956
Vol. 76 [COYTRIBUTIOS FROM
THE
CHEMISTRY LABORATORIES O F xORTHWESTERN UNIVERSITY]
Steric Effects and the Stability of Complex Compounds. IV. The Chelating Tendencies of C-Substituted Ethylenediamines with Copper(I1) and NickelIII) Ions'~' BY F R E D Rr\SOI,O, -ITUN TI C I I E N XiYD
R.
K E K T AIURMilNN'
RECEIVED AUGCST28, 1953 The effect of steric hindrance on the chelating tendencies of diamines with copper(I1) and nickel(I1) ions has been investigated using C-substituted ethylenediamines. The stepwise formation constants of these two ions with HrSCR&R?S H s (R = H , C& or CsH5) have been determined a t 0 and 25'. The chelating tendency of rac-butylenediamine is less than that of meso-butylenediamine as is likewise true of rac-stilbenediamine compared to meso-stilbenediamine. Tetramethylethylenediamine, like meso-stilbenediamine, forms a yellow diamagnetic complex with nickel(I1 j of the type [Ni(AA)*]X*. Several other diamagnetic nickel(I1) complexes have been isolated as the iodide salts.
The effect of steric hindrance on the chelating tendencies of N-substituted ethylenediamines has recently been reported. Carlson, 1lcReynolds and Verhoek5 report that the chelating tendency of propylenediamine is the same as t h a t of ethylenediamine. I t would appear that this should be the case because one methyl group on a carbon atom of ethylenediamine is'not expected to alter greatly the coordinating ability of the diamine. However, i t is of interest t o determine if additional substitutions may not in fact have some effect on the chelating tendency of the bidentate molecule. This paper reports the results of such a study using dl-bn, Inbn, i-bn, dl-stien, nz-stien and Tetralleen6 with copper(I1) and nickel(I1) ions. The stepwise formation constants have been determined for these complex ions a t 0 and 2,j0 in water, except for the water-insoluble complexes of (11-stien and nz-stien in which case a 30-50 dioxane-water mixture was used.
Experimenta 1 Calculation of Formation Constants.-The stepwise formation constants were calculated by the method previously described.*b I n the investigation of dl-stien and nt-stien, lower concentrations of metal ion were employed and it was found more convenient to evaluate the formation constants by the method of Calvin and Melchior.' The dissociation constants reported earlier6 were used in these calculations of the formation constants. Reagents.-Except for Trilleen the syntheses of the diamines used have already been described.* Attempts to prepare sufficient quantities of TriMeen for these studies were not successful. However, it was possible to reduce 2-methyl-2-amino-3-ketoxitnebutane and isolate a small amount of the diamine and its nickel(I1) comples from thiq reaction mixture. 1 1 ) This investigation 5vvn5 supported h y a grqnt-in-aid from the National Institutes of Health, Grant K o G-7239 ( 9 ) Presented in part a t t h e 1.0s Angeleb RIeetinfi of the American Chemical Society. hInrch, 1$1>3 ( 3 ) Abstracted in part from a thesis b y R. Kent hliirmanu presented t o t'le graduate faculty of Sorthwestern University. in partial fulfillment of the requirements for the P h D . degree, 1954. 4) (a) H . Irving, Paper So 4, " A Discussion o n Coiirdination Chemistry," Butterwick Research Laboratories. I . C . I . , S e p t . . 21-22, 1 9 5 0 ; ( b ) F. Basolo and R li I l u r r n a n n , THISJ O U R K . A I . , 7 4 , 5243 ~ 1 % 5 2 ) ;(c) 76, 211 fl95.3) Carlsirn, J . P. \IcReynold.; and I;. 13. Verhoek. i b i , i . , 67,
Sitrosyl chlorideQwas added slowly t o a n excess of trimethylethylene10 a t -20'. The white crystals of 1chloro-1-methyl-2-nitrosobutane were removed on a filter and added t o a saturated solution of ammonia in alcohol a t 0". The temperature was then raised and the mixture refluxed for 6 hr. with continuous addition of ammonia. After removal of the solvent a t reduced pressures the residue was extracted first with hot benzene and then with sec-butyl alcohol. The benzene solution yielded the desired product, 2-methyl2-amino-3-ketoximebutane (12%); m .p. 98.0-98.5". Anal. Calcd. for CjHl?N?O: C , 51.65; N, 24.15; H , 10.34. Found: C, 51.48; h-, 24.42; H, 10.11. An additional yield of 62% of the hydrochloride (m.p. 192-194') was obtained from the sec-butyl alcohol solution. Attempts to prepare the desired diamine by a catalytic (PtO?) hydrogenation of the ketoxime hydrochloride and by electrolytic reduction were not successful. Reduction with tin( 11) chloride in concentrated hydrochloric acid gave almost a quantitative yield of 2,2,3,5,5,6-hexamethyl-2,5-dihydropyrazine" (m.p. 64-65'). It was finally possible to isolate only 1.5 g. of the diamine (b.p. 138-142') and 2.2 g. of [Ki(TriMeen)?]I? from the reduction of 100 g. of 2methyl-2-amino-3-ketoximebutanehydrochloride with sodium amalgam and alcohol. Dioxane was purified by the method of Lf'eissburger and Proskauer.'? The other chemicals were all of reagent grade. Standard Solutions.-Aqueous solutions of the diamines, except dl-stien and wi-den, were standardized by titration with standard acid. The dioxane-mater (50y0by volume) solutions of dl-stien and 7%-stien of approximately 0.b XI concentration were standardized by conductometric titration with standard perchloric acid. Standard aqueous solutions of copper(I1) nitrate and nickel(I1) nitrate were prepared which contained approximately 0.05 M metal ion, 0.1 AI nitric acid and 0.5 2 ' potassium nitrate. A 50% dioxane-water solution of approximately 0.05 JI copper perchlorate was standardized by electrodeposition. A 100-cc. portion of this standard solution was diluted with an approximate amount of standard perchloric acid solution to result in a mixture containing 0.005 AI copper(I1) ion anti 0.025 M perchloric acid. Similarly a standard 50% dioxane-water solution of 0.005 JI nickel(I1) perchlorate and 0.025 -11 perchloric acid was prepared. This choice of perchlorate as the anion in the dioxane-water solutions was made as the perchlorate complexes were somewhat more soluble than the nitrate salts. Furthermore, no supporting electrolyte was used in these mixed solvents. Titrations.-The apparatus and procedures employed were the same as described in paper I of this series.4b However, it was necessary t o prepare a calibration curve to correct for changes in the activity coefficient of the hydrogen ion in the dioxane-water solutions. The results indicated that the actual pH meter readings were correct within esperimental error below a PH of 9, which for the systems reported here was well outside the range of complex formation. Continuous Variation Studies.-The method of Vosburgh and Cooper13 was applied to solutions of nickel(I1) ions and
T h e diamines are desixnatcd as follons: en = ethylenediamine; .~ i!l-hn = i'nr-%.X-diaminobutane, In-hn = iiii,so-2.3-diaminobutane, z-hn = 2 - m e t h ~ l - 1 , 2 - d i n m i n o ~ ~ r ~ ~ ~ a (!a)n e : G. H . Coleman, G. A. Lillis a n d G . B.Goheen, "Inorganic Syntheses," T'ol. I , John Wiley and Sons, Inc., Piew York, N . Y . ,1939, p . .55. T r t r a I l e e n = L',X-dirnethyl-2.X-diaminohutane;di-stien = r o r - l . ? - d phenylethylenediamine; in-stien = mesn-1.2-diphenylethylenediamine: (10) A , Michael a n d F. Zeidler, A w n . , 385, 259 (1911). TriXIeen = 2-methgl-?.3-diaminobutane. (11) S. Gabriel, Ber., 44, 61 (1911). ( 7 , 11. Calvin and pi C I l e l r h i m T H I S J o u n Y A l , 70, 8270 (12) A IVeisshurger a n d E. S . Prmkauer, "Organic SoI\-ents," ( I 9-18) Oxford Press. 1935, p. 184. (13) V.7 C . Voshurgh and G. H . Cooper, T H I S J O I J K P i . a . , 63, 437 ( 8 ) F. Hasolo R. Kent I l u r m a n n and l'un 'Ti Chen. i b i ) l . . 7 6 , 1178 (1953) 1041 ) . IO)
p r l = propylenediamine;
Feb. 20, 1954
ETHYLENEDTAMINES WITH COPPER(II) AND NICKEL(II) IONS
TetraMeen. The absorption spectra and optical density measurements a t selected wave lengths were made with a Beckman DU spectrophotometer on solutions of from t o IO-* molar concentation. Absorption Spectrum and Formation Constant of [Ni( T e t r a M e e ~ ~+Z.-The )~] absorption spectrum of an aqueous solution of [ Ni(TetraMeen)z]( ~\'OS)Z was determined and observed to obey Beer's law upon dilution. However, when an excess of acid is added t o the yellow solution the complex dissociates a t a measurable rate (half-life of approximately 142 sec. a t 25'). Therefore, several solutions of the complex containing varying amounts of acid were thermostated a t 15, 25 and 39" until equilibrium was reached, a t which time both the PH and optical density of the solutions were measured. These data were used t o calculate the total formation constant of the complex for the reactionI4 M+*
+ 2A
[MAz]+~
0.7
0.6
0.5
h; 0.4 . . : M c
where
0.3
Since it was shown (Figs. 1 and 2) that the concentration of [MA] 12 must indeed be very small, it follows that
0.2
CM
C.4 = [A]
=
[M"]
+ [MAz+']
+ [ A H + ] + [AHs+'] + 2[MA*+']
and
CH = [ H + ]
+ [AH+] + 2[AHz+']
(2) (3)
0.1
(4)
From the dissociation constants of the diamine
together with equations 2, 3 and 4 it follows that
and
Therefore substituting equations 2, 7 and 8 into 1, the formation constant of the complex is calculated. The concentration of [MAz]+z can also be obtained directly from the spectrophotometric data since
The molecular extinction coefficient, E, is obtained from the optical density of solutions containing known amounts of nickel(I1) and an excess of the diamine. Yellow, Diamagnetic Complexes of Nickel(II).-The addition of excess TetraMeen to a nickel(I1) solution yielded a compound which could be isolated either as the insoluble iodide or the slightly soluble chloride. Anal. Calcd. for NiC12H32K4C12: X, 16.24; c , 39.8; H , 8.85; N, 15.5; C1, 19.63. Found: S i , 16.35; C, 39.6; H, 8.64; N, 15.4; CI, 19.68. Conversion t o the more soluble [Xi(Tetrame en)^] ( 5 0 3 ) was ~ accomplished by grinding the iodide with slightly less than the calculated amount of concentrated aqueous silver nitrate and extracting the reaction mixture with water a t 80'. Recrystallization from a n ethanolwater mixture produced the pure nitrate which has an absorption band a t 434 m p with a molecular extinction coefficient, e, of 63.8. A yellow solution was formed when solutions of nickel(I1) and TriMeen were mixed a t a pH of 9. The yellow iodide, [Ni(TriMeen)?]IZ was precipitated upon the addition of excess potassium iodide and was recrystallized from ethanolwater. Anal. Calcd. for KiC1~Hz8N4IZ: Xi, 11.37; C, 23.55; H , 5.41; h*,10.82; I, 49.08. Found: Ni, 11.32; C, (11) Symbols used are the same as those in reference 4h.
957
440 480 520 560 mp. Fig. 1.-Absorption spectra of solutions containing different mole ratios of TetraMeen. Concentration of Xi(h'03)zkept the same a t 0.01 Jf whereas that of TetraMeen was changed to give the mole ratios designated. Measurements were made a t room temperature with a Beckman Du spectrophotometer using 1-cm. quartz cells. 360
400
23.88; H, 5.47; N , 10.80; I , 49.1. ( e a t 440 mp is 54.7.) Essentially the same procedure was used for the prepara-
bnj2]12. An aqueous solution containing one equivalent of nickel(I1) nitrate and two equivalents of diamine dichloride was adjusted to a p H of 9 by the addition of potassium hydroxide and then saturated with potassium iodide. Upon standing overnight yellow crystals deposited which were recrystallized from acetone. Anal. Calcd. for S T i C ~ H ~ 4 K 4C1,~19.62; : H, 4.91; N, 11.45. Found for [Ni(m-bn)2]Ig: C, 20.43; H, 4.93; S,11.53. Found for [Xi(i-bn)2]12: C, 19.68; H , 4.92; N, 11.49. ( e a t 448 m p is 21.8.) The complex [Si(dl-bn)*]I?
0.30
* ii"
-
0.25 0.20 0.15 0.10 0.05 0.00
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Mole fraction of TetraMeen. Fig. 2.-Continuous variations studies on the system TetraMeen.
X + 2 -
FREDBASOLO,krUN TI CHEN AND R.KENT1fURIIZANN
958
was instead recrystallized from a concentrated aqueous solution of potassium iodide. Anal. Found: C , 20.00; H , s.09; S , 11.56. Calorimetric Measurements.-These measurements of heats of reaction for the formation of [Cu(TetraMeen)n]+? and [Ni(TetraMeen)2]+ 2 salts were made in a Dewar-type calorimeter by the procedure described previously.4c
Results The formation constants for copper(I1) and nickel(I1) complexes are shown in Table I and 11, respectively. The data on the stilbenediamine compounds were obtained in tj070 dioxane-water solution and these are not comparable to the values obtained in aqueous systems. TABLE I FORMATION CONSTANTS O F COPPER(I1) WITH e-SUBSTITUTED ETIIYLEXEDIAMISES 03C. log
log
Kl
en pn dl-bn m-bn i-bn TetraMeen dl-Stien rn-Stien
11.34 11.65 12.22 11.50 11.31 12.22 11.03 9.30
2.j°C.
lo,.
Kr
KT
9.95 10.12 10.65 10.05 9.88 10.88 10.21) 8.20
21.29 21.77 22.87 21.55 21.19 23.10 21.23 17.50
log
log
Kr
k'i
10.76 9.37 10.78 9 . 2 8 11.39 9.82 10.72 9 . 3 1 10.53 9.05 11.63 1 0 . 2 1 10.07 9 . 1 1 8.46 7.50
log
Kr
20.13 20.06 21.21 20.06 19.L5X
21.87 19.18 Iq5.9fi
TABLE I1 FORMATION COSSTASTS OF SICKEL(II) WITH ETHYLESEDIAMISES ki
log
Kr
25'C.
log KJ KT 3 . 3 6 20.05 4.92 19.77 5 00 20 34 3 21 16 74 2 77 15 118 log
log
Ki
Calorimetric data obtained for the reactions of nickel(I1) and copper(I1) ions with TetraMeen are shown in Table 111. Likewise included for comparison in this table are the thermodynamic data for the analogous reactions with ethylenediamine. Magnetic susceptibilities of some yellow-orange complexes are given in Table IV. TABLE111 THERMODYNAMIC DATA [hI(H20)xla,+2 2e11,,, [Mena(H20),-aInil L 2 iH,O [M(H2O)sIns+* 2TetraMeena, [M(TetraMeen)S];,,,+2 6H20
+
+
Amine
en TetraLTeen
AFO -18.6" -20.0
+
=
Sickel AHQ
+
AS0
- 1 B Xa
+7
AFO
-I4 1
+20
-26.6 -2!j 9
Copper AH0
-24 -24
AS0
64'
+
4
+I9
i
" G . H . McIntyre, J r , , P h . D . Thesis (19S3), Pennsylvania State College. TABLE IT MAGNETICSUSCEPTIBILITYOF SOME[Si(AA)2)Xz CoxComplex
PLEXES~ Color
[Si(TetraMeen)2]C12 [ Si(m-bn)z]I? ISi(i-bn)2]Iy [ Si(dl-bn)2]12 [ S i (TriMeen)2]12
Yellow Yellow Orange Yellow Yellow
x, -0,filE - ,428 - ,205 - ,306
X 10P X X 10F6 X 10-0
.... ... . . . . .
Magnetic measurements were made using the Gouy C-SUBSTITUTED method by Mr. Steve Adler.
00C. log
Vol. '76
log Kr
log
Ka
log
KT
Discussion
The results of this study show that generally Csubstitution on ethylenediamine has only a relatively small effect on its coordination tendency tod-bn v-hn ward copper(I1) and nickel(I1) ions. However, it i-bn is of interest that the complexes of dl-bn are more TetraMeen stable than those of m-bn which is also true of dltil-stien i 08 7 98 4 '44 19 I i l 7 . 3 7 6 00' 4 . 3 8 stien compared to m-stien. These results on dl4 !li 3 72 8 6!4 4 . 7 0 3 . 5 2 ril-stlen stien and m-stien are in accord with the qualitative a I'alue of log K Land log Ky obtained from spectrophotometric data and PH measurements. This value a t 15' is observations of Lifschitz, Bos and Kijkema.'j X 11.87 a d .it 39" it is 14.32. consideration of molecular models shows that the five-membered chelate of the racemic bases are free The absorption spectra for solutions containing of internal strain as the methyl or phenyl groups different ratios of nickel(I1) and Tetrahfeen were de- have ample room being on opposite sides of the ring. termined between 360 and 800 mp. I n all of these However, the meso bases form rings with the subsolutions only one absorption band was found and stitution on the same side and there would appear that at 4.74 mw. The absorption at this wave length to be considerable internal strain resulting from of a solution containing equivalent amounts of crowding of the adjacent groups. nickel(I1) and Tetrahleen is, within experimental In order to test this hypothesis the chelating error, exactly half that of one containing the same tendencies of Tetrahfeen and i-bn were determined. amount of nickel(I1) and two equivalents of dia- It should be noted that the resulting five-membered mine (Fig. l). I t would therefore appear that either chelate ring of TetraMeen has adjacent methvl virtually all the diamine is being used to form [Ni- groups on both sides of the ring, whereas with i-bn (Tetra?lIeen)e\ +? or that [?\Ti(TetrabIeen)(Hz0)4]+? the two methyl groups are on the same carbon. has the same absorption spectra as the bis complex. Since the complexes of Tetrahleen are more The latter possibility does not appear very probable stable while those of i-bn are less stable than since the absorption of the mono complex may be those of either dl-bn or m-bn, i t may appear expected to more closely resemble that of [Ni-en- as if any strain within the chelate ring in these ( H 2 0 ) 4 ] + 2 . That the complex in solution is in fact systems must be playing a minor role. It is is supported by the re- of interest, however, that these results closely the ion [Ni(TetralIeen)2]L2 sults of a continuous variation study shown in Fig. parallel the findings of Kuhn16 on the intra2 . Similar studies with copper(I1) and TetraLIeen I. Lifschitz, J. G. Bos and K . M. Kijkema, Z . a n o y g . allgpm. reveal that both the mono and bis complexes are C h(15) r m . . 242, 97 (1939); I . Lifschitz and J. G. Bos, Rev. frnu. chim., 59, formed in appreciable concentrations as is also true 407 (1940); I. Lifschitz and K. ?J.Kijkema, ibid., 60, 581 (1941). of nickel(I1) with m-stien and with Trilleen. (16) I,. P. Kuhn, THISJ O U R N A L , 74, 2492 (1952). en
pn
7.92 6.77 8 05 6 81) 8 30 7 04 i 37 6 16 7 . 2 7 5 94
7.60 7 43 7.71 7 04 6 77
6.48 5.03 6 . 1 9 4 27 63.48 4 . 3 1 5 . 7 0 2.89 5 40 2 . 2 5
10.11 17.89 18.50 15.63 14.42 1 4 , 6Sa 17 84 8 22
Feb. 20, 1954
ETHYLENEDIAMINES WITH COPPER(I1) AND NICKEL(II)I O N S
molecular hydrogen bonding in alcohols. This process of hydrogen bonding involves the formation of a ring and is somewhat analogous to the chelation which occurs during the coordination of diamines with metal ions. Kuhn reports that the extent of hydrogen bonding in d-butane-2,3-diol and also in tetramethylethylene glycol is greater than that in meso-butane-2,3-diol. An interesting observation was made during the titrations of nickel(I1) salt solutions with alkylethylenediamines. The color of the solutions in the p H region or maximum concentration of the bis complex, changed from the typical violet-blue of tetramminenickel(I1) ion through greenish-yellow to a bright yellow as the extent of C-substitution on the diamine increased. Except for solutions containing either TetraMeen or m-stien which remained bright yellow, the addition of excess diamine produced the characteristic violet color of hexamminenickel(I1) ion. ' In other words, nickel(I1) does not appear to coordinate with more than two molecules of either TetraMeen or m-stien. I n fact a sample of [Ni(TetraMeen)z]( N O & !has remained unchanged in anhydrous TetraMeen a t room temperature for one month. -4 consideration of molecular models show that coordination of a third molecule of TetraMeen would involve considerable steric hindrance. Several of these yellow salts have been isolated and found t o be diamagnetic (Table IV). It therefore appears that they are of the dsp2-typeand thus have a planar configuration. It is of interest to speculate as to why this tendency toward the formation of a planar complex should be enhanced b y the increased C-substitution on ethylenediamine. One possible interpretation is that the increased alkyl substitution renders the bis complex more hydrophobic such that the coordinated water molecules in the complex [Ni(AA)z(H20)2]+2are repelled to yield the tetracovalent ion [Ni(AA)2]+2. This assumption receives some support from the fact that even the ethylenediamine complex [Nien2(HsO)2]+2has been shown" to produce an orange diamagnetic salt of the ion [Niens]+ 2 upon removal of the coordinated water. That the formation constant Kz is appreciably larger than Kl for the stepwise formation of [Ni(TetraMeen)z]+ 2 is not without precedent. For example, K 3 is larger than either Kl or Kz for the formation of tris-(l,10-phenanthroline)-ion(II).'8 I t has been suggested4' that unusual behaviors of this type may result from sudden changes in bond type or orbital hybridization. For example the (17) H. Glaser and P. PfeilTer, J . p r a k l . Chern.. 163, 300 (1939); C M . Harris, Pror Roy. SOC.N . S.Wales, 85, 142 (1953). (18) T. S. Lee, I. M.Kollhoffand D. L. Leussing, THISJOURNAL, 70, 2348,3596 (1948); l a , 2173 (1950).
968
mono and bis iron(I1) complexes may be primarily ionic (sp3d2)and on coordination of a third molecule of 1,lO-phenanthroline from the covalent d?sp3tris complex. In an analogous fashion one may suggest that the nickel(I1) complex [NiTetraMeenchanges to the more covalent ( H ~ 0 ) 4 ] +-sp3d2 ~ bond structure [Ni(TetraMeen)s]+ 2 -dsp.* An alternative e ~ p l a n a t i o n 'to ~ account for the large value of K , as compared to Kl is that a greater increase in entropy is associated with the second step [SiTetraMeen(H10)4] +z
+ TetraMeen If
[Si(TetraMeen),] +z
+ 4Hp0
than the first step
+
[Ni(HZO),]+2 TetraMeen [SiTetraMeen(HzOh 1 f 2
+ 2H20
The greater increase in entropy may be expected to result from the release of four molecules of water in the second step as compared to only two for the first step. Consistent with this view is the fact that an entropy increase of +'i cal./deg. was observed for the reaction
+ 2en
[Si(H,0)ti]+2
Whereas a value of +PO (Table 111) for
+
[Sienz ( H Z O ) ~ ] + 4H20 ~
cal./deg. was obtained
+
[ i V i ( H ~ 0 ) ~ ] + 2TetraMeen ~
+ 6H20
[Si(TetraMeen)?If2
It is believed that the greater increase in entropy for the TetraMeen reaction compared to that of the ethylenediamine (en) reaction can in part a t least be attributed to the release of two more molecules of water. However, approximately the same difference in entropy was observed for the analogous reactions of copper(I1). Since the coordination number of copper(I1) is generally believed to be four, it should follow that in both of these reactions [Cu(H,0),lr2 + 2AA JJ [CU(AA)Z]+' 4Hz0
+
AA = TetraMeen or en
four molecules of water are released. It would therefore appear that either the amount of water released does not make a major contribution to the changes in entropy or that in aqueous solution there are in fact six molecules of water associated with the copper(I1)ion. The latter reason would seem to be the more plausible one as there have been several reports of hexacovalent copper (11) complexes.20 EVANSTON. ILLINOIS (19) M. Calvin, private communication. (20) P. Pfeiffer and K. Pimmer, Z. anoig. Chctn., 48, 98 (1905): F. Rosenblatt, i b i d . , 204, 351 (1932); J. M. Jaeger and J , H. V a n D y k , Proc. Koninkl. Akad. Amsterdam, 37, 395 (1934).
