A novel method for synthesizing crystalline copper carbodiimide, CuNCN. Structure determination by X-ray rietveld refinement

Article abstract of DOI:10.1515/znb-2005-0601

Well-crystallized copper carbodiimide, CuNCN, was synthesized by the slow oxidation of a copper(I) cyanamide precursor under aqueous conditions. The X-ray powder data evidence the orthorhombic system and space group Cmcm with a = 2.9921(1), b = 6.1782(1), c = 9.4003(2) ?, V = 173.769(5) ?³ and Z = 4. There is a strongly distorted octahedral Cu ²⁺ coordination reflecting a typical first-order Jahn-Teller effect, with interatomic distances of 4 x Cu-N = 2.001(2) ? and 2 x Cu-N = 2.613(3) ?; the NCN^2- unit adopts the carbodiimide shape with C-N = 1.227(4) ?. Despite the formal d⁹ electron count of Cu²⁺, CuNCN exhibits a small temperature-independent paramagnetism and is likely to be a metallic conductor.

Full text of DOI:10.1515/znb-2005-0601

A Novel Method for Synthesizing Crystalline Copper Carbodiimide,
CuNCN. Structure Determination by X-Ray Rietveld Refinement
Xiaohui Liu, Martial Aime Wankeu, Heiko Lueken, and Richard Dronskowski
Institut fu¨r Anorganische Chemie der RWTH Aachen, Landoltweg 1, D-52056 Aachen, Germany
Reprint requests to Prof. Dr. R. Dronskowski. E-mail: drons@HAL9000.ac.rwth-aachen.de
Z. Naturforsch. 60b, 593 – 596 (2005); received March 18, 2005
Well-crystallized copper carbodiimide, CuNCN, was synthesized by the slow oxidation of a
copper(I) cyanamide precursor under aqueous conditions. The X-ray powder data evidence the or-
˚
thorhombic system and space group Cmcm with a = 2.9921(1), b = 6.1782(1), c = 9.4003(2) A,
3
2+
˚
V = 173.769(5) A and Z = 4. There is a strongly distorted octahedral Cu coordination reflect-
˚
ing a typical first-order Jahn-Teller effect, with interatomic distances of 4 ×Cu–N = 2.001(2) A and
2−
˚
˚
2 × Cu–N = 2.613(3) A; the NCN unit adopts the carbodiimide shape with C–N = 1.227(4) A.
Despite the formal d9 electron count of Cu2+, CuNCN exhibits a small temperature-independent
paramagnetism and is likely to be a metallic conductor.
Key words: Copper, Cyanamide, Carbodiimide, Rietveld Refinement, Jahn-Teller Distortion
Introduction
lead to the metal, metal nitride, C and gaseous N2 but
not to “MNCN” phases. To look at this synthetic prob-
lem from an alternative point of view, we have recently
finished an extensive quantum-chemical study of the
complete set of “MNCN” (M = Mn, Fe, Co, Ni, Cu) hy-
pothetical phases [24] from density-functional theory
(GGA), and this approach rendered all quasi-binary
compounds as unstable in terms of both formation en-
thalpy ∆Hf and Gibbs formation energy ∆Gf. Using
that information, we have recently accomplished the
successful synthesis of the first magnetic carbodiimide,
MnNCN, using a metathesis route involving zinc car-
bodiimide and manganese(II) chloride as starting ma-
terials. The synthetic idea lies in the low melting and
also sublimation temperature of the product ZnCl2 be-
cause the latter can be removed from the equilibrium
directly upon formation, thereby enforcing MnNCN
formation [25].
The design and synthesis of new classes of mag-
netic materials starting from the combination of para-
magnetic transition-metal centers with diamagnetic,
bidentate bridging ligands to generate extended net-
works is a very attractive goal for solid-state and
coordination chemists. As small nitrogen-containing
ligands, the basic anions (NCNH−, NCN2−) of the
cyanamide (H2NCN) molecule have been successfully
used to prepare a large number of salts comprising al-
kali metals [1 – 3], alkaline-earth metals [4 – 7], main-
group elements [8 – 11], d10 transition metals [12– 18],
and also rare-earth metals [19– 21] via widely dif-
fering synthetic routes. Nonetheless, the synthesis of
cyanamides or carbodiimides of the genuine magnetic
transition metals with a non-d10 electron count ap-
peared as a great challenge, even more so because the
structure elucidation and further characterization of the
products was expected to broaden our basic knowledge
of such a novel class of materials.
Because CuNCN has been predicted to be more un-
stable than MnNCN, even softer chemical routes had
to be found for a successful synthesis. As mentioned
According to our experience, transition-metal M2+ before, Cu2+ cations merely form coordination com-
ions (d4 − d9) prefer to form metal-cyanamide com- plexes such as [Cu(NCNH2)4]2+ from which hydro-
plexes such as [M(NCNH2)4]2+ ions in aqueous solu- gen atoms can not be abstracted; the alternative re-
tion, and it is very difficult to further remove the hy- action of an aqueous copper(II) solution with anionic
drogen atoms from the complex ions in order to yield cyanamide units yields a black material whose amor-
“MNCN” compounds [22, 23]. Other plausible high- phous nature renders its structure determination im-
temperature routes are also unsuitable because they possible [26, 27]. Fortunately, we have recently pre-
c
0932–0776 / 05 / 0600–0593 $ 06.00 ꢀ 2005 Verlag der Zeitschrift fu¨r Naturforschung, Tu¨bingen · http://znaturforsch.com
Unauthenticated
Download Date | 11/10/17 2:29 PM

594
X. Liu et al. · A Novel Method for Synthesizing Crystalline Copper Carbodiimide, CuNCN
Fig. 1. Observed, calculated, and differ-
ence intensities for the X-ray Rietveld re-
finement of CuNCN. The vertical mark-
ers show the positions of the Bragg reflec-
tions.
Table 1. Crystallographic data and structure refinement for
CuNCN.
pared a novel ammine copper(I) cyanamide com-
pound, Cu4(NCN)2NH3 [18], which appeared as a
lucrative oxidizable substrate for the target phase
CuNCN. Indeed, a slightly modified synthetic route
for Cu4(NCN)2NH3 using a large excess of cyanamide
within an ammonia solution leads to perfectly crys-
talline CuNCN.
Formula; molar mass:
Crystal color and form:
Lattice parameters:
CuNCN; 103.55 g/mol
black powder
˚
a = 2.9921(1) A
˚
b = 6.1782(1) A
˚
c = 9.4003(2) A
173.769(5) A
3
˚
Cell volume:
Space group; formula units:
Number of reflections:
Cmcm (no. 63); 4
98
Rp, Rwp, RBragg
:
0.054, 0.079, 0.045
Experimental Section
Table 2. Positional and isotropic displacement parameters for
CuNCN with standard deviations in parentheses.
A clear dark-green aqueous solution (5 ml) of CuCl2 ·
2H2O (1.70 g, 10 mmol) and H2NCN (0.82 g, 20 mmol)
was prepared in a round-bottom flask, and 5 ml of a solution
of Na2SO3 (1.26 g; 10 mmol) was added with stirring; the
solution became light-green. Then, a concentrated ammonia
solution (25 wt.%) was used to attain a pH around 9 where-
upon pale white Cu4(NCN)2NH3 precipitated. Upon stirring
the suspension over night, Cu4(NCN)2NH3 was slowly oxi-
dized at room temperature by atmospheric oxygen to yield
the black CuNCN. It was separated by filtration, washed
with water, and dried in vacuo. The microcrystalline CuNCN
was very well suited for powder X-ray diffraction. Yield:
0.84 g (80% with respect to CuCl2 · 2H2O). Analytical data
for CuNCN (103.55): calcd. Cu 61.37, C 11.60, H 0, N 27.05;
found Cu 61.30, C 11.62, H 0.14, N 27.00. IR (KBr, cm−1):
νas(NCN) = 2035 (s), δ(NCN) = 595 (s) and 690 (m).
2
˚
Atom
Wyckoff site
x
y
z
Biso (A )
Cu
N
C
4a
8b
4c
0
0
0
0
0
1.35(2)
0.77(7)
0.66(11)
0.3826(4)
0.3889(8)
0.3815(3)
1/4
in 2θ with individual steps of 0.01◦. The X-ray powder dia-
gram was indexed using an orthorhombic unit cell with space
group Cmcm. The primary structural model (atomic posi-
tions) came from a theoretical prediction [24]. The back-
ground of the data set was manually subtracted by linear in-
terpolation, and the FULLPROF program package [28] was
used for the Rietveld refinement with a pseudo-Voigt pro-
file function. Using isotropic displacement parameters for all
atoms, the residual values arrive at 0.054 (Rp), 0.079 (Rwp),
The X-ray diffraction data of CuNCN were recorded at and 0.045 (RBragg) for a total of 15 variables and 98 Bragg
room temperature by means of a calibrated STOE STADI- reflections. Tables 1 and 2 contain all important crystallo-
P powder diffractometer with strictly monochromatized Cu- graphic data. Additional structural details may be obtained
Kα1 radiation and a linear position-sensitive detector (PSD) from FIZ Karlsruhe, D-76344 Eggenstein-Leopoldshafen, on
with a flat sample holder; because of the excellent crys- quoting the depository number CSD-415 220. The course of
tallinity, the range of measurement was a very wide 8 – 130◦ the final Rietveld refinement is shown in Fig. 1.
Unauthenticated
Download Date | 11/10/17 2:29 PM

X. Liu et al. · A Novel Method for Synthesizing Crystalline Copper Carbodiimide, CuNCN
595
Fig. 3. Top view of a corrugated layer of Cu2+ and NCN2−
ions with a square planar Cu2+ coordination by nitrogen
atoms.
two nitrogen atoms are generated by only one crys-
tallographic N site; the anionic shape is very close
to D∞h (carbodiimide) symmetry. In accord with
that, the infrared spectrum of CuNCN exhibits only
a strong carbodiimide-type asymmetrical vibration,
νas(NCN) = 2035 cm−1, plus a strong deformation vi-
bration, δ(NCN) = 595/690 cm−1. In contrast, no sym-
metrical νs band (around 1200 cm−1) is observed be-
cause such a breathing mode is IR-forbidden for the
[N=C=N]2− carbodiimide unit although it would be al-
lowed for a cyanamide [N–C≡N]2− anion.
Fig. 2. View of the CuNCN structure along the a axis
with Cu atoms in black, N in white and C in grey. The
◦
˚
bond lengths (A) and angles ( ) are: Cu–Cu = 2.992(1) and
3.432(1), 4 × Cu–N = 2.001(2), 2 × Cu–N = 2.613(3),
C–N = 1.227(4), N–C–N = 176.4(5), N–Cu–N = 180,
N–Cu–N = 95.2(1), N–Cu–N = 84.8(1), N–Cu–N = 96.8(1),
N–Cu–N = 83.2(1), and C–N–Cu = 117.0(3).
Results and Discussion
The title compound has been prepared employing
the procedure presented in eq. (1). Black crystals were
obtained in 80% yield and their structure determined
by powder X-ray diffraction and Rietveld refinement.
In previous contributions [11, 25] we have discussed
the influence of the cationic hardness on the shape
of the NCN2− anion. While chemically hard cations
(e.g., Ca2+) favor ionic bonding to the symmetrical
[N=C=N]2− anion, softer cations (e.g., Pb2+ or Ag+)
bind covalently to one side of the NCN2− unit and
induce the less symmetrical [N–C≡N]2− cyanamide
shape, a finding easily rationalized by Pearson’s HSAB
concept [30]. Just like for MnNCN, however, the soft
Cu2+ ion in CuNCN induces the carbodiimide shape,
the second exception from the above shape rule. Fur-
ther research is needed to clarify whether or not the
H NCN,NH
O
2
2
3
4CuCl2 −−−−−−−→ Cu4(NCN)2NH3 −→ 4Cu(NCN)
Na SO [18]
2
3
(1)
The crystal structure of CuNCN (Fig. 2) may be con-
sidered a layer structure in which each corrugated
layer consists of alternating Cu2+ and NCN2− ions,
somewhat similar to the situation in MnNCN. In an
alternative description, the Cu2+ ions form cationic
sheets parallel to the ab face with Cu–Cu distances
presence of spin polarization (high-spin d 5 for Mn2+
)
or metallicity (d9 for Cu2+, see below) has a decisive
influence on the shape of the NCN2− unit.
2+
˚
of 2.992(1) and 3.432(1) A. Each Cu ion is coor-
dinated by six nitrogen atoms in the spirit of a typical
˚
The magnetic susceptibility of CuNCN has been de-
first-order Jahn-Teller effect (4×Cu–N = 2.001(2) A, termined by SQUID magnetometry (MPMS-5S, Quan-
˚
2 × Cu–N = 2.613(3) A). Methods to computation- tum Design, San Diego, USA) in the temperature range
ally model the Jahn-Teller effect found for Cu(II) between 100 and 300 K at an applied field of B0 =
compounds (tetragonally distorted octahedral coordi- 0.5 T. Allowing for corrections by diamagnetic incre-
nation vs. square planar coordination) have been sug- ments of Cu2+ and NCN2− (−40×10−11 m3 mol−1 in
gested [29]. The NCN2− units (Fig. 3) run paral- SI units), a small temperature-independent paramag-
lel with the c-axis, and the complex anion is prac- netic contribution of 20(3) × 10−11 m3 mol−1 results
tically linear (N–C–N = 176.4(5)◦). The C–N bond which is on the order of the magnitude of temperature-
˚
length arrives at 1.227(4) A, and because the cen- independent paramagnetism (TIP) and spin paramag-
tral C atom coincides with the two-fold axis, the netism of the conduction electrons. The black appear-
Unauthenticated
Download Date | 11/10/17 2:29 PM

596
X. Liu et al. · A Novel Method for Synthesizing Crystalline Copper Carbodiimide, CuNCN
ance of the material also points towards metallicity Acknowledgements
which certainly comes as a surprise. Presently, the
It is a pleasure to thank L. Stork for analytical help. The
financial support by the Fonds der Chemischen Industrie is
gratefully acknowledged.
magnetic behavior below 100 K and the field depen-
dence of the magnetic susceptibility as well as the elec-
tronic structure by first-principles methods are under
thorough investigation.
[16] M. Becker, M. Jansen, Z. Anorg. Allg. Chem. 626,
1639 (2000).
[17] X. Liu, P. Mu¨ller, P. Kroll, R. Dronskowski, Inorg.
Chem. 41, 4259 (2002).
[18] X. Liu, P. Mu¨ller, R. Dronskowski, Z. Anorg. Allg.
Chem. 631, 1071 (2005).
[19] O. Reckeweg, F. J. DiSalvo, Z. Anorg. Allg. Chem.
629, 177 (2003).
[1] M. G. Down, M. J. Haley, P. Hubberstey, R. J. Pul-
ham, A. E. Thunder, J. Chem. Soc. Dalton Trans. 1407
(1978).
[2] A. Harper, P. Hubberstey, J. Chem. Res. (S) 7, 194
(1989).
[3] M. Becker, J. Nuss, M. Jansen, Z. Anorg. Allg. Chem.
626, 2505 (2000).
[4] N. G. Vannerberg, Acta Chem. Scand. 16, 2263 (1962).
[5] U. Berger, W. Schnick, J. Alloys Compd. 206, 179
(1994).
[20] R. Srinivasan, M. Stro¨bele, H.-J. Meyer, Inorg. Chem.
42, 3406 (2003).
[21] W. Liao, C. Hu, R. K. Kremer, R. Dronskowski, Inorg.
Chem. 43, 5884 (2004).
[6] O. Reckeweg, F. J. DiSalvo, Angew. Chem. Int. Ed. 39,
412 (2000).
[22] X. Liu, P. Kroll, R. Dronskowski, Z. Anorg. Allg.
Chem. 627, 1682 (2001).
[7] W. Liao, R. Dronskowski, Acta Crystallogr. E60, 124
(2004).
[23] X. Liu, R. Dronskowski, Z. Kristallogr. NCS 217, 118
(2002).
[24] M. Launay, R. Dronskowski, Z. Naturforsch. 60b, 437
(2005).
[8] R. Dronskowski, Z. Naturforsch. 50b, 1245 (1995).
[9] R. Riedel, A. Greiner, G. Miehe, W. Dreßler, H. Fueß,
J. Bill, F. Aldinger, Angew. Chem. Int. Ed. Engl. 36,
603 (1997).
[25] X. Liu, M. Krott, P. Mu¨ller, C. Hu, H. Lueken, R. Dron-
skowski, Inorg. Chem. 44, 3001 (2005).
[26] M. J. Cooper, Acta Crystallogr. 17, 1449 (1964).
[27] O. Reckeweg, A. Simon, Z. Naturforsch. 58b, 1097
(2003).
[10] M. J. Cooper, Acta Crystallogr. 17, 1452 (1964).
[11] X. Liu, A. Decker, D. Schmitz, R. Dronskowski, Z. An-
org. Allg. Chem. 626, 103 (2000).
[12] F. P. Bowdem, H. M. Montagu-Pollock, Nature 191, 556
(1961).
[28] J. Rodriguez-Carvajal, Fullprof2000, Version 1.9b,
Laboratoire Le´on Brillouin (2000).
[13] M. Becker, J. Nuss, M. Jansen, Z. Naturforsch. 55b, 383
(2000).
[29] P. Comba, M. Zimmer, Inorg. Chem. 33, 5368 (1994).
[30] R. G. Pearson, Chemical Hardness, Wiley-VCH, Wein-
heim, New York (1997).
[14] M. Becker, M. Jansen, Acta Crystallogr. C57, 347
(2001).
[15] G. Baldinozzi, B. Malinowska, M. Rakib, G. Durand,
J. Mater. Chem. 12, 268 (2002).
Unauthenticated
Download Date | 11/10/17 2:29 PM