COMMUNICATIONS TO THE EDITOR
With the further extension of the molecule, this model leads to the one proposed by us, namely, where the molecule was fully extended. It is impossible to show that such complete extension as used in our calculation would take place in any finite time interval; however, we can see no reason to expect breakage in the center of the molecule without some orientation of the structure along the flow lines, and for such a situation the dimensions of the molecule must be greater than for a random coil. This statement is our major rebuttal to the criticisms of Harrington and Zimm. In spite of the fact that it is not possible to carry out the kinetic calculations in detail, it has been shown with model systems that flexible fibers in laminar flow will not maintain a random c~nfiguration,~ and in fact the shear dependence of the viscosity of many polymeric materials shows that they do not. Recent experimental evidence shows that the preferential rupture of DNA molecules in the middle of the polymer under hydrodynamic shear holds even in the case of denatured molecules where the rigidity of the chains is much lower than in the case of native DNA.6 Thus, we would conclude that if a t the moment of shear even flexible molecules are to some degree extended, then the stiff native D S A molecules may be close to full extension. Our model would predict that the rupturing shear stress is independent of solvent but not of solvation, and we cannot deduce how a change of solvent would affect the rate at which the molecule can become extended in the brief periods it comes under tension as it is tumbling. I n fact, Harrington and Zimm found that the molecules were more readily broken in a poor solvent than in a good solvent. Their observation is directly opposed to the behavior predicted by their model which would assume a greater extension and therefore a lower critical shear for molecules in a good solvent. Finally, Harrington and Zimm criticized the use of the Lippencott and Schroeder potential function for calculating the breaking strength of the covalent bonds in the DNA backbone. It is true that the equation is properly applicable to the gas phase and its relevance to normal solvated reaction mechanisms can be questioned. Howevw, there is no evidence that shearing proceeds by a hydrolytic process. For polymers in organic solvents, it is known that degradation induced by sonic irradiation and shearing with high speed stirrers proceeds extensively if not entirely with the liberation of the radicals at the points where the polymers break.’ There is also evidence for free radical phenomena in sonic degradation of polymers in aqueous solutions, and in view of the striking parallel
465
between sonic degradation and hearing,^^^ it is probable that shear scission in aqueous systems also occurs with the liberation of free radicals, clearly a highly energetic process. A free-radical generating process is, we believe, compatible with tearing the molecule asunder, scarcely with the facilitation of a hydrolytic process. Both the bond st,ress calculated on our model and the bond strength we calculated are maximal, but it appears to us that both values must be in the right order of magnitude. On the other hand, Harrington and Zimm are forced to postulate unknown mechanisms involved in shear degradation since the activation energy which they calculate is provided by the hydrodynamic forces is too weak to provide significant facilitation of normal hydrolysis. The difficulty in interpretation stems, we believe, from an inappropriate calculation. (5) 0. L. Forgacsand S. G . Mason, J . Colloid Sci., 14,473 (1959). (6) P. F. Davison and D. Freifelder, J . Mol. Biol., 16, 490 (1966). (7) W. R. Johnson and C. C. Price, J . Polymer Sci., 45,217 (1960). (8) D. Freifelderand P. F. Davison, B w p h y s . J . , 2,235 (1962). DEPARTMENT OF BIOLOGY CYRUSL E V ~ N T H A L INSTITUTE OF TECHNOLOGY PETER F. DAVISON MASSACHUSETTS CAMBRIDQE, MASSACHUSETTS 02139 RECEIVED OCTOBER 6. 1966
A 2:l Solid-state Complex of HMBaTCNE’ Sir: Although pyrene and pyromellitic dianhydride tend to form a 1 : 1 crystalline molecular complex, Ilmet and Kopp have recently discovered that, using rather extreme conditions, the 2 : 1 crystalline complex may be obtained.2 They proposed that the 2 : l crystal is composed of alternating stacks of the 1: 1 complex and pure pyrene, but this certainly has not been confirmed. In a related study, we have found that well-defined needle-shaped crystals of a 2 : 1 complex of hexamethylbenzene (HMB) and tetracyanoethylene are easily deposited from an ether solution. I n this case, however, the infrared spectra and, in particular, the infrared dichroism exhibited by these crystals show conclusively that the extra HMB molecules are incorporated into the complex stacks so that the simplest and most likely structure can be represented as -D.A.D-
D*A*D-. The 2 : 1 complex crystals were prepared by mixing 2 mmoles of HMB with 1 mmole of TCISE in 50 ml of ether within a standard 100-ml beaker followed by (1) Supported by the National Science Foundation. (2) I. Ilmet and L. Kopp, J . P h y s . Chem., 70, 3371 (1966).
Volume 71,Number B January 1967
COMMUNICATIONS TO THE EDITOR
466
1:l 2:l
mt -or -plane
1:
-1
-
:1
1600
1400
1200
1000
Figure 1. Infrared spectra in the 1100-1600-cm-1 range for the 1:1 and 2 : 1 crystalline complexes of HMB and TCNE.
evaporation at the rate of 10 ml/hr. During evaporation, parallel needle crystals of the 2: 1 complex were continuously deposited on a polished NaCl window positioned upright in the center of the beaker. Under tt microscrope the resulting pattern appeared to be as well ordered as for the 1:l crystal which had been studied p r e v i ~ u s l y . ~ The infrared spectra in Figure 1 are confirmation of the orderly arrangement of the crystals, as nearly perfect separation of in-plane and out-of-plane infrared activity was invariably achieved by orienting the needle
The Journal of Physical Chemistry
axes alternately perpendicular to and parallel to the electric vector of the polarized infrared radiation. The most interesting spectral feature relates to the 1295-cm-' totally symmetric C-CH, stretching mode of HMB. Ferguson and Chang have shown that this normally inactive mode is strongly activated in the HR/IB.Iz complex in solution4 and have properly attributed this activity to a vibronic interaction unique to such c~mplexes.~The charge oscillation between donor and acceptor, characteristic of this vibronic interaction, is symmetry canceled in the -D .A D .Astacks in the 1 : l crystal of H14B.TCNE so that, as can be seen from Figure 1, no absorption occurs a t 1295 em-'. However, in the proposed structure for the 2 : 1 crystal, the symmetry about the donor molecules is lowered so that symmetry cancellation of the vibronic charge oscillation is no longer possible. One would thus predict and, in fact, observes strong activity a t 1295 em-' in the 2 : 1 complex. The existence of this strong absorption, having the predicted out-of-plane character, plus the obvious stacked character of the crystal is considered strong evidence for the proposed crystal structure. It is also noteworthy t,hat the infrared activity at 1560 cm-l, attributable to the totally symmetric double bond mode of TCT\TE,3 is quite similar in the 2: 1 and 1: 1 crystals. This is consistent with the expectation that the symmetry about the acceptor is comparable in the two crystals. (3) J. Stanley, D. Smith, B. Latimer, and P. Devlin, J . Phys. Chem.. 70, 2011 (1966). (4) E. E. Ferguson and I. Y . Chang, J . C h m . Phgs., 34, 628 (1961). ( 5 ) H. B. Friedrichs and W-.B. Person, ibid., 44, 2161 (1966).
BOBBY HALL DEPARTMENT O F CHEMISTRY J. PAUL DEVLIN OKL.ZHOMA ST.\TEUNIVERSITY STILLWATER, OKLAHOMA RECEIVED NOVEMBER 21, 1966
