In the Laboratory
Experiments with Tris(ethylenediamine)cobalt(III) Compounds:
59Co NMR and the Resolution of Enantiomeric [Co(en) ]3+ Ion
                                                     3
and Analysis by Formation of Diastereomeric Ion Pairs                                                                            W
L. L. Borer,* J. G. Russell, and R. E. Settlage
Department of Chemistry, California State University, Sacramento, Sacramento, CA 95819-6057; *borer@csus.edu
R. G. Bryant
Department of Chemistry, University of Virginia, Charlottesville, VA 22901
Background
      Cobalt(III) complexes are used in upper-division under-
graduate inorganic laboratory to demonstrate techniques in
coordination chemistry. The experiment described in this re-
port involves the synthesis and resolution (1) of enantiomeric
tris(ethylenediamine)cobalt(III) compounds and the analy-
sis of enantiomeric purity by optical rotation and 59Co NMR
spectroscopy through the formation of diastereomeric ion
pairs (2).
      Cobalt-59 has a nuclear spin number, I, of 7/2, an isotopic
abundance of 100%, and resonance frequency of 71 MHz
at 7.0 T, which makes 59Co NMR spectroscopy a practical and          Figure 1. 59Co NMR spectrum of 10 mM (+/᎑)-[Co(en)3]Cl3 solution.
rapid experiment. The only caveat stems from the nuclear
quadruple moment of 59Co, which requires the cobalt atom to
occupy a highly symmetric environment in order to yield line
widths of 100–200 Hz. The cobalt ion in the cobalt complex
[Co(en)3]3+ resides in an octahedrally symmetric environment
and meets the symmetry criterion for a relatively narrow 59Co
NMR line width. Although a line width of 100–200 Hz is
much larger than the 0.5–1.0 Hz of 1H and 13C NMR, the
cobalt NMR line width is not a problem at 7.0 T, as the
cobalt(III) chemical shift range is on the order of 15,000 ppm
(3). This extremely large chemical shift range implies an
extreme sensitivity of the chemical shift of a cobalt atom to
small perturbations in its environment. Indeed, the cobalt
chemical shift for the cobalt atom in [Co(en)3]3+ hydrogen–          Figure 2. 59Co NMR spectrum of 10 mM (+/᎑)-[Co(en)3]Cl3 solution
deuterium isotopomers show about 5 ppm shift for each                that is 50 mM is sodium d-tartrate.
hydrogen atom that is replaced by a deuterium atom (4 ).
      Iida et al. reported that racemic [Co(en)3]3+ ion pairs with
d-tartrate ion, resulting in diastereomeric ion pairs having
clearly separated 59Co NMR resonances that differ in chemical        M is (+)- or (᎑)-[Co(en)3]3+, T is d-tartrate ion, M*T is the
shift by about 5–6 ppm (2). Figure 1 shows the 59Co NMR              [Co(en)3]3+–d-tartrate ion paired species, δ is the chemical
spectrum of 10 mM (+/᎑)-[Co(en)3]3+ ion and Figure 2 shows           shift and δ obs the observed chemical shift, N is the ion or ion
the two 59Co NMR resonances that result from the addition of         pair mole fraction, and K is the ion pair association constant.
sodium d-tartrate to a concentration of 50 mM. The spectrum               The ion pairs between d-tartrate ion and (+)-[Co(en)3]3+
reflects the rapid equilibrium between the free ions and the         and (᎑)-[Co(en)3]3+ ions are diastereomeric and would be
ion paired species, which results in mole fraction weighted          expected to have different association constants, K(+) and K(᎑),
averaged chemical shift values, as described in eqs 1 and 2.         respectively, and 59Co chemical shifts, δ ip. The association
Equation 3 gives the ion pair association constant, K.               constants (5; Settlage, R. E.; Russell, J. G.; Bryant, R. G.
                        Mf + Tf = M*Tip                       (1)    unpublished results) are on the order of 20–60 and the ratio
                                                                     K(+)/K(᎑) is approximately equal to 1.1, indicating that ion pair
                      δobs = Nf δ f + Nip δ ip                (2)    formation between d-tartrate and (+)-[Co(en)3]3+ is favored
                                                                     slightly but at 0.1–1.0 M tartrate ion more than one tartrate
                                M*Tip
                         K=                                   (3)    associates with the [Co(en)3]3+ ion (2). This multiple associa-
                               Mf Tf                                 tion of [Co(en)3]3+ with the tartrate counter ion does not affect
                                                                     the application described here and we are not attempting to
where the subscripts f and ip denote free and ion-paired species,    calculate K or K (+)/K(᎑).
494       Journal of Chemical Education • Vol. 79 No. 4 April 2002 • JChemEd.chem.wisc.edu
                                                                                                               In the Laboratory
Figure 3. 59Co NMR spectrum of sample 5-d, 10 mM (+)-[Co(en)3]          Figure 4. 59Co NMR spectrum of sample 5-1,10 mM (᎑)-[Co(en)3]I3
d-tartrate Cl to which sufficient sodium d-tartrate was added to give   to which sufficient sodium d-tartrate was added to give a solution
a solution 50 mM in d-tartrate ion. The resonances from both            50 mM in d-tartrate ion. The spectrum shows the resonances from
diastereomeric ion pairs are seen, showing 92% (+)-enantiomer.          both diastereomeric ion pairs, with 88% (᎑)-enantiomer. Resolution
The resolution was incomplete.                                          was incomplete.
      The 59Co chemical shift of the d-tartrate–(+)-[Co(en)3]3+         at 71 MHz on stationary samples with outfield-frequency lock
ion pair is chemically shifted more from the 59Co chemical              using a Bruker AC-300 equipped with a 10-mm broadband
shift for reference of 10 mM (+/᎑)-[Co(en)3]Cl3 solution than           probe temperature-controlled at 298 K. The spectra were
is the 59Co chemical shift of the d-tartrate–(᎑)-[Co(en)3]3+            acquired by signal averaging 256 scans obtained with a 90°
ion pair (2). This is reflected in Figure 2, which shows a ᎑30          pulse, 80-ms pulse recycle time, 50 kHz spectral width, and 8 k
ppm shift on addition of sodium d-tartrate to 10 mM (+/᎑)-              data points zero-filled to 32 k points. The resulting spectra were
[Co(en)3]Cl3 solution (cf. Fig. 1). Of the two 59Co NMR reso-           processed on a PC using NUTS.1 The relative peak areas were
nances shown in Figure 2, the more chemically shifted one at            obtained by the NUTS line-fitting routine after fitting and
᎑30.85 ppm is that of the ion pair between d-tartrate ion               flattening the spectrum baseline. The 59Co chemical shifts are
and (+)-[Co(en)3]3+; the resonance at ᎑26.74 ppm is that of the         referenced to an external 10 mM (+/᎑)-[Co(en)3]Cl3 solution.
ion pair between d-tartrate ion and (᎑)-[Co(en)3]3+, in agreement             NMR spectrometers operating at a field strength of
with the reported assignment (2). This can be verified by com-          approximately 7.0 T and equipped with 5- or 10-mm broad
paring Figures 3 and 4, which show the 59Co NMR spectra                 band probes are common even at undergraduate institutions.
of incompletely resolved (+)-[Co(en)3]⭈d-tartrate⭈Cl⭈5H2O               The high field strength is required for resolution of the broad
                                                                        59
with sodium d-tartrate added and incompletely resolved                    Co NMR lines. A broad band NMR spectrometer based on
(᎑)-[Co(en)3]I3 with sodium d-tartrate added. The spectrum              an electromagnet will not have a high enough magnetic field
of the solution containing mostly (+)-[Co(en)3]3+ ion shows             strength to adequately resolve the broad 59Co NMR lines. To
a larger, more chemically shifted peak that can be assigned             accommodate those with electromagnet systems, we have made
to the 59Co resonance of the (+)-[Co(en)3]3+-containing species,        several cobalt spectra FID files available online.W
as the specific rotation, [α], is positive for the cobalt complex
solute. The 59Co NMR spectrum of the solution containing                Optical Rotation
mostly (᎑)-[Co(en)3]3+ ion can similarly be assigned. The larger,            The optical rotation of the resolved cobalt complexes
less chemically shifted peak is assigned to the (᎑)-[Co(en)3]3+-        were determined using a Rudolf Auto polarimeter, 1-dm
containing species.                                                     sample cells, and solutions containing approximately 0.025 g
      The percent enantiomeric purity of a tris(ethylenedi-             of cobalt compound per 1.0 mL of solution.
amine)cobalt(III) sample can be obtained from the d-tartrate
ion pair solutions using 59Co NMR peak areas, which are                 Hazards
obtained using a line-fitting routine. The observation of two
59
  Co NMR peaks provides a concrete corroboration of enan-                    Wear protective eyewear and gloves at all times. Keep
tiomeric purity calculated from optical rotation.                       flammable liquids away from open flames. All reactions should
                                                                        be carried out in a hood or well-ventilated area. Exercise
Experimental Procedure                                                  caution when handling ethylenediamine and the barium
                                                                        compounds. Barium sulfate and cobalt wastes must be put
Synthesis                                                               in separate receptacles in the hood.
    The cobalt compounds were synthesized and resolved
according to known procedures (1).                                      Results
NMR Spectra                                                                  The optical rotation results for several student samples
    The NMR samples were 10 mM in cobalt complex to                     of (+)-[Co(en)3]⭈d-tartrate⭈Cl⭈5H2O and (᎑)-[Co(en)3]I3 are
which sufficient sodium d-tartrate was added to give a solution         collected in Tables 1 and 2, respectively. The percent optical
50 mM in d-tartrate ion. The 59Co NMR spectra were acquired             purity, %op, and the percent enantiomer purity, %enant, were
                            JChemEd.chem.wisc.edu • Vol. 79 No. 4 April 2002 • Journal of Chemical Education                        495
 In the Laboratory
calculated from the experimental specific rotation, [α]exptl, using                 Table 1. Student Optical Rotation Data
eqs 4 and 5, respectively (6 ):                                                       for (+)-[Co(en)3]ⴢd-TartrateⴢClⴢ5H2O
                                                                         Sample               [α]exptl       Optical             (+)-Enantiomer
                              α   exptl                                   No.                  (deg)        Purity (%)              Purity (%)
                     %op =                × 100                (4)
                               α   lit
                                                                              1-d             +88              86                     93
                                                                              2-d            +101              99                    100
                                         100 – %op                            3-d            +103            100                     100
                 %enant = %op +                                (5)            4-d            +100              98                     99
                                             2
                                                                              5-d             +87              87                     93
where [α]lit is the literature value (1) of +102° for the specific            6-d             +92              90                     95
rotation for (+)-[Co(en)3]⭈d-tartrate⭈Cl⭈5H2O and ᎑90° for
(᎑)-[Co(en)3]I3.
     Typical 59Co NMR spectra after the addition of sodium
d-tartrate to a student sample making a solution 50 mM in
d-tartrate ion are shown in Figures 3 and 4. Figure 3, the spec-                    Table 2. Student Optical Rotation Data
trum of (+)-[Co(en)3]⭈d-tartrate⭈Cl⭈5H2O with added tartrate                                   for (–)-[Co(en)3] I3
ion, sample 1-d, shows the experimental spectrum and the                 Sample               [α]exptl       Optical             (–)-Enantiomer
two calculated peaks that were fitted to it. Similarly, Figure            No.                  (deg)        Purity (%)              Purity (%)
4 shows the experimental spectrum of (᎑)-[Co(en)3]I3 with                     1-l             ᎑90            100                     100
added tartrate ion, sample number 5-l, and the two calculated
                                                                              2-l             ᎑35              39                     69
peaks that were fitted to it. The NMR data and percent enan-
tiomeric purity for (+)-[Co(en)3]⭈d-tartrate⭈Cl⭈5H2O and (᎑)-                 3-l             ᎑70              78                     89
[Co(en)3]I3 with added tartrate ion are collected in Tables 3                 4-l             ᎑45              50                     75
and 4, respectively. The sample numbers of the NMR data                       5-l             ᎑71              79                     90
and results correspond to the sample numbers for the optical                  6-l             ᎑20              22                     61
rotation results collected in Tables 1 and 2.
Discussion
     A comparison of the percent enantiomeric purity in                         Table 3. Student 59Co NMR Data for
Tables 1 and 2 and Tables 3 and 4, calculated from the specific        (+)-[Co(en)3]ⴢd-TartrateⴢClⴢ5H2O with Na2[d-Tartrate]
rotation data and the 59Co NMR relative peak areas, are
                                                                                      (+)-[Co(en)3]3+         (–)-[Co(en)3]3+
within ± 2% of each other, in good agreement. The detection
                                                                      Sample                     Relative                Relative      (+)-Enanti-
limit of the 59Co NMR peak area is estimated to be 2–3%                No.          δ (ppm)       Peak      δ (ppm)       Peak          omer (%)
of the minor enantiomer for peaks with widths at half-height
                                                                                                  Area                    Area
in the range of 150–200 Hz and chemical shift differences
                                                                       1-d          ᎑28.79         1.00     ᎑24.79        0.08              93
of 4 ppm.
     For each set of diastereomeric ion samples there is a range       2-d          ᎑29.38         1.00        —           nil             100
of chemical shifts. For example, the d-tartrate ion pairs derived      3-d          ᎑29.10         1.00        —           nil             100
from (+)-[Co(en)3]⭈d-tartrate⭈Cl⭈5H2O show a shift of ᎑28.75           4-d          ᎑28.75         1.00        —           nil             100
to ᎑31.64 ppm on the addition of sodium d-tartrate, and                5-d          ᎑31.64         1.00     ᎑27.56        0.09              92
(᎑)-[Co(en)3]I3 samples have a range of ᎑20.35 to ᎑22.59 ppm.
                                                                       6-d          ᎑30.75         1.00     ᎑26.70        0.06              94
The range of shifts may be due to the solutions’ having
different pH values, which would affect the concentration of
the tartrate ion. However, it was not necessary to adjust the
pH of the sample solutions for this analysis.
     The ion-pair 59Co NMR spectra obtained from [Co(en)3]I3
                                                                      Table 4. Student 59Co NMR Data for (–)-[Co(en)3] I3 with
samples are not chemically shifted as much as those of the
                                                                                          Na2[d-Tartrate]
[Co(en)3]Cl3 reference or [Co(en)3]⭈d-tartrate⭈Cl⭈5H2O. The
chemical shift of the d-tartrate (+)–[Co(en)3]3+ ion pair in the                      (–)-[Co(en)3]3+         (+)-[Co(en)3]3+
[Co(en)3]Cl3 reference sample is ᎑30.85 ppm, while the minor          Sample                     Relative                Relative       (–)-Enanti-
amount of ion-paired (+)-[Co(en)3]3+ in the (᎑)-[Co(en)3]I3            No.          δ (ppm)       Peak      δ (ppm)       Peak           omer (%)
samples has chemical shifts ranging from ᎑24.21 to ᎑26.59                                         Area                    Area
ppm. The chloride and iodide ions in these samples interact             1-l         ᎑20.88         1.00        —            nil            100
to different degrees with the [Co(en)3]3+, and in the d-tartrate        2-l         ᎑20.35         1.00     ᎑24.21        0.46               68
ion-paired solutions they compete with the tartrate ion for the         3-l         ᎑21.58         1.00     ᎑25.47        0.13               88
[Co(en)3]3+, though not as effectively as the tartrate ion. The         4-l         ᎑20.96         1.00     ᎑24.87        0.30               77
59
  Co chemical shift of 6.7 ppm for 10 mM (᎑)-[Co(en)3]I3
                                                                        5-l         ᎑22.59         1.00     ᎑26.59        0.13               88
before addition of d-tartrate is evidence for the different degree
of interaction of chloride and iodide with [Co(en)3]3+.                 6-l         ᎑21.22         1.00     ᎑25.15        0.45               67
496       Journal of Chemical Education • Vol. 79 No. 4 April 2002 • JChemEd.chem.wisc.edu
                                                                                                          In the Laboratory
     These experiments using polarimetry and NMR spectros-        Note
copy are an excellent addition to experiments of synthesis and
                                                                     1. Acorn NMR Inc., 46560 Fremont Blvd #418, Fremont,
resolution of Co complexes in an upper-division chemistry
                                                                  CA 94538-6491; http://www.acornnmr.com/.
laboratory (7). Students can assess their laboratory techniques
by determining the degree of resolution achieved on the cobalt
                                                                  Literature Cited
samples. Students profit from analyzing data that have been
collected from more than one type of instrumentation. Critical     1. Broomhead, J. A.; Dwyer, F. P.; Hogarth, J. W. In Inorganic
thinking skills must be used for interpretation and integration       Synthesis; Rochow, E. G., Ed.; McGraw-Hill: New York, 1960;
of data. Students who have “less-than-perfect” data usually           Vol. VI, p 183.
gain the most understanding from this experiment.                  2. Iida, M.; Mieuno, Y.; Koine, N. Bull. Chem. Soc Jpn. 1995,
                                                                      68, 1337.
                                                                   3. Brevard, C.; Granger, P. Handbook of High Resolution Multi-
Acknowledgment                                                        nuclear NMR; Wiley: New York, 1981; p 124.
                                                                   4. Peterson, S. H.; Bryant, R. G.; Russell, J. G. Anal. Chim. Acta
    Partial support for the purchase of the Bruker AC-300
                                                                      1983, 154, 211.
NMR spectrometer was provided by the National Science
                                                                   5. Taur, T.; Nakazawa, H.; Youeda, H. Inorg. Nucl. Chem. Lett.
Foundation under grant USE-9051141.
                                                                      1977, 13, 3.
                                                                   6. Pavia, L.; Lampman, G. M.; Kriz, G. S. Introduction to Or-
W
    Supplemental Material                                             ganic Laboratory Techniques, 3rd ed.; Saunders: New York,
                                                                      1988; pp 648–654.
    Several cobalt spectra FID files and instructions for stu-     7. Girolami, G. S.; Rauchfuss, T. B.; Angelici, R. J. Synthesis and
dents on conducting the experiment and preparing the lab              Techniques in Inorganic Chemistry, 3rd ed.; University Science
report are available in this issue of JCE Online.                     Books: Sausalito, CA, 1999; pp 131–157.
                         JChemEd.chem.wisc.edu • Vol. 79 No. 4 April 2002 • Journal of Chemical Education                       497