Vol. 77
NOTES
where the brackets iiidicatc niolar colicelitration and X corresponds to the ethanolamine used. The reciprocal slope values given in Table I have a wide range of values and this causes some uncertainty in the determination of the values of Kf. Therefore, the K fvalues in Table I1 are given tCJ only one significant figure. For each of the ethanolamines studied, the value of p obtained was four which indicates that the complexes contain four ligand molecules. Since copper(I1) ordinarily has a coordination number of four, this shows that the coordinating molecules are monodentate. I t is assumed that the coordinate bond occurs through the amine nitrogen. The diethanolamine complex found was quite similar in behavior to the monoethanolamine and ethylethanolamine complexes, and this points to the likelihood of amino coiirdiiiation. Diethanolamine has one more hydroxyl than either monoethanolmiine or ethylethanolatniiie. Hence, for h>= droxyl coiirdination, diethanolaiiiine should behave differently from the latter two compounds. All three compounds should act in a similar manner in the formation of complexes through the amine nitrogen. The fact that it was found that four molecules of the ethanolaniines were coordinated in all three cases substantiates the assumption that the linkage is through the amine nitrogen. Some further studies with diethylethanolarnine and triethanolamine in this Laboratory indicate that the corresponding complexes are not formed in the same concentration range as the others given here, but show a tendency toward coordination of only two or three molecules Iier copper(1Ij. The square-planar configuration of copper(I1) complexes might produce steric hindrance if the coordination occurs through the tertiary anline nitrogen. This would riot be the situation if i t were the hydroxyl group which coordinates. This is in agreement with the beliefs of other authors. Rreckenridge arid Hodgiris-l and l l a n n j have described complexes of 1,8-diatnino-2 propanol with copper(TI) for which they believe the coordination took place through the amino groups only. Hreckeiiridge6 states that for I\--hydroxyethylethyleriediamine the hydroxyl group does not coordinate. Harvey, et describe a number of complexes of K-hydroxyethylethylenediaminestudied spectrophotometrically. Among the structures mentioned is [Cu(HOC2H4NHCzH4NHaj4]1' for which they reason that only the primary amino group is cotirdinated. They also give evidence for the existence of a coinplex of diethanolamine with copper(I1) in which the amine group, rather than the hydroxyl group, is coiirdinated. If the difference in coijrdinating ability of the aiiiino aiid hydroxvl groups is substantially different, then one should expect to find only one type of group coordinated when an excess of the cornplexing agent is present. This assumes that there are no complications such as steric effects. Thus the experimental conditions of this research favored the formation o f the [CuX4]+ + cniiiplex. However I 11 J . ( > . H r r c k e i i r i i l c r .3:4 I (I!l3!lj
( ( i j J . (;.
iiii(l
l 3 r e c k c i i r i ~ l q r ,i
1 . LV
,it,
.i
B17,
I
B26, 1 I
(l!ilS).
in the presence of a limited amount of the ethaliolamines, particularly in non-aqueous solutions where water does not compete, i t is possible t h a t one may find chelates formed with both the amino arid hydroxyl groups being coordinated. One can find support for this hypothesis in the results of earlier workers based on analyses of compounds prepared under these conditions. It would be of interest to compare the forniatioii constants for the ethanolamine complexes of c o p per(I1) with the corresponding simple amine cornplexes, but the data on these do not appear to be available. However the silver complexes have been reported and it has been found that two moles of the complexing agent are coordinated per atom of silver(1). The formation constants reported7 for the monoethanola~iiine,diethanolamine, ethylairline and diethylamine complexes of silver are, respectively, 4.S X 1 06, 0.30 X IO6, 0.2 x IO6 and I .(iX IO6. These values indicate that the simple ~ I I I iries and the hydroxj- substihted amines are not very different in coiirdinating ability. I t is suggested that the water solubility of these ethanolamine complexes is enhanced by the presence of thc hydr groups extending into solution. This may ain the fact that certain I C hydroxyethylethylenedianiinetriacetic acid coniplexes are more soluble than the corresponding complexes with ethylenediaminetetraacetic acid. ( 7 ) J. Bjerrurn, C ' i f ~ l i r R r i s , 46, DEPARTMENT ( I Y HEMI IS TRY
381 (19:O).
1v.4YSE y S I \ - E R S I T V
DETROIT, hlr CH I GA s ..
..~_ . ~~~
On the Mechanism of the Acid-catalyzed Rearrangement of Siloxane Linkages in Organopolysiloxanes BY DALLAS T . H C R D RECEIVED JASCARY : 3 , 19L5
I t has beeii well established that the actiL-c agent in the base-catalyzed rearrangement of siloxane linkages in polydiinethylsiloxanes is the oxygen atom, or ionic form derived therefrom, which acts as an electron donor to silicon in the siloxane linkage exchange niechariisi~i.~,'Since the rate of siloxane rearrangement depends on the effective strength of the oxygen as an electron donor, determined in turn by the nature of the group or groups attached to it, it seeins improbable that the acidcatalyzed rearrangement oi siloxane linkages would follow a course entirely analogous to that of basecatalyzed rearrangement. I t is known that, like basic catalysts, protoiiic acid catalysts do react with siloxane niolecules in the course of the rearrangeinent reaction and the products formed are formally arialogous to those formed with basic catalysts."-" The reaction, either with base or acid, involves a cleavage of the siloxane (1) D. T. Hurd, R.C Osthoff a n d 11,I,. Corrin, T H I SJ O U R N A L , 76, 249 (1954). (0) XV. T. Griibb and I
