by rearranging and expanding the equanot known. I n order to throw some light on these tions of Bigeleisen and Mayer6 without introducing 1,1,2,2-tetradeuterio-l-iodomercuri-2-any new assumptions or approximations. questions methoxyethane (11) was prepared, and its elimiI n eq. 2 , K is a transmission coefficient, m+ is nation rate in aqueous perchloric acid was com- a reduced mass for motion along the reaction pared with that of its undeuterated analog ( I ) . coordinate, u is a symmetry number, vi is a fundaThe isotope effect turns out to be small and its mental frequency and f is a vibrational partition magnitude is discussed. f~nction.~ The ratio K H / K D is assumed to be unity and tunneling has been neglected. Both of Results As before, good pseudo first-order kinetics were these seem to be fairly valid approximations for ~ ~ ~symmetry ~ numobserved for both I and 11. Three new determi- secondary isotope e i T e ~ t s .The nations of kl and k:! for I were made in watera a t bers of all the molecules and transition states being 25.0'. They gave an average value of 3.35 X discussed are unity. On the basis of the model can be obtained by aslo-? 1. mole-' sec.-l for k2 with an average devia- an estimate of ( m D * /mH*)'I2 suming that the two hydrogen atoms on the carbon 1. mole-I set.-'. tion from the mean of 0.08 X whose bond to oxygen is breaking move along with This can be compared with a value of 3.31 X 1. mole-' set.-' with a probable error5 of 0.02 1. that carbon and that no atoms other than these and mole-' set.-' obtained p r e v i ~ u s l y . ~The new and the oxygen atom move. Thus mH* is obtained by old measurements combined gave an average value multiplying the mass of the oxygen atom by the mass of the CH2 group and dividing this product 1. mole-' (thirteen measurements) of 3.32 X sec.-l with a probable error5 of 0.02 X lo-* 1. by the sum of these same two masses. In this way thevalue 1.035 is obtained for (mnT/mH*)'/2. It is mole-' sec. -I. Eight separate determinations of kz were made neither a maximum nor a minimum. The maxifor I1 under exactly the same conditions and using mum value, 1.069, is obtained by assuming that the same perchloric acid. These gave an average the mass moving along with the oxygen (methyl 1. mole-' sec.-' for k2 with group and solvent molecules) is infinite by comvalue of 3.14 x a n average deviation from the mean of 0.08 X parison with the mass of the methylene group. The minimum value, 1.000, would be obtained by (1) This work was supported, in part, by the Air Force Office of assuming that the hydrogen atoms do not move Scientific Research through Contract no. AP 49(638:711 and, in part, along with the carbon atom. by the National Science Foundation through grant no. NSF-G5434. Reproduction in whole or in part is permitted for any purpose of the In past work of this sort the vibrational freUnited States Government. quencies have been divided into two classes: the (2) Maurice M . Kreevoy and Frances R. Kowitt, THISJOURNAL, carbon-hydrogen stretching and bending frequen82, 739 (1960); see 0 . W. Berg, W.P . Lay, A. Rodgman and G . F . cies, for which i t has been assumed that V H is Wright, Can. J . Chem., 36, 359 (1958), for a contrary view, however. (3) Maurice M. Kreevoy and Leroy T. Ditsch, J . Ore. Chcni., 26, 1 . 3 5 ~ ~and ; all other frequencies, which have 134 (1960). been assumed to be unchanged by the isotopic
(4) Compound I and I1 were dispensed as solutions in methanol, so the aqueous solutions referred to actually contained 2y0 of methanol. This much methanol has been shown t o have a negligible effect on the rate of such reactions and, in any event, was a constant factor in the present measurements. (5) R. Idvingston, "Physico Chemical Experiments," The Macmillan Co., New York, N. Y.,1957, ChPp. I.
( 6 ) (a) J Bigeleisen and M. G. Mayer, J. Chem. P h y s . , 16, 261 (1947); (h) J. Bigeleisen, ibid., 17, 675 (1949); (c) J . Bigeleisen and M . Wolfsherg, Ado. Chem. Phys., 1, 15 (1958). (7) A. A. Frost and R . G. Pearson, "Kinetics and Mechaoi+ni," John Wiley and Sons, Inc., New York, h-.Y., 1953, p. 8 5 . ( 8 ) Reference 7, p. 94.
6128
IfAURICE
M. KREEVOY AND LEROY T. DITSCH
Vol. 82
Two limiting assumptions can be made for the four carbon-hydrogen stretching frequencies, the eight deformation frequencies of the methylene groups and the carbon-carbon stretching frequency. I t can be assumed that their starting state frequencies are unchanged except that the five frequencies in I1 (bending, wagging and carboncarbon stretching) which lie close to the carbonoxygen stretching frequencies and have the same symmetry are multiplied by a factor of (1,085)-1/6. Or i t can be assumed that the transition state frequencies resemble those of a metal-olefin complex.'* The latter resemble each other12 and also those of ethylenel3 quite closely. I*,The factor (1.085)-'/~0.9838-enters because the Teller-Redlich product d e l 4 requires that the pre-exponential frequency ratio be very close to HaJ unity for each symmetry class. The transition state molecular weights are nearly identical with those of the starting states and the moments of inFig. 1.-A rough pictorial representation of the reaction ertia must be very similar because there is little coordinate for acid deoxymercuration. The arrows are change in geometry. When the methylene deformavectors representing the direction and, crudely, the relative tions and the carbon-carbon stretching frequency are uncoupled from the carbon-oxygen stretching size of the atomic motions. frequencies, the former must change sufficiently so s u b s t i t ~ t i o n . ~Most of the fundamental frequen- that the product rule will still be obeyed. The incies of I and I1 have been identified,l" so that troduction of the factor 0.9838 distributes this these approximations need not be made in the change evenly over those frequencies of I1 in the present case. The 23 assigned vibrational fre- neighborhood of the carbon-oxygen stretching frequencies of I and I1 have been used (including the quencies which have the proper symmetry. It three skeletal deformations which occur in the would make very little difference to the calculated methylene deformation region) and only the ten isotope effect (a factor of about 1.005) if the deunassigned frequencies have been assumed equal crease were concentrated in one or two frequencies in I and 11. The unassigned frequencies are all instead of being evenly distributed. low-lying skeletal vibrations which should not I n the second transition state model the ethylenebe strongly coupled to the methylene deforma- like frequencies were taken from those of Ziese's tions. It was assumed that the 1110 cm.-' fre- salt -K[Pt(C2H4)C13].H20.12 The ratios VH/VD quency in I and the 1050 cm.-l frequency in I1 are were assumed to be the same as in e t h ~ 1 e n e . l ~ the methylene-oxygen stretching frequencies and In Ziese's salt there is a torsional oscillation which that these go to zero in the two transition states." is a molecular rotation in ethylene. The isotopic The carbon-hydrogen vibrations of the methyl ratio in that case was assumed to be 1.408, which group do not undergo significant isotopic shifts in makes the pre-exponential frequency ratio unity. the starting states and i t seems reasonable that With the first set of assumptions the partition this will be true also in the transition states. function ratio is 1.000 and the exponential zeroThe methyl-oxygen frequency and the hydrogen- point energy factor is 1.005, so that the calculated oxygen frequencies in the transition state are isotopic ratio, k H / k D , is 1 . 0 4 0 . With the second coupled to methylene groups only through the set of assumptions the partition function ratio is carbon-oxygen bond which is breaking. They 0.942 and the exponential factor is 0.632, SO that should, therefore, undergo very little change when the calculated isotopic rate ratio is 0.616. The the methylene groups are deuterated and it is as- former is in very reasonable agreement with the sumed that they undergo none. The low-lying observed value (1.06 =k 0.02) while the latter is structural deformation frequencies are assumed very far removed. It can, therefore, be conto be the same in the two transition states. cluded that the transition state probably resembles (9) A. Streitwieser, Jr., R . H. Jagow, R . C. Fahey and S. Suzuki' the protonated substrate, and bears very little THISJOURNAL, 80, 2320 (1968). resemblance to the mercuric-olefin complex. (10) M. M. Kreevoy and L. T . Ditsch, i b i d . , 83, 6124 (1960); The second model fails primarily because of the The vibrational assignments were made for the solids but the spectroscopic results strongly suggest that the bulk of the material retains the substantial increase in the carbon-hydrogen stretchsolid state geometry and vibrational frequencies a t least in solutions in ing frequencies that accompanies the conversion carbon tetrachloride and carbon disulfide. of the substrate to a metal-olefin complex (or to (11) The carbon-carbon stretching frequency, the methylen-oxyethylene, for the matter). The agreement with the gen stretching frequency and the methyl-oxygen stretching frequency were not originally identified10 but it seems reasonable that the carbonpredictions of the first model is not critically de-
ti
\
I
carbon frequency should undergo the largest isotopic shift, the methylene-oxygen frequency the next largest, and the methyl-oxygen frequency the smallest. Accordingly the 1090 cm. -1 frequency in I and the 975 cm. -1 frequency in I1 are considered the carbon-carbon stretching frequencies, 1110 cm.-l and 1050 cm.-l the methyleneoxygen stretching frequencies, and 1135 cm.-l and 1106 c n - 1 the methyl-oxygen stretching frequencies.
(12) D . B. Powell and N. Sheppard, Sficorochim. Acto. 18, 6Q (1958). (13) B . L. Crawford, J. E. Lancaster and R. E. Inskeep, J . Chcm. Phys., 81, 678 (1953). (14) G. Herzberg, "Infrared and Raman Spectra of Polyatomic Molecules," D. Van Nostrand Co., Inc., New York, N . Y., 1945, P. 232
Dec. 5, 1960
MONOHALO~RGANOSILANES WITH MAGNESIUM
pendent e n the assigned frequencies of I and I1 because most of these cancel out. It is interesting to note that most of the calculated isotope effect is temperature independent. For a number of reactions involving carbonium ion-like transition states i t has been shown that k R / k D is substantially greater than unity for both a- and &deuterium s u b s t i t ~ t i o n . ~ J ~ -Values, ’~ ranging from 1.05 to 1.3 per isotopic substitution have been reported. The small isotope effect found in the present case (about 1.015 per substitution) along with the ability of a substrate-like (15) A. Streitwieser, Jr., Chcm. Rcos., 66, 571 (1956). (16) K. B. Wiberg, ibid., 66, 713 (1955). (17) V. J. Shiner and S. Cross, THISJOURNAL, 79, 3599 (1957).
[CONTRIBUTIONFROM
THE
6129
model to reproduce the effect militates strongly against any large degree of carbonium ion character in the transition state. If the small excess of observed effect over that predicted is real, however, i t may be due to a slight degree of carbonium ion character in the transition state. Experimental The preparation of I and I1 have been described Kate constants were measured as previously. before. Acknowledgment.-The authors are pleased to acknowledge the financial support of the National Science Foundation and the Air Force Office of Scientific Research. 3910
CHEMICAL LABORATORY OF IOWA STATE UNIVERSITY,
AMES, IOWA]
Reactions of Monohaloijrganosilanes and Magnesium in Tetrahydrofuran BY WALTERSTEUDEL AND HENRY GILMAN RECEIVED JUNE 8, 1960 Monohaloorganosilanes react with magnesium in tetrahydrofuran to give essentially two classes of compounds. Products of tetrahydrofuran ring cleavage are obtained when no aromatic groups are linked to the silicon atom, whereas aryl-containing halosilanes form the corresponding disilanes. By the latter reaction, a new class of disilanes of the general formula PhRHSiSiHRPh has become readily available. The preparation of 1,1,2,2-tetraphenyldisilane,1,2-dimethyl-1,2-diphenyldisilane and 1,2-diethyl-1,2-diphenyldisilaneis described.
Reactions of halosilanes with magnesium have gnard reagent, just as they appear necessary for the been of special interest because of the possible formation of organosilylalkali compounds.”s In formation of silyl Grignard reagents with silicon this Laboratory, the direct coupling of triphenyldirectly linked to magnesium. No such reagent chlorosilane by magnesium in diethyl ether proved has been isolated ; however, the intermediate for- to be unsuccessful,6 but was eventually achieved mation of such compounds has been proposed by by the use of tetrahydrofuran as the ~ o l v e n t . ~ Anderson and SprungEinvestigated the reaction several workers.’-3 Emileus, Maddock and Reid’ suggested the for- of trimethylchlorosilane with magnesium in tetramation of a n unstable Grignard compound SiH3- hydrofuran in the presence of ethyl iodide or magMgI from the reaction of iodosilane with magnesium nesium iodide and found that 4-trimethylsilylin diisoamyl ether to account for the formation of butoxytrimethylsilane and 3-butenoxytrimethylmonosilane and hydrogen. No disilane could be silane are formed as products of the cleavage of detected in their experiments. An analogous re- tetrahydrofuran. They propose a mechanism action between bromosilane and magnesium is re- which involves the cleavage of the tetrahydrofuran ported by Van Artsdalen and Gravis2 Eaborns ring by magnesium iodide in the presence of magreported possible formation of a transient Grignard nesium. The general usefulness of tetrahydrofuran as a reagent from triethyliodosilane and magnesium solvent for the preparation of organometallic comin ether. Recently, strong evidence for the triphenyl- pounds gave the impetus to this investigation of silyl Grignard reagent has been shown by work the behavior of various monohaloorganosilanes of Selin and Wests4 Triphenylchlorosilane was toward magnesium in this solvent. Under the concoupled by cyclohexylmagnesium bromide and ditions employed there are two courses the reaction other Grignard reagents in tetrahydrofuran to give may follow good yields of hexaphenyldisilane. Trimethyl- 2RaSiX T H F -+RoSi(CH&OSiRa (Hz0) chlorosilane, under the same conditions, gave no R&i( CH&OH d- R3SiSiRt reaction, but l,l,l-trimethy1-2,2,2-triphenyldisi-Mg lane was obtained when a mixture of trimethylOne of these leads to the formation of products chlorosilane and triphenylchlorosilane was treated from the cleavage of tetrahydrofuran to give, on with the Grignard reagent. The authors state hydrolysis, 4-organosilyl-substituted 1-butanols as that “apparently aromatic groups on silicon are the main products. This mode of reaction was necessary for the stabilization of the silyl Gri- always observed when the organohalosilane did not
1-
(1) H. J. EmCleus, A. G . Maddock and C. Reid, J . Chcm. Soc.. 353 (1941). (2) E. R . Van Artsdalen and J. Gavis, THISJOURNAL74, 3195 (1952). (3) C. Eaborn, J . Chcm. SOC.,2755 (1949). (4) T. G. Selin and R . West, Tetrahedron, 5, 97 (1959).
__f
(5) H. Gilman and T. C. Wu, THISJOURNAL, 75, 4031 (1951). (6) H. Gilman and T. C. Wu, J . Org. Chcm., 18, 753 (1953). (7) M. V. George, D. J. Peterson and H. Gilman, THILI JOURNAL, 82, 403 (1960). (8) R. P. Anderson and M. M. Sprung, WADC Technical Report 69-61, 47 (1959).
