Supporting information
Crystallographic Information File (CIF) https://doi.org/10.1107/S0108270103001264/bc1004sup1.cif | |
Structure factor file (CIF format) https://doi.org/10.1107/S0108270103001264/bc1004Isup2.hkl | |
Portable Document Format (PDF) file https://doi.org/10.1107/S0108270103001264/bc1004sup3.pdf |
Tetrapropylammonium bromide (Aldrich, 532 mg, 2 mmol) was dissolved in absolute ethanol (10 ml). Potassium dicyanoargentate(I) (Aldrich, 1194 mg, 6 mmol) was dissolved in water (15 ml). These two solutions were combined and layered on top of an aqueous solution (10 ml) of manganese(II) nitrate hydrate (Aldrich, 358 mg, 2 mmol). Colorless hexagonal plates of KMnAg_{3}(CN)_{6} formed within 1 d.
The crystal structure was initially solved by direct methods in space group P3. After inspection with the routine MISSYM (Le Page, 1987, 1988) in the program PLATON (Spek, 1990), we added a vertical mirror plane and concomitant twofold axis to arrive at space group P31m. Refinement in this space group (with half-occupied disordered K atoms) converged at R_{1} = 0.053, wR_{2} = 0.218, with peaks on the difference map of +2 (near Ag) and −3 e Å^{−3} (near C). Subsequently, the inversion center was removed and only one of the now inequivalent K positions retained, leading to the final refinement in space group P312. We also explored the effect of exchanging the C and N atoms in space group P312. This led to convergence at R_{1} = 0.028, wR_{2} = 0.088, peaks of +1 (near C) and −1 e Å^{−3} (near N), and highly dissimilar atomic displacement parameters for the C and N atoms, so the assignment listed in the tables here was accepted as correct. The chirality of the crystal resulted in a Flack parameter (Flack, 1983) of 0.18 (4), which was interpreted and included in the refinement as racemic twinning of the same volume fraction. The 234 Friedel pairs were not merged in the final refinement.
Data collection: SMART (Siemens, 1995); cell refinement: SAINT (Bruker, 2001); data reduction: SAINT; program(s) used to solve structure: SHELXTL (Sheldrick, 2001); program(s) used to refine structure: SHELXTL; molecular graphics: SHELXTL; software used to prepare material for publication: SHELXTL.
C_{6}Ag_{3}KMnN_{6} | D_{x} = 2.819 Mg m^{−}^{3} |
M_{r} = 573.77 | Mo Kα radiation, λ = 0.71073 Å |
Trigonal, P312 | Cell parameters from 772 reflections |
Hall symbol: P 3 2" | θ = 3.4–28.2° |
a = 6.9219 (10) Å | µ = 5.48 mm^{−}^{1} |
c = 8.1465 (16) Å | T = 293 K |
V = 338.03 (10) Å^{3} | Hexagonal plate, colorless |
Z = 1 | 0.46 × 0.32 × 0.03 mm |
F(000) = 263 |
Siemens SMART CCD area-detector diffractometer | 558 independent reflections |
Radiation source: fine-focus sealed tube | 551 reflections with I > 2σ(I) |
Graphite monochromator | R_{int} = 0.025 |
area–detector ω scans | θ_{max} = 28.2°, θ_{min} = 2.5° |
Absorption correction: integration (XPREP in SHELXTL; Sheldrick, 2001) | h = −8→9 |
T_{min} = 0.155, T_{max} = 0.829 | k = −8→9 |
3417 measured reflections | l = −10→10 |
Refinement on F^{2} | Secondary atom site location: difference Fourier map |
Least-squares matrix: full | w = 1/[σ^{2}(F_{o}^{2}) + (0.0183P)^{2} + 0.008P] where P = (F_{o}^{2} + 2F_{c}^{2})/3 |
R[F^{2} > 2σ(F^{2})] = 0.012 | (Δ/σ)_{max} < 0.001 |
wR(F^{2}) = 0.032 | Δρ_{max} = 0.25 e Å^{−}^{3} |
S = 1.11 | Δρ_{min} = −0.35 e Å^{−}^{3} |
558 reflections | Extinction correction: SHELXTL (Sheldrick, 2001), Fc^{*}=kFc[1+0.001xFc^{2}λ^{3}/sin(2θ)]^{-1/4} |
30 parameters | Extinction coefficient: 0.058 (2) |
0 restraints | Absolute structure: Flack (1983) |
Primary atom site location: structure-invariant direct methods | Absolute structure parameter: 0.18 (4) |
C_{6}Ag_{3}KMnN_{6} | Z = 1 |
M_{r} = 573.77 | Mo Kα radiation |
Trigonal, P312 | µ = 5.48 mm^{−}^{1} |
a = 6.9219 (10) Å | T = 293 K |
c = 8.1465 (16) Å | 0.46 × 0.32 × 0.03 mm |
V = 338.03 (10) Å^{3} |
Siemens SMART CCD area-detector diffractometer | 558 independent reflections |
Absorption correction: integration (XPREP in SHELXTL; Sheldrick, 2001) | 551 reflections with I > 2σ(I) |
T_{min} = 0.155, T_{max} = 0.829 | R_{int} = 0.025 |
3417 measured reflections |
R[F^{2} > 2σ(F^{2})] = 0.012 | 0 restraints |
wR(F^{2}) = 0.032 | Δρ_{max} = 0.25 e Å^{−}^{3} |
S = 1.11 | Δρ_{min} = −0.35 e Å^{−}^{3} |
558 reflections | Absolute structure: Flack (1983) |
30 parameters | Absolute structure parameter: 0.18 (4) |
Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. |
Refinement. Refinement of F^{2} against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F^{2}, conventional R-factors R are based on F, with F set to zero for negative F^{2}. The threshold expression of F^{2} > σ(F^{2}) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F^{2} are statistically about twice as large as those based on F, and R- factors based on ALL data will be even larger. |
x | y | z | U_{iso}*/U_{eq} | ||
Ag | 0.49498 (2) | −0.01004 (4) | 0.5000 | 0.05041 (11) | |
Mn | 0.0000 | 0.0000 | 0.0000 | 0.02307 (12) | |
K | 0.6667 | 0.3333 | 0.0000 | 0.0394 (2) | |
C | 0.3426 (3) | −0.0042 (4) | 0.2843 (2) | 0.0391 (3) | |
N | 0.2590 (3) | 0.0040 (3) | 0.16637 (16) | 0.0376 (3) |
U^{11} | U^{22} | U^{33} | U^{12} | U^{13} | U^{23} | |
Ag | 0.05723 (13) | 0.05722 (15) | 0.03678 (13) | 0.02861 (8) | −0.02267 (6) | 0.000 |
Mn | 0.02715 (16) | 0.02715 (16) | 0.0149 (2) | 0.01357 (8) | 0.000 | 0.000 |
K | 0.0383 (3) | 0.0383 (3) | 0.0418 (5) | 0.01913 (15) | 0.000 | 0.000 |
C | 0.0378 (10) | 0.0437 (8) | 0.0346 (7) | 0.0195 (9) | −0.0068 (7) | 0.0036 (9) |
N | 0.0364 (7) | 0.0450 (7) | 0.0308 (6) | 0.0199 (7) | −0.0060 (6) | 0.0052 (8) |
Ag—C | 2.0599 (16) | K—N | 2.9266 (18) |
Mn—N | 2.2365 (13) | C—N | 1.138 (2) |
Ag···Ag^{i} | 3.3568 (7) | K···C | 3.258 (2) |
Mn···K | 3.9964 (6) | ||
C^{ii}—Ag—C | 178.05 (15) | N—K—N^{viii} | 66.75 (5) |
N^{iii}—Mn—N | 93.76 (9) | N—K—N^{ix} | 100.27 (4) |
N^{iii}—Mn—N^{iv} | 178.78 (10) | N—C—Ag | 178.3 (2) |
N—Mn—N^{iv} | 87.08 (5) | C—N—Mn | 159.36 (14) |
N^{iv}—Mn—N^{v} | 92.08 (9) | C—N—K | 96.52 (14) |
N^{vi}—K—N | 95.15 (7) | Mn—N—K | 100.58 (5) |
N^{vii}—K—N | 161.53 (6) |
Symmetry codes: (i) −y, x−y−1, z; (ii) −x+y+1, y, −z+1; (iii) −y, −x, −z; (iv) −y, x−y, z; (v) −x+y, y, −z; (vi) −x+y+1, y, −z; (vii) −y+1, −x+1, −z; (viii) x, x−y, −z; (ix) −x+y+1, −x+1, z. |
Experimental details
Crystal data | |
Chemical formula | C_{6}Ag_{3}KMnN_{6} |
M_{r} | 573.77 |
Crystal system, space group | Trigonal, P312 |
Temperature (K) | 293 |
a, c (Å) | 6.9219 (10), 8.1465 (16) |
V (Å^{3}) | 338.03 (10) |
Z | 1 |
Radiation type | Mo Kα |
µ (mm^{−}^{1}) | 5.48 |
Crystal size (mm) | 0.46 × 0.32 × 0.03 |
Data collection | |
Diffractometer | Siemens SMART CCD area-detector diffractometer |
Absorption correction | Integration (XPREP in SHELXTL; Sheldrick, 2001) |
T_{min}, T_{max} | 0.155, 0.829 |
No. of measured, independent and observed [I > 2σ(I)] reflections | 3417, 558, 551 |
R_{int} | 0.025 |
(sin θ/λ)_{max} (Å^{−}^{1}) | 0.665 |
Refinement | |
R[F^{2} > 2σ(F^{2})], wR(F^{2}), S | 0.012, 0.032, 1.11 |
No. of reflections | 558 |
No. of parameters | 30 |
Δρ_{max}, Δρ_{min} (e Å^{−}^{3}) | 0.25, −0.35 |
Absolute structure | Flack (1983) |
Absolute structure parameter | 0.18 (4) |
Computer programs: SMART (Siemens, 1995), SAINT (Bruker, 2001), SAINT, SHELXTL (Sheldrick, 2001), SHELXTL.
Ag—C | 2.0599 (16) | K—N | 2.9266 (18) |
Mn—N | 2.2365 (13) | C—N | 1.138 (2) |
C^{i}—Ag—C | 178.05 (15) | N—K—N^{vii} | 66.75 (5) |
N^{ii}—Mn—N | 93.76 (9) | N—K—N^{viii} | 100.27 (4) |
N^{ii}—Mn—N^{iii} | 178.78 (10) | N—C—Ag | 178.3 (2) |
N—Mn—N^{iii} | 87.08 (5) | C—N—Mn | 159.36 (14) |
N^{iii}—Mn—N^{iv} | 92.08 (9) | C—N—K | 96.52 (14) |
N^{v}—K—N | 95.15 (7) | Mn—N—K | 100.58 (5) |
N^{vi}—K—N | 161.53 (6) |
Symmetry codes: (i) −x+y+1, y, −z+1; (ii) −y, −x, −z; (iii) −y, x−y, z; (iv) −x+y, y, −z; (v) −x+y+1, y, −z; (vi) −y+1, −x+1, −z; (vii) x, x−y, −z; (viii) −x+y+1, −x+1, z. |
The construction of three-dimensional and interpenetrating lattices is a contemporary challenge for crystal engineering (for a review of interpenetrating networks, see Batten & Robson, 1998). Transition metals can be conveniently linked through the use of pseudohalides, leading to the formation of such polymeric networks. The magnitude of magnetic superexchange in pseudohalide-based coordination polymers has remained an area of active research interest. While the cyanide and dicyanamide anions have been extensively studied in this regard, much less is known about related structures incorporating the dicyanoargentate anion, the high aspect ratio of which makes it ideal for the formation of interpenetrating structures. We are interested in the crystallization of magnetic anionic networks and chose to study the dicyanoargentate anion, a nearly linear analog to the bent dicyanamide ligand, as a building block for cyano-bridged heterobimetallic complexes.
Pauling & Pauling (1968) suggested that very large open cubic frameworks of the type M[Ag(CN)_{2}]_{3} can be prepared through the incorporation of large neutral molecules into the center of the cubes. Alternatively, doubly or triply interpenetrating lattices can be formed to fill the voids. While attempting to crystallize the three-dimensional anionic network Mn[Ag(CN)_{2}]_{3}^{−}, in which a tetrapropyl ammonium cation would fill the cubic void, the salt KMnAg_{3}(CN)_{6} formed instead.
The structure of KMnAg_{3}(CN)_{6} consists of three interpenetrating trigonally distorted cubic lattices of composition MnAg_{3}(CN)_{6}^{−} and intercalated K^{+} ions. The six-coordinate Mn atoms occupy the corners of the cubes, whereas the two-coordinate Ag atoms are found in the middle of the cube edges. The cyanide groups bridge between the Mn and Ag atoms. Alternatively, the structure can be thought of as being composed of close-packed double layers of cyanide ions, with one-third of the octahedral holes filled by Mn, and another third by K atoms. The C atoms of the cyanide groups protrude out of the double layer in a tilted direction (along the a+c diagonal), and Ag^{+} ions then bridge the double layers via C—Ag—C bonds.
We have carefully established (see Experimental) that the cyanide groups bind to Mn through the N atoms and to Ag through the C atoms. This assignment is also consistent with the observed bond lengths of 2.060 (2) Å for Ag—C and 2.236 (1) Å for Mn—N. These short Ag—C and relatively long Mn—N bond lengths are consistent with other cyanide-bridged Mn—Ag systems, e.g. catena-(tetrakis{µ_{2}-cyano-bis[2-(4-pyridyl)-4,4,5,5- tetramethylimidazoline-1-oxyl-3-oxide]disilvermanganese(II)} Please check brackets − a closing) is missing (Dasna et al., 2001), where the corresponding values are 2.060–2.072 and 2.252–2.253 Å, respectively. The site symmetry on the Mn center is 32, with a small trigonal distortion from octahedral symmetry corresponding to a rotation of the `top' versus the `bottom' (with respect to the ab plane) N3 triangle of 1.5 (1)° and a 5.0 (2)% elongation along the c axis. The distortion at the K site is much more severe, i.e. a rotation of 20.9 (1)° and a 19.8 (2)% trigonal flattening. Except for the K atom, the deviations from P31m symmetry are slight, e.g. the Ag(CN)_{2}^{−} group is essentially linear (Table 1), and the trans-N—Mn—N angles are 178.8 (1)°.
A similar triply interpenetrating MAg_{3}(CN)_{6} network has been found in a number of compounds, e.g. CoAg_{3}(CN)_{6} (Pauling & Pauling, 1968; Ludi & Güdel, 1968), K_{2}NaAg_{3}(CN)_{6} (Zabel et al., 1989), and RbCdAg_{3}(CN)_{6} (Hoskins et al., 1994), but the title compound is the first example containing the Mn^{2+} ion at the six-coordinate site. In the first two examples, the symmetry is P31m, because the two extra octahedral sites are either both empty or both filled. In contrast, RbCdAg_{3}(CN)_{6} is isomorphous with the title compound, as is gold-containing KCoAu_{3}(CN)_{6}, with a demonstrated piezoelectric effect as proof for the noncentrosymmetric crystal symmetry (Abrahams et al., 1980). In KAg(CN)_{2} (Hoard, 1933) and KAu(CN)_{2} (Rosenzweig & Cromer, 1959), all octahedral holes are filled, but the sequence of double layers contains additional shifts, therefore leading to a doubling and tripling, respectively, of the c axis. Interestingly, CoAg_{3}(CN)_{6} is the only salt in the series where the Co ion was assumed, without discussion by either group of authors (Pauling & Pauling, 1968; Ludi & Güdel, 1968), to be bonded to the C atom of the cyanide group. It would be interesting to examine whether this assumption holds in a rigorous refinement of diffraction data obtained by the use of modern instrumentation.