NOTES
3738
*
compounds: S O 2 9 8 RhCls(s) = 27.1 3.0 eu, obtained RuC13(s) = 30.5 by Bell, Tagami, and nIerten;g Sozgs k 2.,i eu, obtained by Bell, Garrison, and Merten;6 and S O 2 9 8 CrC13(s) = 29.4 f 0.2 eu, given by Kelley and King.8 The entropies of RhClo(s) and RuCla(s) were obtained by essentially the same technique as was the present value for IrC13. The entropy for CrCls(s) was based on heat capacity data. Acknowledginents. The authors wish to thank R. E. Inyard for assistance in the experimental work and P. T i . Gantzel and J. 11. Dixon for assistance in X-ray analyses. ~
~~
(9) W. E. Bell, 51. Tagami, and U. Merten, J . P h y s . C h m . , 66, 490 (1962).
Rotational Transitions in Neopentyl Alcohol and Neopeiityl Glycol
by J. A. Faucher, J. D. Graham, J. V. Koleske, E. R. Santei:, Jr., and E. R. Walter Research and Development Department, Cnwn Carbide Corporation, Chemicals Diviaion, South Charleston, W e s t V i r g i n i a (Received Febrztnry 21, 1966)
In recent years, a rapidly growing amount of research has been conducted on rotational transitions in solids.' However, most workers in this area have utilized only a single physical method, such as nuclear magnetic resonance (nmr) or heat capacity. Having discovered the occurrence of rotational transitions in neopentyl alcohol, CHsC(CH3)2CH20H,and neopentyl glycol, HOCHZC(CHs)2CH~OH, me felt that it would be of some interest to compare the results of several methods, as applied to these compounds. Because of their roughly spherical shape, neoprntyl derivatives are likely materials to have rotational transitions, but relatively little work has been done with such compounds. In particular, the alcohol and glycol have not been studied in this connection, although a recent paper of Dannhauser and co-workers2 suggests that the alrohol is a plastic crystal. Experimental Section and Discussion The samples used were commercial materials from llatheson Coleman and Bell (neopentyl glycol mp 127') and Aldrich Chemical Co. (neopentyl alcohol nip 53') ; they were used without further purification. The initial investigation was made with a recording torsion pendulum3 which is more commonly employed for polymeric samples. Materials which are not T h e Journal o j Physical Chemistry
self-supporting, however, can be easily used with the aid of a cellulose blotter strip.4 The blotter is saturated with the liquid sample, cooled in liquid nitrogen, and then mounted in the pendulum. Damping measurements are taken at regular intervals as the temperature is slowly raised. With this technique, transitions of almost any kind in the sample will produce peaks in the mechanical loss. The method is particularly useful for a quick survey of transition behavior. The curves obtained in this manner for the two compounds are plotted in Figure 1. Instead of the loss, Q-I, a related quantity, the loss modulus, G", has been used since it shows the transition behavior more clearly (in part due to the fact that a logarithmic scale is employed). For both cases the rotational transitions are clearly indicated by sharp peaks in the loss modulus, at -45' for neopentyl alcohol and at 40' for neopentyl glycol. I n addition, there are clear indications of "secondary" transitions at about - 120 and -40°, respectively. Such secondary transitions have been observed before by this method, for example, in the normal aliphatic alcohols by Illem6 They have been interpreted as indications of small segment mobility in polymers or the onset of limited molecular motion in monomeric materials such as the present case. Additional information mas obtained by wide-line nuclear magnetic resonance spectroscopy. A Varian DP-40 spectrometer equipped for variable temperature measurements was employed. Figure 2 shows the line-width curves for both materials. In neopentyl alcohol, the observed temperature dependence of line width from -170 to -45' may be attributed to a gradual expansion of the crystal lattice and a concomitant increase in the frequency of molecular reorientation. At - 170' there is some residual motion which is probably due to methyl group osdations. At -43' a small but sharp line-width transition occurs, after which further motional narrowing takes place. The curve for neopentyl glycol IS somewhat different in character, remaining nearly constant from -170 to 25', where a decrease in line width begins. The observed second moment of 23 gauss2 at - 143O, obtained by numerical integration of the derivative curve, is a reasonable rigid lattice value. Possibly, the methyl (1) For a recent review, see J. G. Aston in "Physics & Chemistry of the Organic Solid State," Vol. 1, Interscience Publishers, Inc., New I'ork, N. Y., 1963, Chapter 9 . (2) W.Dannhauser, L. W.Bahe, R. Y. Lin, and A. F. Flueckinger, J . Chem. Phys., 43, 257 (1965). (3) L. E. Nielsen, Rev. Sci. Instr., 22, 690 (1951). (4) J. J. S t r a t h , F. P. Reding, and J. A. Faucher, J . Polymer Sci., A2,5017 (1964). (5) K. H. Illers, Rheol. A c t a , 3, 185 (1964).
NOTES
3739
I NEOPENTYL GLYCOL cy3
- 'j - CHI-OH
HO-HzC
cn3
-
I
32
28
-200
.-3
-100 0 TEMPERATURE, O C .
Figure 3.
36 40 44 TEMPERATURE ('C)
48
Specific volume of neopentyl glycol.
Figure
NEOPENTYL GLYCOL
v)
W 1 0 , 3 , 4
i
- I50
-100
-
50 TEMPERATURE,*C
0
+ so
Figure 2. Line width us. temperature for neopentyl alcohol and glycol.
groups in the molecule could be undergoing torsional motions of such low frequency and amplitude that the line ividth and second moment would be unaffected at this temperature. A t about 30 to 40' there is a sharp reduction in line width corresponding to a considerable increase in molecular motion. At 48.5' the second moment has decreased to approximately 2.5 gauss2 This suggests that, in addition to reorientation of various groups in the molecules, over-all molecular reorientation about one or more molecular axes takes p1ac.e. -1s the temperature is increased above SO", the line width decreases further; and this reduction may be ascribed to general reorientation and selfdiffusion of the molecules. Similar line-width curves
have been observed for other molecules with a neopentyl-like structure.6 It should be noted that as one progresses through the series neopentane, neopentyl alcohol, and neopentyl glycol the probability of hydrogen bonding increases and crystal stability should also increase. The observed nmr spectra are in accord with this view, since measurable reorientation sets in a t progressively higher temperatures as one goes from neopentane6 to neopentyl glycol. Examination was also made by means of X-ray diffraction. Powder diagrams were taken at -60 and 0" for neopentyl alcohol. The patterns obtained were clearly different. Although a detailed analysis of structure was not made, the spacings obtained at 0" are consistent with a cubic lattice-a typical result for the high-temperature phase of a plastic crystal. The -60' pattern shows a more complex structure, probably monoclinic. Seopentyl glycol is similar; in this case, diffraction patterns were taken at 0 arid 5.5'. Again, the high-temperature phase is consistent with a cubic lattice, while the low-temperature phase is complex and may be of a monoclinic structure as indicated by the work of Zannetti.7 Specific volume measurements were attempted on both materials through their transition ranges. In the case of neopentyl glycol, a 50-mg sample was placed in an aluminum foil pan and evacuated under DC-510 silicone oil to remove air bubbles. The pari, ~
~~~~
(6) J. G. Powles and H.
~
~~~~~~~~
~
L. Gutowsky, J . Chem. Phys.,
21, 1695
(1953). (7) R. Zannetti. Acta Cryst., 14, 203 (1961)
Volume 70. Szimber 11 AYozember1966
3740
suspended from the beam of a microdensity balance, was submerged in a vessel of silicone oil and the vessel heated a t a rate of about O.lo/min beginning a.t 2 5 O . From the immersed weight recorded by the balance, a specific volume us. temperature curve was calculated (see Figure 3). A sharp increase in volume was recorded in the range 40-41'. The total volume increase at the transition was about 5%; this is almost as great as the volume change of fusion, which was found by the same method to be about 6.701,. Hence, a rather substantial change of structure a t the transition is indicated. Unfortunately, this technique did not work with neopentyl alcohol. This compound is miscible with silicone oils, and indeed with all organic liquids which were tried for the purpose. Such miscibility plus the inconvenient temperature of transition (-45') make very stringent requirements for a dilatometric fluid, and so far no suitable material has been found.
Acknowledgment. We are indebted to Mr. G. D. hlcclanahan for the measurements of specific volume.
Bond Angles and Bonding in Group IIa Metal DihalidesIa
NOTES
Table I: Geometry of Group IIa Dihalidea Halide
Metal
F
c1
Br
Be
15 1 b b b
1 1
1 1
1
1 1
1 1 1
b
b
Mg
Ca Sr Ba
b b
I
1
' 1, linear; b, bent.
order to make such a prediction. The purpose of this note is to give a simple explanation of the observed geometries and thereby obtain some insight into the factors which are important in the bonding in these species. Our method is the traditional, naive method of summing up orbital energies and examining how the sum changes with angle. Of course, it is well known that the sum of the orbital energies, ef, is not equal to the total energy, but Boer, Newton, and Lipscombs have found that the sum of the orbital energies plus a constant, say A , is approximately equal to the total energy, E, i.e.
&+A=E z
by Edward F. Hayeslb Department of Chemistry, The Johns Hopkins University, Baltimore, Maryland (Received April 16, 1966)
Recently, Klemperer, et ~ l . , ~have - ~ used the electric quadrupole deflection of molecular beams to detect molecules possessing permanent dipole moments. I n these experiments, several symmetrical triatomic dihalides were investigated. Since only a bent symmetrical triatomic species can possess a permanent dipole moment, these authors were able to distinguish between linear and bent species. In the case of the group I I a dihalides, they found that the geometry depended upon the metal-halogen combination; i.e. , the linear form is favored by the light metal-heavy halogen combination, whereas the bent form is favored by the heavy metal-light halogen combination. Table I summarizes the observed geometries. At present, there seems to be no completely satisfactory explanation of the observed trends in the geometry of the group I I a dihalides. A priori prediction of the bond angle for any of these species would be very difficult. I n fact, it is not certain at what level of sophistication one would have to operate in The Journal of Physical Chemistry
where the sum i is over all the space orbitals. Now it is the hope that the approximate nature of this equality does not vary too greatly with angle so that we can at least gain some insight into how the total energy might change by considering the sum of the orbital energies. I n 1953, Walsh6 presented diagrams of valence molecular orbital energies as a function of angle for AB2 type molecules. These diagrams were constructed by considering what symmetry-orbitals could be formed using s and p orbitals on atoms A and B. Then, with the aid of intuition and experimental data, Walsh inferred how the various symmetry orbitals (1) (a) Aided by a research grant to the Johns Hopkins University from the National Science Foundation; (b) National Defense Education Act Fellow. (2) (a) L.Wharton, R. A. Berg, and W. Klemperer, J . Chem. Phye., 39, 2023 (1963); (b) A. Bachler, J. L. Stauffer, W. Klemperer, and L. Wharton, ibid., 39, 2299 (1963). (3) A. Bfichler, J. L. StaufTer, and W. Klemperer, ibid., 40, 3471 (1964). (4) A. Bllchler, J. L. Stauffer, and W. Klemperer, J . Am. Chem. SOC.,86, 4544 (1964). (5) F. P. Boer, M. D. Newton, and W. N. Lipscomb, PTOC.Natl. Acad. Sei. U . S., 52, 890 (1964). For another point of view, see S. D. Peyerimhoff, R. J. Buenker, and L. C. Allen, J . Chem. Phys., 45, 734 (1966). (6) A. D. Walsh, J . C h m . SOC.,2266 (1953).
