Evaluation of diimine ligand exchange on Cu(I)

Article abstract of DOI:10.1021/ic0000606

Four ligands have been prepared, 8,8-dimethyl-6,7,9-trihydropyrido[1,2-b]acridine and three 4,4′,6,6′-tetrasubstituted derivatives of 2,2′-bipyrimidine where the substituents are methyl, phenyl, and p-tolyl. The corresponding [CuL₂]⁺ salts of these ligands evidence nonequivalent NMR signals that allow an estimation of the ligand exchange barrier in both acetonitrile and chloroform solution. Lower barriers are found in the former solvent and attributed to solvent participation in the exchange process. Corresponding differences in the oxidation potentials of the complexes are explained in a similar manner. The electronic absorption properties of the complexes are also consistent with the steric and electronic properties of the ligands. [Cu(2c)₂](PF₆), where 2c = 4,4′,6,6′-tetraphenyl-2,2′-bipyrimidine, was analyzed by X-ray diffraction and found to crystallize in the space group Pccn with a = 14.761(2) A, b = 15.007(2) A, c = 24.407(4) A, and Z = 4. The internal and external phenyl rings are disposed quite differently, with the internal rings interacting strongly with the orthogonal ligand.

Full text of DOI:10.1021/ic0000606

Inorg. Chem. 2001, 40, 2541-2546
2541
Evaluation of Diimine Ligand Exchange on Cu(I)
Elvira Riesgo, Yi-Zhen Hu, Fre´de´ric Bouvier, and Randolph P. Thummel*
Department of Chemistry, University of Houston, Houston, Texas 77204-5641
ReceiVed January 18, 2000
Four ligands have been prepared, 8,8-dimethyl-6,7,9-trihydropyrido[1,2-b]acridine and three 4,4′,6,6′-tetrasubstituted
derivatives of 2,2′-bipyrimidine where the substituents are methyl, phenyl, and p-tolyl. The corresponding [CuL2]+
salts of these ligands evidence nonequivalent NMR signals that allow an estimation of the ligand exchange barrier
in both acetonitrile and chloroform solution. Lower barriers are found in the former solvent and attributed to
solvent participation in the exchange process. Corresponding differences in the oxidation potentials of the complexes
are explained in a similar manner. The electronic absorption properties of the complexes are also consistent with
the steric and electronic properties of the ligands. [Cu(2c)2](PF6), where 2c ) 4,4′,6,6′-tetraphenyl-2,2′-bipyrimidine,
was analyzed by X-ray diffraction and found to crystallize in the space group Pccn with a ) 14.761(2) Å, b )
15.007(2) Å, c ) 24.407(4) Å, and Z ) 4. The internal and external phenyl rings are disposed quite differently,
with the internal rings interacting strongly with the orthogonal ligand.
Introduction
Albrecht-Gary and co-workers have recently explored the
importance of steric and electronic effects on the stability of
Over the past few decades chemists have become increasingly
concerned with the employment of metal-ligand interactions
in the construction of organized supramolecular assemblies.1
The Cu(I) ion lends itself particularly well to this task through
the formation of bis-diimine complexes with a variety of
judiciously designed bridging ligands. The advantage of these
systems stems from their preference for tetrahedral geometry,
allowing ligands to be oriented in orthogonal planes, and the
kinetic lability of the Cu(I) complexes which allows ligand-
metal exchange to occur readily in solution. Formation of the
thermodynamically favorable, and readily predictable, complex
often results.
mono- and bischelate Cu(I) complexes of 2,9-disubstituted
derivatives of 1,10-phenanthroline.7 Insights were gained pri-
marily through cyanide-assisted demetalation kinetic studies.
We have examined the Cu(I) complexes of a series of 2,2′-
biquinolines where the biquinoline could be replaced by the
stronger binding 2,9-dimethyl-1,10-phenanthroline.8 This ex-
change could be monitored spectrophotometrically, but, due to
complexities imposed by the involvement of partially exchanged
intermediates, these systems did not allow a quantitative
assessment of the ligand exchange barrier. This paper will
examine several ligand systems which do allow an estimate of
this exchange barrier, and the effects of structure and solvent
will be discussed.
An early example of molecular architecture involving Cu(I)
was the “molecular box” constructed by Osborn and Youinou.2
The bridging ligand, 3,6-di(2′-pyridyl)pyridazine, assembles
around four Cu(I) centers to afford the box, which is the simplest
arrangement of the ligand and metal that allows every binding
site on both species to be occupied without undue distortion.
Interestingly, stabilizing π-stacking interactions between the
ligands may also assist in the assembly process.3 Baxter, Lehn,
and co-workers have exploited this self-assembly theme using
a variety of polypyridine bridging ligands as well as an
assortment of metals. The construction of rods,4 ladders,5 and
larger grids6 has emphasized the importance of facile metal-
ligand exchange.
For [Cu(L)2]+ complexes, where L is an unsymmetrical
diimine-type ligand, stereoisomers are possible at the metal
center.9 This chirality at Cu(I) creates a situation where geminal
groups on the cyclohexeno ring of 1a become diastereotopic in
the corresponding complex [Cu(1a)2]+. Thus, one should be able
1
to observe independent H NMR resonances for such geminal
protons. In practice, the chemical shift difference between these
geminal protons is too small to allow clear resolution and
interpretation. If one introduces a pair of geminal methyl groups
as in 1b, however, their 1H resonances should be more readily
distinguishable, depending on the integrity of the complex.
Where ligand exchange is rapid on the NMR time scale, only
(1) Lehn, J.-M. Supramolecular Chemistry: Concepts and PerspectiVes,
VCH: 1995, Chapter 9.
(2) Youinou, M.-T.; Rahmouni, N.; Fischer, J.; Osborn, J. A. Angew.
Chem., Int. Ed. Engl. 1992, 31, 733.
(3) Riesgo, E. C.; Bouvier, F.; Thummel, R. P. Inorg. Chem., in press.
(4) (a) Wa¨rnmark, K.; Baxter, P. N. W.; Lehn, J.-M. Chem. Commun.
(Cambridge) 1998, 993. (b) Sleiman, H.; Baxter, P. N. W.; Lehn, J.-
M.; Airola, K.; Rissanen, K. Inorg. Chem. 1997, 36, 4734. (c) Sleiman,
H.; Baxter, P. N. W.; Lehn, J.-M.; Rissanen, K. J. Chem. Soc., Chem.
Commun. 1995, 715.
(5) (a) Baxter, P. N. W.; Hanan, G. S.; Lehn, J.-M. Chem. Commun.
(Cambridge) 1996, 2019. (b) Baxter, P.; Lehn, J.-M.; De Cian, A.;
Fischer, J. Angew. Chem., Int. Ed. Engl. 1993, 32, 69.
(6) (a) Baxter, P. N. W.; Lehn, J.-M.; Kneisel, B. O.; Fenske, D. Chem.
Commun. (Cambridge) 1997, 2231. (b) Baxter, P. N. W.; Lehn, J.-
M.; Kneisel, B. O.; Fenske, D. Angew. Chem., Int. Ed. Engl. 1997,
36, 1978. (c) Baxter, P. N. W.; Lehn, J.-M.; Rissanen, K. Chem.
Commun. (Cambridge) 1997, 1323. (d) Baxter, P.; Lehn, J.-M.;
Fischer, J.; Youinou, M.-T. Angew. Chem., Int. Ed. Engl. 1994, 33,
2284.
(7) Meyer, M.; Albrecht-Gary, A.-M.; Dietrich-Buchecker, C. O.; Sauvage,
J.-P. Inorg. Chem. 1999, 38, 2279.
(8) Jahng, Y.; Hazelrigg, J.; Kimball, D.; Riesgo, E.; Wu, F.; Thummel,
R. P. Inorg. Chem. 1997, 36, 5390.
(9) Riesgo, E. C.; Credi, A.; De Cola, L.; Thummel, R. P. Inorg. Chem.
1998, 37, 2145.
10.1021/ic0000606 CCC: $20.00 © 2001 American Chemical Society
Published on Web 04/20/2001

2542 Inorganic Chemistry, Vol. 40, No. 11, 2001
Riesgo et al.
one signal will be observed, while slow exchange should
evidence a separate peak for each methyl group.
A related series of ligands which shows symmetry differences
in the complexed and uncomplexed states are the 2,2′-bipyr-
imidines 2a-d. In the free ligand, substituents at the 4,4′- and
6,6′-positions will be equivalent, but when two ligands are bound
in a tetrahedral fashion to a single metal, one pair of these
substituents points in toward the metal and one points away,
making them nonequivalent and potentially distinguishable by
1H NMR.
Figure 1. Upfield region of the 300 MHz 1H NMR of [Cu(1b)2](ClO4)
in CD3CN (δ 1.94, H2O: δ 2.15).
Results and Discussion
Table 1. Ligand Exchange Data for Cu(I) Complexes Studied by
VT NMRa
The phenanthrolines 1a,b were prepared in a straightforward
fashion by the Friedla¨nder condensation of 8-amino-7-quinoline-
carbaldehyde (3)10 with either cyclohexanone or 3,3-dimethyl-
cyclohexanone. It is noteworthy that, in the latter reaction, only
the 8,8-dimethyl isomer is formed to the total exclusion of the
6,6-dimethyl species. To obtain the 6,6-dimethyl isomer would
have required that the enamine, initially formed between 3 and
4b, subsequently condense at the much more hindered 2-position
of 4b.
Tc
((2 K)
kc
∆Gq
complex
(s-1, (20%)
((0.2 kcal/mol)
CD3CN
[Cu(1b)2](ClO4)
[Cu(2b)2](PF6)
[Cu(2c)2](PF6)
[Cu(2d)2](PF6)
265b
248b
273c
288c
15
313
344
350
14
11.6
12.8
13.5
d
CDCl3
[Cu(1b)2](ClO4)
[Cu(2d)2](PF6)
283b
318b
318c
11
510
528
15.2
14.7
14.7
a Calculated at the coalescence temperature Tc according to ∆Gq )
4.58Tc[10.32 + log(Tc/kc)] cal/mol, where kc (ligand exchange rate) )
(π × ∆υ)/x2 and ∆υ is the peak separation in the absence of exchange.
b Based on coalescence of CH3 signals. c Based on coalesence of Ar-H
signals. d Peaks for 2b complex were too broad and there was
insufficient sample of 2c complex.
The parent 2,2′-bipyrimidine (2a) was commercially available;
however, it was suspected that complexation-induced differences
between the 4,4′- and 6,6′-protons might be difficult to discern.
Such differences should be more apparent for the substituted
derivatives 2b-d which were all prepared by a Ni(0)-promoted
coupling11 of the corresponding 2-chloropyrimidine. The tetra-
p-tolylbipyrimidine 2d, which has not previously been reported,
was prepared in three steps from diketone 5.
afford the corresponding [Cu(L)2]+ salt in good yield. These
complexes were readily characterized by their 1H NMR spectra.
For the complex of 1b, the aliphatic region of the NMR is
displayed in Figure 1 and clearly shows three well-resolved
methylene signals. At 3.14 ppm a triplet appears for H6. The
adjacent non-benzylic H7 shows an upshifted triplet at 1.59 ppm.
The H9 protons appear as a singlet at 2.46 ppm, shielded by
0.73 ppm due to proximity of the Cu(I) center.
The complexes of 1b and 2b-d were then analyzed at various
temperatures in acetonitrile-d3 and chloroform-d to determine
the coalescence temperature for their diastereotopic NMR
signals. For [Cu(1b)2](ClO4) in CD3CN at 25 °C, the geminal
methyl groups of 1b evidence a sharp singlet at 0.65 ppm. Upon
cooling to -40 °C, the ligand exchange becomes sufficiently
slow such that the two diastereotopic methyl groups become
clearly resolved (Figure 1). The methylene signals, however,
remain essentially unchanged. A similar effect is observed for
the 4,4′,6,6′-substituents on the bipyrimidine ligands in com-
plexes of 2b-d. The observed coalescence temperatures and
resulting exchange rate and free energy change data are collected
in Table 1. It is noteworthy that, for [Cu(2d)2](PF6), the ∆Gq
value is identical when either the methyl or aryl protons of the
p-tolyl group are monitored.
The ligands 1a,b and 2b-d were treated with either
[Cu(CH3CN)4](ClO4) or [Cu(CH3CN)4](PF6) in acetonitrile to
(10) Riesgo, E. C.; Jin, X.; Thummel, R. P. J. Org. Chem. 1996, 61, 3017.
(11) (a) Tiecco, M.; Testaferri, L.; Tingoli, M.; Chianelli, D.; Montanucci,
M. Synthesis 1984, 736. (b) Semmelhack, M. F.; Helquist, P. M.; Jones,
L. D. J. Am. Chem. Soc. 1971, 93, 5908.
The ligand exchange barriers in acetonitrile fall within the
fairly narrow range of 11.6-14.0 kcal/mol. Three interesting

Evaluation of Diimine Ligand Exchange on Cu(I)
Inorganic Chemistry, Vol. 40, No. 11, 2001 2543
and self-consistent effects are in evidence. First, we can correlate
binding strength with ligand basicity. With four nitrogens, 2,2′-
bipyrimidine (pKa ) 0.6) is considerably less basic than 1,10-
phenanthroline (pKa ) 5.2)12 and thus should form a weaker
coordinative bond, explaining its lower exchange barriers as
compared to 1b. Part of this difference may also be due to the
flexibility of bipyrimidine about the 2,2′-bond which should
facilitate its stepwise coordination and decoordination.
For the two systems which could be clearly measured in both
solvents, we observe that, in both cases, the barrier is 1.3 kcal/
mol higher in chloroform than in acetonitrile. This observation
is consistent with the fact that the latter is a stronger coordinating
solvent and thus can facilitate ligand exchange. It seems likely
that the exchange process will be assisted by solvent binding
to partially exchanged intermediate species. Frei and Geier have
similarly observed that ligand exchange for [Cu(biq)2]+ (biq )
2,2′-biquinoline) occurs more rapidly in acetonitrile than in
acetone or methanol.13
Finally, if we examine the series [Cu(2b-d)2]+ in acetonitrile,
the exchange barrier increases 1.2 kcal/mol in going from the
tetramethyl- to the tetraphenyl-substituted bipyrimidine. This
barrier increases an additional 0.7 kcal/mol when p-tolyl replaces
phenyl. If solvent coordination is important to the ligand
exchange process, as the size of the substituents on the 2,2′-
bipyrimidine is increased, these substituents will interfere to a
greater extent with solvent coordination and exchange will
become more difficult.
Figure 2. Drawing of the cation of [Cu(2c)2](PF6) with atomic
numbering scheme.
Table 2. Selected Bond Lengths, Bond Angles, and Dihedral
Angles for [Cu(2c)2](PF6)a
One can test the relative binding strengths of the ligands
examined in this study by a competitive exchange experiment.
In principle, more weakly binding ligands should be replaced
by more strongly binding ones and judicious use of this
knowledge could result in the construction of predictable self-
assembled arrays. The relative affinities of 1b and 2b for Cu(I)
can be estimated from the data in Table 1. Considering that the
exchange rate kc will increase as the temperature is raised, we
predict that ligand 2b will exchange much more rapidly than
1b and thus should be replaced by this species. This prediction
was borne out by an experiment in which 2 equiv of the free
ligand 1b was added to a CD3CN solution of [Cu(2b)2]+. The
NMR spectrum showed the complete disappearance of signals
for 1b, with the appearance of uncomplexed 2b and [Cu(1b)2]+.
Bond Lengths (Å)
Dihedral Angles (deg)
Cu-N1
Cu-N19
2.059(4)
2.058(4)
N1-C2-C2′-N1′
22.3(7)
24.5(7)
N3-C2-C2′-N3′
N19-C20-C20′-N19′ 21.8(7)
Bond Angles (deg)
N21-C20-C20′-N21′ 23.6(7)
N1-Cu-N1′
80.8(2)
N19-Cu-N19′ 81.7(2)
N21-C22-C25-C26
C23-C22-C25-C30
15.5(7)
19.9(8)
6.3(7)
N1-Cu-N19
N1-Cu-N19′
N1′-Cu-N19
N1′-Cu-N19′
125.64(16) N3-C4-C7-C8
124.70(15) C5-C4-C7-C12
124.70(15) N1-C6-C13-C14
125.64(16) C5-C6-C13-C18
C23-C24-C31-C32
4.4(8)
47.3(7)
48.3(7)
42.9(7)
44.2(7)
N19-C24-C31-C36
a Numbering pattern from Figure 2 with esd’s in parentheses.
Being aware that Cu(I) complexes are capable of considerable
distortion from ideal tetrahedral geometry,8 we became inter-
ested in the possibility of intramolecular π-stacking in the two
tetraaryl derivatives [Cu(2c,d)2]+. The dihedral angle between
these aryl groups and the parent pyrimidine will determine the
degree of conjugative interaction between these rings, possible
steric effects on coordination, and the existence of π-stacking
which may exert a stabilizing influence. Thus we undertook a
single-crystal X-ray analysis of [Cu(2c)2](PF6), and a drawing
of the cation is illustrated in Figure 2 and selected geometric
features are summarized in Table 2.
bonds angles are nearly equal at about 125.2°, indicating that
the two ligands occupy essentially perpendicular planes.
The bound ligand can evidence rotation about any of its three
nonequivalent C-C single bonds, and this rotation is reflected
by the dihedral angle between adjacent aryl rings. These dihedral
angles may be approximated by the average of two measure-
ments. For the central bipyrimidine, we average N1-C2-C2′-
N1′ and N3-C2-C2′-N3′ and find a twist angle of approxi-
mately 23.4°. This compares exactly with the angle between
the mean planes of the pyrimidines, which also measures 23.4°.
The “internal” and “external” phenyl rings are disposed quite
differently, and, especially for the external phenyls, the two
ligands are nonequivalent, possibly due to crystal packing forces.
One ligand shows an external phenylpyrimidine dihedral angle
of 17.4°, and the other shows only 5.4°. Both of these angles
are relatively small and indicate that conjugative effects encour-
age coplanarity of the rings. Figure 2 illustrates that, other than
interaction with other molecules in the lattice, the phenyl rings
are able to adopt the more favorable coplanar conformation.
The internal phenyls of each ligand interact strongly with
the orthogonal ligand and hence rotate about the phenyl-
pyrimidine bond to alleviate unfavorable repulsions. We had
The geometry around Cu(I) appears quite normal. The two
Cu-N bonds are equal in length and within the range normally
associated with such complexes.14 The N-Cu-N′ bite angles
average 81.3°, which is typical and the nonchelating N-Cu-N
(12) Bly, D. D.; Mellon, M. G. Anal. Chem. 1963, 35, 1386.
(13) (a) Frei, U. M.; Geier, G. Inorg. Chem. 1992, 31, 187. (b) Frei, U.
M.; Geier, G. Inorg. Chem. 1992, 31, 3132.
(14) (a) Healy, P. C.; Engelhardt, L. M.; Patrick, V. A.; White, A. H. J.
Chem. Soc., Dalton Trans. 1985, 2541. (b) Klemens, F. K.; Fanwick,
P. E.; Bibler, J. K.; McMillin, D. R. Inorg. Chem. 1989, 28, 3076. (c)
Geoffroy, M.; Wermeille, M.; Buchecker, C. O.; Sauvage, J.-P.;
Bernardinelli, G. Inorg. Chim. Acta 1990, 167, 157.

2544 Inorganic Chemistry, Vol. 40, No. 11, 2001
Riesgo et al.
Table 3. Long-Wavelength Absorption Maxima and Oxidation
Potentials for Cu(I) Complexes
E1/2 (ox)b
a
complex
λmax
CH3CN
CH2Cl2
[Cu(1a)2](ClO4) 458 (0.079)
[Cu(1b)2](ClO4) 455 (0.061)
0.44 (90) 0.65 (256)
0.42 (90) 0.60 (110)
0.83 (77) 0.97 (146)
[Cu(2b)2](PF6)
[Cu(2c)2](PF6)
[Cu(2d)2](PF6)
439 (0.045)
420 (0.064), 544 (0.035) 0.96 (94) 1.10 (82)
420 (0.063), 553 (0.040) 0.87 (83) 1.01 (106)
a 10-5 M CH2Cl2; absorptivities in parentheses. b Potentials are in
volts vs SCE; solutions were 0.1 M TBAP; T ) 25 ( 1 °C; the sweep
rate was 200 mV/s; and the number in parentheses is the difference
(mV) between the anodic and cathodic waves.
initially expected dihedral angles close to 90° which might lead
to π-stacking interactions with the orthogonal pyrimidine.
However, these dihedral angles measure 43.6° and 47.8°,
averaging near the compromise angle of 45°. It should further
be noted that these four phenyl groups quite effectively protect
the central copper ion from solvent attack.
An important characteristic of Cu(I) complexes is their
oxidative stability. Upon the loss of an electron, the tetrahedral
Cu(I) center, in going to Cu(II), becomes approximately square
planar and the ligands associated with the metal can have a
profound influence on this process. The half-wave oxidation
potentials for the complexes being studied are summarized in
Table 3 in both acetonitrile and dichloromethane solution. We
expect little effect from the substitution of dichloromethane for
chloroform as a noncoordinating solvent used in the NMR
measurements.
It has been well established for Cu(I) complexes involving
ligands related to 2,2′-bipyridine that substituents ortho to the
chelating nitrogens will stabilize the complex with respect to
oxidation by providing steric hindrance to the planarization
associated with this process.15 Interestingly, the 1,10-phenan-
throline ring in ligands 1a,b possesses only one such stabilizing
substituent, the 2,3-fused cyclohexeno ring. With the other side
of the ligand unencumbered, a favorable pathway exists for
planarization wherein the two phen rings rotate such that their
bulky substituents move apart and do not sterically interfere
with each other. The availablity of this oxidative pathway is
reflected by the low oxidation potential of +0.42-0.44 V, which
is consistent with the value of +0.37 V which we measured
for the analogous, and more bulky, 2,3-pinenyl-fused derivative.9
The bipyrimidine complexes of 2b-d have both their ortho
positions substituted and hence are considerably more difficult
to oxidize with potentials of +0.83-0.96 V. The fact that the
more bulky p-tolyl derivative oxidizes more readily than the
phenyl analogue is somewhat curious and could reflect the
slightly better electron-donating ability of the p-tolyl group.
There is a solvent effect on the oxidation of these Cu(I)
complexes which is consistent with the ligand exchange data
discussed earlier. For all the systems studied, oxidation in
dichloromethane occurs at higher potential than for aceto-
nitrile, and the waves are much less reversible. For the series
[Cu(2b-d)2]+, an increment of +0.14 V is observed for all three
complexes. This difference is attributed to the coordinating
ability of acetonitrile which could occupy the apical binding
site in the square pyramidal geometry normally associated with
the Cu(II) state. It should be noted that Miller and co-workers
have suggested that solvent does not bind to the Cu(II) form
Figure 3. Electronic absorption spectra of [Cu(2b)2](PF6) (‚‚‚),
[Cu(2c)2](PF6) (- - -), and [Cu(2d)2](PF6) (s); 10-5 M CH2Cl2.
when it is complexed to a highly congested ligand such as 2,9-
diphenyl-1,10-phenanthroline.16
These Cu(I) complexes exhibit a long-wavelength electronic
absorption attributed to a metal-to-ligand charge transfer
(MLCT) in which the metal is photooxidized in the excited state
and an electron is promoted to a vacant π*-orbital on one of
the ligands. The electronic absorption data is collected in Table
3. The MLCT transitions have been found to be subject to effects
dictated by ligand structure much in the same manner that the
redox properties are affected.17 The two phenanthroline systems
[Cu(1a,b)2]+ show MLCT absorptions at 455-458 nm, which
is in the range expected for such systems. Not surprisingly the
absorptivity for the unsubstituted derivative 1a is greater than
for the dimethyl derivative 1b, which would experience a weaker
ligand field. The tetramethyl bipyrimidine derivative [Cu(2b)2]+
exhibits a slightly higher energy absorption at 439 nm due to
the lower basicity of the ligand and its accordingly poorer ability
to stabilize the photooxidized state of the metal. Interestingly,
the corresponding Ru(II) complexes of 1a and 1,10-phenan-
throline show nearly identical absorption energies,18 arguing
against the importance of ligand basicity in these analogous d6
complexes.
A second, lower energy band appears at 544 and 553 nm,
respectively, for the complexes of the tetraaryl-substituted
derivatives 2c,d. Cunningham and co-workers have argued that,
for such complexes, the appearance of a pronounced shoulder
at about 550 nm suggests considerable distortion of the expected
D
2d symmetry.19 Such distortion in the complexes of 2c,d might
be attributed to intramolecular π-π interactions associated with
the aryl substituents. The analogous tetramethyl derivative
[Cu(2b)2]+ shows no long wavelength shoulder (Figure 3). The
solid state structure analysis of [Cu(2c)2]+ is, however, incon-
sistent with this explanation, evidencing relatively little distortion
from D2d symmetry. Crystal packing forces very likely override
the π-stacking effects which may be responsible for distortion
in solution. Excitation of CH2Cl2 solutions of the complexes
under study into their long-wavelength absorption band did not
produce any detectable luminescence at room temperature.
(16) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. Inorg. Chem. 1998,
37, 2285.
(17) (a) Miller, M. T.; Gantzel, P. K.; Karpishin, T. B. Inorg. Chem. 1999,
38, 3414. (b) Eggleston, M. K.; McMillin, D. R.; Koenig, K. S.;
Pallenberg, A. J. Inorg. Chem. 1997, 36, 172.
(18) Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.; von
Zelewsky, A. Coord. Chem. ReV. 1988, 84, 85.
(19) Cunningham, C. T.; Cunningham, K. L. H.; Michalec, J. F.; McMillin,
D. R. Inorg. Chem. 1999, 38, 4388.
(15) (a) Eggleston, M. K.; Fanwick, P. E.; Pallenberg, A. J.; McMillin, D.
R. Inorg. Chem. 1997, 36, 4007. (b) Pallenberg, A. J.; Koenig, K. S.;
Barnhart, D. M. Inorg. Chem. 1995, 34, 2833.

Evaluation of Diimine Ligand Exchange on Cu(I)
Inorganic Chemistry, Vol. 40, No. 11, 2001 2545
chromatographed on alumina (40 g), eluting with CH2Cl2/hexanes (1:
1), followed by EtOAc/hexanes (1:1) to obtain 1b (155 mg, 51%) as
white crystals, mp 144-145 °C: 1H NMR (CDCl3) δ 9.20 (d, 1H, J )
3.0 Hz, H9), 8.23 (d, 1H, J ) 6.9 Hz, H7), 7.95 (s, 1H, H4), 7.70 (AB
quartet, 2H, J ) 2.4, 9.0 Hz, H5/H6), 7.60 (quartet, 1H, J ) 3.6 Hz,
H8), 3.96 (broad s, H2O), 3.19 (broad s, 2H, -CH2-), 3.09 (t, 2H, J )
6.9 Hz, -CH2-), 1.73 (t, 2H, J ) 6.9 Hz, -CH2-), 1.09 (s, 6H, CH3);
13C NMR (CDCl3) δ 159.0, 150.0, 145.9, 144.2, 135.9, 135.3, 131.3,
128.2, 127.2, 126.2, 125.3, 122.4, 47.6, 35.4, 30.3, 28.2, 25.9. Anal.
Calcd for C18H18N2‚0.25H2O: C, 81.06; H, 6.94; N, 10.51. Found: C,
81.21; H, 6.30; N, 10.44. MS m/e 262 (M+).
In conclusion, we have shown that a judicious choice of
ligands can provide a means for estimating the barrier for ligand
exchange in Cu(I) diimine type complexes using VT NMR
techniques. For two relatively different types of ligands, 1,10-
phenanthroline and 2,2′-bipyrimidine, these values span a fairly
narrow range of 11.6-14.0 kcal/mol in acetonitrile. This
coordinating solvent facilitates ligand exchange as well as
oxidation of the complex. Bulky substituents which impede
solvent attack lead to higher exchange barriers. The resulting
better understanding of the ligand exchange process should assist
in the design of more efficient and specific self-assembling
ligand-metal complexes. Ideally one can hope for eventual
quantification of factors such as ligand basicity, steric bulk, and
conformation to allow structure-based predictions of self-
assembling tendencies in Cu(I)-based diimine ligand systems.
Further efforts are being directed along these lines.
4,6-Di(p-tolyl)-2-hydroxypyrimidine (6a). To urea (0.6 g, 9.5
mmol) dissolved in absolute EtOH (10 mL) were added 1,3-di(p-tolyl)-
1,3-propanedione (1.2 g, 4.8 mmol) and concentrated HCl (1.2 mL),
and the mixture was refluxed for 24 h. Additional urea (0.3 g, 4.7 mmol)
and concentrated HCl (0.3 mL) were then added and the reflux
continued for an additional 24 h. After cooling, the precipitate was
filtered, washed with CHCl3, triturated with hot acetone, and dried to
yield 6a (0.6 g, 46%) as a yellow powder, mp >270 °C: 1H NMR
(DMSO-d6) δ 8.07 (d, 4H, J ) 8.1 Hz, ArH), 7.60 (s, 1H, dCH-),
7.45 (d, 4H, J ) 8.1 Hz, ArH), 4.80 (s, 1H, -OH), 2.42 (s, 6H, -CH3).
4,6-Di(p-tolyl)-2-chloropyrimidine (6b). A mixture of 6a (410 mg,
1.48 mmol) and N,N-dimethylaniline (3 drops) in POCl3 (12 mL) was
refluxed for 7 h. Excess POCl3 was then removed by distillation under
vacuum. The resulting oil was dissolved in cold water and extracted
with CH2Cl2 (3 × 40 mL). The combined organic layers were dried
over MgSO4, concentrated, and chromatographed on alumina (25 g),
eluting with CH2Cl2, to afford 6b (100 mg, 23%) as a white solid, mp
142-145 °C: 1H NMR (CDCl3) δ 8.05 (d, 4H, J ) 8.1 Hz, ArH),
7.96 (s, 1H, dCH-), 7.33 (d, 4H, J ) 8.1 Hz, ArH), 2.44 (s, 6H,
-CH3).
4,4′,6,6′-Tetra(p-tolyl)-2,2′-bipyrimidine (2d). A mixture of an-
hydrous nickel chloride (89 mg, 0.68 mmol), triphenylphosphine (714
mg, 2.72 mmol), and zinc powder (63.5 mg, 0.97 mmol) was kept under
Ar for 10 min. Freshly distilled DMF (4 mL) was then added, and the
mixture was stirred at 50 °C for 30 min. To the dark red solution was
added 6b (200 mg, 0.68 mmol) in DMF (3 mL). Stirring was continued
for 6 h at 50 °C. After cooling, the reaction mixture was poured into
a mixture of NH4OH (28-30%, 14 mL) and water (36 mL), stirred
overnight, and extracted with CH2Cl2 (3 × 50 mL) in the presence of
EDTA disodium salt (3 g) to break the emulsion. The organic layer
was washed with water (30 mL), dried over MgSO4, and concentrated.
The resulting orange-yellow liquid was chromatographed on silica gel
(25 g) eluting with hexane/EtOAc (4:1). Recrystallization from EtOH
yielded 2d (55 mg, 31%) as a white powder, mp 271-273 °C: 1H
NMR (CDCl3) δ 8.25 (d, 8H, J ) 8.1 Hz, ArH), 8.18 (s, 2H, dCH-),
7.38 (d, 8H, J ) 7.8 Hz, ArH), 2.46 (s, 12H, -CH3); 13C NMR (CDCl3)
δ 165.8, 143.7, 141.2, 134.5, 129.6, 127.5, 111.7, 21.5. Anal. Calcd
for C36H30N4‚1.5H2O: C, 79.27; H, 6.05; N, 10.28. Found: C, 79.37;
H, 5.60; N, 10.15.
Experimental Section
Nuclear magnetic resonance spectra were recorded on a General
1
Electric QE-300 spectrometer at 300 MHz for H NMR and 75 MHz
for 13C NMR. Chemical shifts are reported in parts per million
downfield from Me4Si. Electronic spectra were obtained on a Perkin-
Elmer 330 spectrophotometer. Mass spectra were obtained on a Hewlett-
Packard 5989B mass spectrometer (59987A Electrospray) using the
atmospheric pressure ionization (API) method at a temperature of 160
°C for the complexes and atmospheric pressure chemical ionization
(APCI) at 300 °C for the ligands. Cyclic voltammograms were recorded
using a BAS CV-27 voltammograph or an EG & G Princeton Applied
Research Potientiostat/Galvanostat model 263A and a Houston Instru-
ments model 100 X-Y recorder according to a procedure which has
been described previously.20 All solvents were freshly distilled reagent
grade. Melting points were measured with a capillary melting point
apparatus and are not corrected. Elemental analyses were performed
by National Chemical Consulting, Inc., Tenafly, NJ. The 8-amino-7-
quinolinecarbaldehyde (3),10 4,4′,6,6′-tetramethyl-2,2′-bipyrimidine (2b),21
4,4′,6,6′-tetraphenyl-2,2′-bipyrimidine (2c),22 1,3-di(p-tolyl)-1,3-pro-
panedione (5),23 and [Cu(CH3CN)4]PF6 and [Cu(CH3CN)4]ClO424 were
prepared according to literature procedures. CAUTION! Perchlorate
salts of transition metal complexes containing organic ligands are
potentially explosive and should be prepared in small quantities and
handled with appropriate precautions. While no difficulties were
encountered with the complexes reported herein, due caution should
be exercised.
6,7,8,9-Tetrahydropyrido[1,2-b]acridine (1a). To a solution of
cyclohexanone (57 mg, 0.6 mmol) and 8-amino-7-quinolinecarb-
aldehyde (100 mg, 0.58 mmol) in absolute EtOH (5 mL) was added
saturated ethanolic KOH (0.5 mL). The solution was refluxed under
Ar for 17 h. The solvent was evaporated, and the residue was
chromatographed on alumina (30 g) eluting with CH2Cl2/hexane (7:3)
to give 1a (105 mg, 77%) as a beige solid, mp 116-117 °C: 1H NMR
(CDCl3) δ 9.32 (d, 1H, J ) 3.6 Hz, H9), 8.38 (d, 1H, J ) 7.8 Hz, H7),
8.01 (s, 1H, H4), 7.78 (broad s, 2H, H5/H6), 7.72 (quartet, 1H, J ) 4.5
Hz, H8), 3.58 (broad s, H2O), 3.51 (t, 2H, J ) 6.3 Hz, -CH2), 3.08 (t,
2H, J ) 5.7 Hz, -CH2-), 2.04 (m, 2H, -CH2-), 1.96 (m, 4H, -CH2-
); 13C NMR (CDCl3) δ 157.8, 144.4, 135.5, 133.5, 131.4, 127.8, 127.7,
126.8, 125.2, 125.0, 124.4, 124.4, 33.8, 29.2, 23.5, 23.2; MS m/e 235
(MH+).
[Cu(1a)2]ClO4. Solid [Cu(CH3CN)4]ClO4 (14 mg, 0.043 mmol) was
added to a solution of 1a (20 mg, 0.09 mmol) in CH3CN (7 mL). The
solution became dark red and was stirred at 25 °C under Ar for 1 h.
The solution was concentrated under vacuum and the complex
precipitated as air-stable, deep red crystals (22 mg, 81%): 1H NMR
(CD3CN) δ 8.82 (broad s, 2H, H9), 8.61 (d, 2H, J ) 8.1 Hz, H7), 8.33
(s, 2H, H4), 8.03 (broad s, 4H, H5/H6), 7.84 (quartet, 2H, J ) 4.8 Hz,
H8), 3.08 (t, 4H, J ) 6.0 Hz, -CH2-), 2.71 (broad s, 4H, -CH2-),
2.19 (s, H2O), 1.79 (broad s, 4H, -CH2-), 1.67 (broad s, 4H, -CH2-
); MS m/e 529 (M-ClO4)+. Anal. Calcd for C32H28N4CuClO4: C, 60.87;
H, 4.43; N, 8.88. Found: C, 60.84; H, 4.34; N, 8.70.
8,8-Dimethyl-6,7,9-trihydropyrido[1,2-b]acridine (1b). Following
the procedure described for 1a, 3,3′-dimethylcyclohexanone (147 mg,
1.16 mmol) was condensed with 8-amino-7-quinolinecarbaldehyde (200
mg, 1.16 mmol) in absolute EtOH (12 mL). The crude product was
[Cu(1b)2]ClO4. Following the procedure described for [Cu(1a)2]-
ClO4, 1b (50 mg, 0.2 mmol) was treated with [Cu(CH3CN)4]ClO4 (31
mg, 0.09 mmol) in CH3CN (7 mL) to give the complex as air-stable
deep red crystals (57 mg, 92%): 1H NMR (CD3CN) δ 8.80 (d, 2H, J
) 6.0 Hz, H9), 8.64 (d, 2H, J ) 6.0 Hz, H7), 8.40 (s, 2H, H4), 8.07
(broad s, 4H, H5/H6), 7.84 (quartet, 2H, H8), 3.14 (t, 4H, -CH2-),
2.46 (broad s, 4H, -CH2-), 1.59 (t, 4H, -CH2-), 0.65 (s, 12H, CH3).
Anal. Calcd for C36H36N4CuClO4: C, 62.90; H, 5.24; N, 8.15. Found:
C, 62.79; H, 5.29; N, 7.80.
(20) Goulle, V.; Thummel, R. P. Inorg. Chem. 1990, 29, 1767.
(21) Bly, D. D. J. Org. Chem. 1964, 29, 943.
(22) Nasielski, J.; Standaert, A.; Nasielski-Hinkens, R. Synth. Commun.
1991, 21, 901.
(23) Choschi, T.; Horimoto, S.; Wang, C. Y.; Nagase, H.; Ichikawa, M.;
Sugino, E.; Hibino, S. Chem. Pharm. Bull. 1992, 40, 1047.
(24) (a) Kubas, G. J.; Monzyk, B.; Crumbliss, A. L. Inorg. Synth. 1971,
19, 90. (b) Hemmerich, P.; Sigwart, C. Experientia 1963, 488.

2546 Inorganic Chemistry, Vol. 40, No. 11, 2001
Table 4. Crystallographic Data for [Cu(2c)2](PF6)
Riesgo et al.
information pertinent to data collection and refinement, are listed in
Table 4. The Laue symmetry was determined to be mmm, and from
the systematic absences noted the space group was shown unambigu-
ously to be Pccn. Intensities were measured using the ω-scan technique,
with the scan rate depending on the count obtained in rapid prescans
of each reflection. Two standard reflections were monitored after every
2 h or every 100 data collected, and these showed no significant
variation. During data reduction Lorentz and polarization corrections
were applied; however, no correction for absorption was made due to
the small absorption coefficient.
chemical formula C64H44CuF6N8P
fw
1133.58
Pccn
-50
0.71073
1.393
5.04
0.0436
0.1145
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
Z
14.761(2)
15.007(2)
24.407(4)
90(1)
space group
T (°C)
λ (Å)
F
calcd (g cm-3
)
90(1)
90(1)
5406.8(14)
4
µ (cm-1
R1a
)
wR2b
The structure was solved by the SHELXTL direct methods program,
which revealed the positions of most of the non-hydrogen atoms in
the molecule. Remaining atoms were located in subsequent difference
Fourier syntheses. The usual sequence of isotropic and anisotropic
refinement was followed, after which all of the hydrogens were entered
in ideal calculated positions and constrained to riding motion, with a
single variable isotropic temperature factor for all of them. After all
shift/esd ratios were less than 0.1, convergence was reached at the
agreement factors listed in Table 4. No unusually high correlations were
noted between any of the variables in the last cycle of full-matrix least-
squares refinement, and the final difference density map showed a
maximum peak of about 0.45 e/Å3. All calculations were made using
Bruker’s SHELXS-97 (Sheldrick, 1997) series of crystallographic
programs.25
a ∑||Fo| - |Fc||/∑|Fo|. b [∑w(Fo - Fc2)2/∑w(Fo )2]; w ) [(Fo ) +
2
2
2
(0.0566P)2 + (6.77P)]-1, where P ) (Fo + 2Fc2)/3.
2
[Cu(2b)2]PF6. To a solution of 2b (103 mg, 0.48 mmol) in CH3CN
(20 mL) was added [Cu(CH3CN)4]PF6 (89.2 mg, 0.24 mmol), and the
mixture was stirred for 1 h. The solvent was concentrated, and the
precipitate was dried overnight to provide [Cu(2b)2]PF6 as a red solid
(137.2 mg, 90%): 1H NMR (CD3CN) δ 7.47 (s, 4 H, dCH), 2.45 (s,
24 H, CH3). Anal. Calcd for C24H28N8CuPF6‚0.25H2O: C, 44.94; H,
4.44; N, 17.47. Found: C, 45.30; H, 4.43; N, 17.06.
[Cu(2c)2]PF6. A solution of 2c (52.1 mg, 0.113 mmol) in CH3CN
(15 mL) was gently warmed to ensure the complete dissolution of the
ligand. Solid [Cu(CH3CN)4]PF6 (21 mg, 0.056 mmol) was then added,
and the dark solution was stirred at room temperature for 1 h.
Concentration under vacuum gave [Cu(2c)2]PF6 as an air-stable dark
purple solid (67 mg, 93%): 1H NMR (CD3CN) δ 8.16 (bs, 20H, ArH
and dCH-), 7.42 (bs, 24H, ArH). Anal. Calcd for C64H44N8CuPF6‚
H2O: C, 66.77; H, 4.00; N, 9.73. Found: C, 66.60; H, 3.90; N, 10.12.
[Cu(2d)2]PF6. A solution of 2d (41.1 mg, 0.08 mmol) in CH3CN
(10 mL) and CH2Cl2 (5 mL) was gently warmed to ensure the complete
dissolution of the ligand. Solid [Cu(CH3CN)4]PF6 (14.74 mg, 0.04
mmol) was then added, and the dark purple solution was stirred at room
temperature for 2 h. Concentration under vacuum gave [Cu(2d)2]PF6
as an air-stable purple solid (45 mg, 91%): 1H NMR (CDCl3) δ 8.20
(bs, 8H, ArH), 7.70 (s, 4H, dCH-), 7.63 (bs, 8H, ArH), 7.37 (bs, 8H,
ArH), 6.75 (bs, 8H, ArH), 2.44 (bs, 12H, -CH3), 1.68 (bs, 12H, -CH3).
Insufficient sample was available for CHN analysis.
Acknowledgment. We would like to thank the Robert A.
Welch Foundation (E-621), the National Science Foundation
(CHE-9714998), and the Petroleum Research Fund of the
American Chemical Society for financial support of this work.
We would also like to thank Dr. James Korp for assistance with
the X-ray determination, Julien Wellhoff for assistance with the
electrochemical measurements, and Dr. Debby Roberts for
assistance with the mass spectrometry.
Supporting Information Available: Tables of atomic coordinates,
bond lengths, bond angles, anisotropic displacement parameters, and
H-atom coordinates for [Cu(2c)2](PF6); VT NMR data used to deter-
emine the kinetic parameters outlined in Table 1. This material is
available free of charge via the Internet at http://pubs.acs.org.
Crystal Structure Determination of [Cu(2c)2](PF6). A red-orange
thick plate having approximate dimensions 0.15 × 0.25 × 0.35 mm
was mounted in a random orientation on a Nicolet R3m/V automatic
diffractometer. The sample was placed in a stream of dry nitrogen gas
at -50 °C, and the radiation used was Mo KR monochromatized by a
highly ordered graphite crystal. Final cell constants, as well as other
IC0000606
(25) Sheldrick, G. M. In Crystallographic Computing 3; Sheldrick, G. M.,
Kruger, C., Goddard, R., Eds.; Oxford University Press: Oxford, U.K.,
1985; pp 175-189.