- Email: [email protected]
Resolution of neutral complexes of transition metals by stereoselective adsorption on optically active complexes
&x.unal of Chrqnra:o~raphu.
198 (1980) 172-175
EIsevier Scientific Publishing Company, Amsterdam -
Printed in The Netherlands
CHROM. 12.961
Note Resolution adsorption
of neutral complexes of transition on optically active complexes
il. Partial resolution nato)chromium(lll)
metAs
ef tris(acetylacetonato)cobalt(lII) on cobalt(lll) complexes
by stereoselective
and tris(acetylaceto-
M. B. CELAP+, 1. M. HODZIC and T. J. JANW Institute
of
Chemistry.
Facdty
of Sciences,
Umkersity
of Beograd,
P-0.
Box
550, IIOOl
Beograd
( Yugoslavia)
(First received March 18th. 1980: revised manuscript received May 22nd. 1980)
A large number of papers have described the chromatographic resolution of metal complexes into enantiomers’*L. However, only a few of them refer to the resolution of neutral complexes by stereoselective adsorption on dissymmetric adsorbents acid’, such as qua&, sodium chlorateq, lactose**‘, alumina treated with (+)-tart&c D-sorbitol and D-manitol* and starch9. None of the studies employed an optically active complex of a transition metal as adsorbent_ Hence we have investigated the resolution of neutral complexes using seven optically active cobaltOH) complexes that had been prepared in our laboratories for the first time, i.e.:
I(f)589-KCCO(g~Y)Z(NO*)*110~f ~I(-_),,9-KECo(gly),(N03,1’0 III(+),,9-K~Co(L-a~a),(N03,111
.
IV(-),,,-[Co(L-arg)~(N0~,]C1*2
V(+),,[Co(L-arg),(N0,),lC1’2
VI(-_),,~-[Co(L-arg)~(N03~l-(-_),,-~Co(gly)XN03,1’3
v~I(-_)~~9-CCo(L-arg)~~03~l-~+)~~9-CCo(gly)~(N03~1*3
All these complexes represent very stable cis(NOJ-tram(N) geometrical isomers which suffer neither physical nor chemical changes in the course of time. The neutral complexes investigated were tris(acetylacetonato)cobalt(III) and tris(acetylacetonato)chromium(III), which had been prepared according to known proced~re-&‘~. These complexes were chosen because they are soluble in benzene, whereas the compounds used for their resolution are insoluble. The procedure involved the dissolution of 0.20 g of the neutral complex in 5 cm3 of benzene. The resulting solution was passed through a column cf diameter 1 cm and adsorbent height 60 cm, the adsorbent grain size ranging from 50 to 100 mesh. The complex was first adsorbed to the column and then eluted with benzene. Fractions of 15 drops were collected and then diluted up to 3 cm3 with benzene. The concentration of solutions coming off the co!umn was determined spectrophotometril
glytt = Glycine; alatt = &mine; arg = arginine.
0021-9673/8O/oooo-oooo jSO2.25
0
1980 Ekevier
Scientific Pub!ishing
Company
NOTES
173
tally at a wavelength of 577 nm, using a Beckman DLL2 spectrophotometer. Optical rotation of solutions was measured by means of a Perkin-Elmer 141-MC polarimeter at a wavelength of 546 nm. RESULTS
AND
DISCUSSION
From the results given in Table I it may be seen that a partial resolution of the two neutral complexes has been achieved in all the cases investigated. TABLE I MAXIMAL Adsorbent
VALUES Experiment
OF APPARENT MS
MOLECULAR
ROTATIONS
(degrees)
co(~d%~Z),
wcd%w2
I
1 2
-i- 179.9 + 188.3
- 420.4 - 356.0
- 48.6 - 55.6
+ 66.8 i- 71.8
II
1 2
-
136.0 172.4
+ 379.9 f 369.9
+ 50.4. + 52.4
- 61.6 - 58.2
III
1 2
+ 219.3 + 228.2
- 291.2 - 275.2
- 54.1 - 58.2
+ 58.9 + 55.3
IV
1 2
- 346.6 - 306.0
+ 295.4 + 302.6
+ 71.2 + 68.7
- 72.6 - 79.2
V
1 2
- 226.4 - 209.3
_t 358.8 + 335.7
+ 69.8 + 65.4
- 52.5 - 53.4
VI
1 2 1 2
+ + + +
- 548.7 - 559.7 - 169.9 - 159.6
-
+ + + +
VII
278.3 283.8 284.1 280.2
64.5 66.3 55.3 52.8
50.8 40.2 51.4 48.7
The maximal values of the apparent molecular rotations obtained in the resolution of tris(acetylacetonato)cobalt(lIi) were higher than those obtained with tris(acetylacetonato)chromium(III), using the same adsorbent. This may be due either to a higher actual molecular rotation of the cobalt(II1) complex or to its more complete resolution. However, since the same phenomenon was observed with other adsorbent@, the most likely explanation is a higher molecular rotation of the cobaltIII) complex relative to that of the chromium(II1) complex. It may also be seen (Table I) that the two complexes exhibit opposite elution sequences of the optically active isomers, using the same adsorbent. Fig. 1 shows that the first fractions coming off the column exhibit the greatest values of the apparent molecular rotation, which gradually decreases. Middle fractions are inactive, and they are followed by fractions with increasing molecular rotation of opposite sign to that of the first fractions. Thus in the course of these experiments, only a partial resolution of the tris(acetylacetonato) complexes of cobalt(III) and chromium(III), has been achieved. A similar partial resolution of these complexes has also been effected by the use of
174
Fig. 1. Functional dependence of the apparent molecular rotation for tris(acetylacetonato)cobaIt (-- - -) on fraction number in the course of WI) (--- ) and tris(acetylacetonato)chromium(III) sepantion on a column packed with c~NOr)_rru~NH-)sg~-[Co(L-arg)~(NO,)rlCI.
adsorption columns of lactose6, alumina treated with (+)-tartaric acid’, o-sorbitol and D-mannitoP_ The achievement of only partial resolution cannot be due to racemfzation in the course of the separation, since we have established experimentally that these complexes do not racemize upon standing in benzene solution even for a timeperiod longer than that required for the elution in our resolution experiments. One reason for the incomplete resoluticn is the relatively low plate number of the system used due to the large grain size of the adsorbent. However, the use of adsorbents of smaller grain size was not possible since the elution was not performed under high pressure. Experiments aimed at increasing the efficiency of the column are now being made.
REFERENCES I Y. Yoshikawa and K. Yamasaki, Coord. Chem. Rev., 28 (1979) 205. 2 S.Ya mazaki, T. Yukimoto and H. Yoneda, J_ Chromotogr_. 175 (1979) 317. 3 4 5 6 7
R_ Tsuchida, M. Kobtiyashi and A. Nakamura, Bull. CIrem_ Sot. Jap., $1 (1936) 38. E. Ferroni and R_ Cini, I_ Amer. Chem_ Sot., 82 (1960) 2427. T. MoeiIer and E. Gulyas, J. Inorg. Nucl. Chem., 5 (1958) 245. J. P. Co&tan, R. P, BIair. R. L. Marshall and S. Slade, inorg. Chem., 2 (1963) 576. T. 4;. Piper, J. Amer. Chem. Sot., 83 (1961) 3908. 8 N. S. Bowman, V. GEeva and G. K. Schweitzer, inorg. Chem. Left., 2 (1966) 2.51.
H. Krebs and R. Rasche, Narurwisenschnfren, 41(1954) 63. M. E. &lap, D. J. Radanovi~ and T. J. Tanjie, inorg. Chem., 4 (1965) 1494. M. B. &lap, D. J. Radanovi~, T. I. NikoliC and T. J. JanjE, Inorg. Cfiim. Acta. 2 (1968) 52. B. A. ?Xamberi,M. B. &lap and T. J. Janj& Gkzs. Hem. Drti.. Beograd, 43 (1978) 149. M. B. &!a~, P. N. RadivojZa and G. Djmovi& in preparation. L. F. Auckfeth, (Editor), Inorg. Syn., 5 (1950) 188. 15 L. F. Audrieth (Editor), Borg. Sym, 5 (1950) 130.
9 10 11 12 13 14
198 (1980) 172-175
EIsevier Scientific Publishing Company, Amsterdam -
Printed in The Netherlands
CHROM. 12.961
Note Resolution adsorption
of neutral complexes of transition on optically active complexes
il. Partial resolution nato)chromium(lll)
metAs
ef tris(acetylacetonato)cobalt(lII) on cobalt(lll) complexes
by stereoselective
and tris(acetylaceto-
M. B. CELAP+, 1. M. HODZIC and T. J. JANW Institute
of
Chemistry.
Facdty
of Sciences,
Umkersity
of Beograd,
P-0.
Box
550, IIOOl
Beograd
( Yugoslavia)
(First received March 18th. 1980: revised manuscript received May 22nd. 1980)
A large number of papers have described the chromatographic resolution of metal complexes into enantiomers’*L. However, only a few of them refer to the resolution of neutral complexes by stereoselective adsorption on dissymmetric adsorbents acid’, such as qua&, sodium chlorateq, lactose**‘, alumina treated with (+)-tart&c D-sorbitol and D-manitol* and starch9. None of the studies employed an optically active complex of a transition metal as adsorbent_ Hence we have investigated the resolution of neutral complexes using seven optically active cobaltOH) complexes that had been prepared in our laboratories for the first time, i.e.:
I(f)589-KCCO(g~Y)Z(NO*)*110~f ~I(-_),,9-KECo(gly),(N03,1’0 III(+),,9-K~Co(L-a~a),(N03,111
.
IV(-),,,-[Co(L-arg)~(N0~,]C1*2
V(+),,[Co(L-arg),(N0,),lC1’2
VI(-_),,~-[Co(L-arg)~(N03~l-(-_),,-~Co(gly)XN03,1’3
v~I(-_)~~9-CCo(L-arg)~~03~l-~+)~~9-CCo(gly)~(N03~1*3
All these complexes represent very stable cis(NOJ-tram(N) geometrical isomers which suffer neither physical nor chemical changes in the course of time. The neutral complexes investigated were tris(acetylacetonato)cobalt(III) and tris(acetylacetonato)chromium(III), which had been prepared according to known proced~re-&‘~. These complexes were chosen because they are soluble in benzene, whereas the compounds used for their resolution are insoluble. The procedure involved the dissolution of 0.20 g of the neutral complex in 5 cm3 of benzene. The resulting solution was passed through a column cf diameter 1 cm and adsorbent height 60 cm, the adsorbent grain size ranging from 50 to 100 mesh. The complex was first adsorbed to the column and then eluted with benzene. Fractions of 15 drops were collected and then diluted up to 3 cm3 with benzene. The concentration of solutions coming off the co!umn was determined spectrophotometril
glytt = Glycine; alatt = &mine; arg = arginine.
0021-9673/8O/oooo-oooo jSO2.25
0
1980 Ekevier
Scientific Pub!ishing
Company
NOTES
173
tally at a wavelength of 577 nm, using a Beckman DLL2 spectrophotometer. Optical rotation of solutions was measured by means of a Perkin-Elmer 141-MC polarimeter at a wavelength of 546 nm. RESULTS
AND
DISCUSSION
From the results given in Table I it may be seen that a partial resolution of the two neutral complexes has been achieved in all the cases investigated. TABLE I MAXIMAL Adsorbent
VALUES Experiment
OF APPARENT MS
MOLECULAR
ROTATIONS
(degrees)
co(~d%~Z),
wcd%w2
I
1 2
-i- 179.9 + 188.3
- 420.4 - 356.0
- 48.6 - 55.6
+ 66.8 i- 71.8
II
1 2
-
136.0 172.4
+ 379.9 f 369.9
+ 50.4. + 52.4
- 61.6 - 58.2
III
1 2
+ 219.3 + 228.2
- 291.2 - 275.2
- 54.1 - 58.2
+ 58.9 + 55.3
IV
1 2
- 346.6 - 306.0
+ 295.4 + 302.6
+ 71.2 + 68.7
- 72.6 - 79.2
V
1 2
- 226.4 - 209.3
_t 358.8 + 335.7
+ 69.8 + 65.4
- 52.5 - 53.4
VI
1 2 1 2
+ + + +
- 548.7 - 559.7 - 169.9 - 159.6
-
+ + + +
VII
278.3 283.8 284.1 280.2
64.5 66.3 55.3 52.8
50.8 40.2 51.4 48.7
The maximal values of the apparent molecular rotations obtained in the resolution of tris(acetylacetonato)cobalt(lIi) were higher than those obtained with tris(acetylacetonato)chromium(III), using the same adsorbent. This may be due either to a higher actual molecular rotation of the cobalt(II1) complex or to its more complete resolution. However, since the same phenomenon was observed with other adsorbent@, the most likely explanation is a higher molecular rotation of the cobaltIII) complex relative to that of the chromium(II1) complex. It may also be seen (Table I) that the two complexes exhibit opposite elution sequences of the optically active isomers, using the same adsorbent. Fig. 1 shows that the first fractions coming off the column exhibit the greatest values of the apparent molecular rotation, which gradually decreases. Middle fractions are inactive, and they are followed by fractions with increasing molecular rotation of opposite sign to that of the first fractions. Thus in the course of these experiments, only a partial resolution of the tris(acetylacetonato) complexes of cobalt(III) and chromium(III), has been achieved. A similar partial resolution of these complexes has also been effected by the use of
174
Fig. 1. Functional dependence of the apparent molecular rotation for tris(acetylacetonato)cobaIt (-- - -) on fraction number in the course of WI) (--- ) and tris(acetylacetonato)chromium(III) sepantion on a column packed with c~NOr)_rru~NH-)sg~-[Co(L-arg)~(NO,)rlCI.
adsorption columns of lactose6, alumina treated with (+)-tartaric acid’, o-sorbitol and D-mannitoP_ The achievement of only partial resolution cannot be due to racemfzation in the course of the separation, since we have established experimentally that these complexes do not racemize upon standing in benzene solution even for a timeperiod longer than that required for the elution in our resolution experiments. One reason for the incomplete resoluticn is the relatively low plate number of the system used due to the large grain size of the adsorbent. However, the use of adsorbents of smaller grain size was not possible since the elution was not performed under high pressure. Experiments aimed at increasing the efficiency of the column are now being made.
REFERENCES I Y. Yoshikawa and K. Yamasaki, Coord. Chem. Rev., 28 (1979) 205. 2 S.Ya mazaki, T. Yukimoto and H. Yoneda, J_ Chromotogr_. 175 (1979) 317. 3 4 5 6 7
R_ Tsuchida, M. Kobtiyashi and A. Nakamura, Bull. CIrem_ Sot. Jap., $1 (1936) 38. E. Ferroni and R_ Cini, I_ Amer. Chem_ Sot., 82 (1960) 2427. T. MoeiIer and E. Gulyas, J. Inorg. Nucl. Chem., 5 (1958) 245. J. P. Co&tan, R. P, BIair. R. L. Marshall and S. Slade, inorg. Chem., 2 (1963) 576. T. 4;. Piper, J. Amer. Chem. Sot., 83 (1961) 3908. 8 N. S. Bowman, V. GEeva and G. K. Schweitzer, inorg. Chem. Left., 2 (1966) 2.51.
H. Krebs and R. Rasche, Narurwisenschnfren, 41(1954) 63. M. E. &lap, D. J. Radanovi~ and T. J. Tanjie, inorg. Chem., 4 (1965) 1494. M. B. &lap, D. J. Radanovi~, T. I. NikoliC and T. J. JanjE, Inorg. Cfiim. Acta. 2 (1968) 52. B. A. ?Xamberi,M. B. &lap and T. J. Janj& Gkzs. Hem. Drti.. Beograd, 43 (1978) 149. M. B. &!a~, P. N. RadivojZa and G. Djmovi& in preparation. L. F. Auckfeth, (Editor), Inorg. Syn., 5 (1950) 188. 15 L. F. Audrieth (Editor), Borg. Sym, 5 (1950) 130.
9 10 11 12 13 14