You are on page 1of 13

PAPER

www.rsc.org/crystengcomm | CrystEngComm

Crystal structures of quinacridones{


Erich F. Paulus,a Frank J. J. Leusenb and Martin U. Schmidt*c
Received 8th September 2006, Accepted 24th November 2006 First published as an Advance Article on the web 7th December 2006 DOI: 10.1039/b613059c The crystal structure of the aI-phase of quinacridone was determined from non-indexed X-ray powder data by means of crystal structure prediction and subsequent Rietveld refinement. This aI-phase is another polymorph than the a-phase reported by Lincke [G. Lincke and H.-U. Finzel, Cryst. Res. Technol. 1996, 31, 441452.]. The crystal structures of the b and c polymorphs were determined from single crystal data. The knowledge of the crystal structures can be used for crystal engineering, i.e., for targeted syntheses of pigments having desired properties, especially for the syntheses of new red pigments.

Introduction
Quinacridone (Pigment Violet 19, formula 1) is the most important pigment for redviolet shades. The annual production totals several 1000 tons with a sales volume of more than 100 million euros per year. The b-phase is reddish violet whereas the c-phase is red. Both phases are used for the colouration of laquers and paints, plastics and printing inks.1 In solution quinacridone is yellow (see Fig. 1).

As visible from the different colours of the individual polymorphs, the crystal structure has a large impact on the pigment properties. For any structureproperty relationship, as well as for crystal engineering, knowledge of the crystal structures is required. In earlier publications, the crystal structures of a and c quinacridone have been published,2,3 but the structure of the a-phase may be questionable. The structures of the aI and b phases were published only on conferences4,5 and in a survey article.6 Here we report the crystal structures of the aI, b, and c phases.

Fig. 1 Colours of quinacridone polymorphs (industrial samples) (a): From left to right: aI, aII, b, and c phases. Far right: dichloroquinacridone (Pigment Red 209). (b) Quinacridone in solution (small amounts of quinacridone dissolved in 500 ml boiling DMSO at 189 uC; photo taken at about 185 uC).

1. Polymorphs of quinacridone: A real chaos


Various polymorphs have been described in patents and journals, including the phases a, b, BI, c, c9, cI, cII, cIII, cIV,
Institut fu r Geowissenschaften Facheinheit Mineralogie/ Kristallographie, Johann Wolfgang Goethe-Universita t Frankfurt am Main, Senckenberganlage 30, D-60054, Frankfurt am Main, Germany b Institute of Pharmaceutical Innovation, University of Bradford, Bradford, UK BD7 1DP c Institut fu r Anorganische und Analytische Chemie, Johann Wolfgang Goethe-Universita t Frankfurt am Main, Max-von-Laue-Str. 7, D-60438, Frankfurt am Main, Germany. E-mail: m.schmidt@chemie.uni-frankfurt.de { Electronic supplementary information (ESI) available: Additional crystallographic data for the c phase (Tables S1 and S2, Fig. S1). See DOI: 10.1039/b613059c
a

d, D, e, and f.718 All phases were characterised by X-ray powder diffraction. The phases a, b, and c were found already in 1955.7,8 A closer look at the powder diagrams of the individual phases reveals that in fact b and BI9 describe identical phases, and all c-phases3,1016 belong to only one polymorphic form (c).{ Furthermore, the d-phases are either equal to the c-phase14 or they consist of a mixture of c with a trace of b-phase.17 Also the D-phase18 is the same polymorph as c. To complete the chaos, two e-polymorphs have been described, stating that they clearly differ from each

{ The differences seen in the X-ray powder diagrams of the various c phases result probably from (i) differences in crystal size, morphology, and lattice defects, all resulting in isotropic or anisotropic peak broadening, thus affecting the peak heights and the overlapping of neighbouring peaks; (ii) inadequate measurement conditions, e.g. measurements in reflection mode resulting in preferred orientation effects; (iii) contamination with other polymorphic phases (e.g. cIII contains a); (iv) contamination with additives, starting materials or byproducts, which are incorporated in the crystal lattice, causing lattice distortions and thereby shifts in the peak positions (an investigation on the lattice distortions in 21 different c-quinacridones is given by Lincke.16). Furthermore, peak positions listed in patents can be affected by zero point errors or by admixture of Ka2 radiation. Additionally the peak positions depend on the algorithm for extracting the positions from the diagram.

This journal is The Royal Society of Chemistry 2007

CrystEngComm, 2007, 9, 131143 | 131

other,14,18 but in fact both are c-phases. The f-phase seems to be a mixture of at least three different phases (a, b, and c). Hence, are there only those 3 polymorphs left, which have already been described in 1955, namely a, b, and c? No, there are four polymorphs! Because, what is described as a-phase, are indeed two different phases, which we will denote as aI and aII: aI is the phase described by W. S. Struve,8 Labana and Labana19 and others,14,18 which is formed during synthesis or by grinding with NaCl. aII is the phase investigated by Lincke and Finzel.2 This phase is formed upon recrystallisation in H2SO4 (see chapter 7). The colours of aI and aII are considerably different: whereas I a has a dull reddish-violet shade, aII is red (only slightly more bluish than c-quinacridone), see Fig. 1. Also the powder diagrams of aI and aII are clearly different, as can be seen in Fig. 2. Both a-phases are stable, also at elevated temperatures: e.g. Ogawa et al. obtained the aI-phase from sublimation at 140170 uC,20 and Lincke treated aII crystals in solvents at 7080u for 4 weeks;2 at room temperature the a-phases are stable for many years.21 Harsh conditions are required to transform the a-phases to the b- or c-phase (see following chapter).

2. Industrial syntheses and applications


Quinacridone has been known since 193522 and industrially produced since 1958.23 There are several synthetic routes; the most important one is shown in Scheme 1. The crystal structures of the intermediates sodium aniloate (2), anilic acid (3) and its calcium salt were recently determined.24 The final ring closures are achieved by a treatment in polyphosphoric acid (!) at 120 to 140u, followed by hydrolysis with water (warning: vivid exothermic reaction). The resulting quinacridone precipitates as a fine, insoluble powder. Depending on the synthetic conditions, the syntheses can give the aI, aII , b or c phases, or a mixture of phases.

Scheme 1 Industrial synthesis of quinacridone.

Fig. 2 X-Ray powder diagrams of quinacridone: From the top: aI, aII, b, and c phases. The diagrams were measured in transmission on a STOE-Stadi-P diffractometer with curved Ge[111] monochromator, using Cu Ka1 radiation and a linear position-sensitive detector.

The b and c phases are produced industrially, either directly, or via the a-phases.2528 Tradenames are e.g. 1Hostaperm Red Violet ER02 for the b-phase and 1Hostaperm Red E5B02 for the c-phase. b and c phases do not interconvert; both are stable up to high temperatures. The b and c phases have high photostabilities and high fastness to weathering. Therefore they are used for automotive finishes, powder coatings, paints, plastics and high-grade printing inks. The a-phases are not commercially used (except as intermediates in the syntheses of b and c phases) due to their less than optimal application properties, and because they may convert to the b or c phases during their application in a coating or plastics at elevated temperatures. Quinacridone is internationally registered in the Colour Index as C.I. Pigment Violet 19, independently of the phase and of the producer. The German name for quinacridone, Chinacridon implies an etymological connection with China, but in fact quinacridone is not a Chinacridon but a Chin-acridon, namely the 5,12-dihydroquino[2,3-b]acridine-7,14-dione.
This journal is The Royal Society of Chemistry 2007

132 | CrystEngComm, 2007, 9, 131143

3. Crystallisation of quinacridones
All polymorphs of quinacridone are insoluble or nearly insoluble in water and all other solvents, even at elevated temperatures. This is a typical behaviour for organic pigments, caused by the combination of hydrogen bridges and very dense van der Waals packing, resulting in high lattice energies.29 There are only two ways to recrystallise quinacridone: by vacuum sublimation, and by protonation, e.g. using concentrated sulfuric acid, with subsequent dilution or evaporation. If a solution of quinacridone in concentrated sulfuric acid is placed on a glass slide, the solution absorbs moisture from the air, and the growth of quinacridone crystals can be observed under the microscope within a few minutes.30 By this procedure, the aII-phase is formed as radial bundles of needles, which are not suitable for single crystal X-ray analysis, and even do not give a good X-ray powder diagram. Generally, crystals of quinacridone show many lattice defects and a strong mosaicity; sometimes they are strongly bent (see Lincke31 for impressive photos of such bad crystals).

Table 2

Refinement data for b-quinacridone C20 H12 N2 O2 312.32 293(2) 0.71070 1.538 0.101 324 0.45 6 0.17 6 0.05 2.73 to 24.00u 26 h 6, 0 k 4, 234 l 34 2124/1064 [R(int) = 0.1302] 99.9% to h = 24.00u Full-matrix least-squares on F2 1064/0/109 0.908 R1 = 0.0898, wR2 = 0.2057 R1 = 0.1997, wR2 = 0.2728 0.276 and 20.300 e A23

Empirical formula Formula weight Temperature/K Wavelength/A Calculated density/Mg m23 Absorption coefficient/mm21 F(000) Crystal size/mm3 h range for data collection Limiting indices Reflections collected/unique Completeness Refinement method Data/restraints/parameters Goodness-of-fit on F2 Final R indices [I . 2s(I)] R indices (all data) Largest diff. peak and hole

Table 3 Atomic coordinates and equivalent isotropic displacement 2) for b-quinacridone. Ueq is defined as one third parameters (in 1023 A of the trace of the orthogonalized Uij tensor x y 0.4159(15) 20.1576(17) 0.221(2) 0.125(2) 20.050(2) 20.143(2) 20.062(2) 0.133(2) 20.076(2) 0.110(2) 0.228(2) 0.181(2) z 0.0873(2) 0.0778(2) 0.1677(2) 0.2043(3) 0.1986(2) 0.1572(2) 0.1197(2) 0.1242(2) 0.0385(2) 0.0414(2) 0.0847(2) 0.0021(2) Ueq 59(2) 47(2) 57(2) 59(2) 56(2) 53(2) 46(2) 46(2) 43(2) 41(2) 43(2) 43(2)

4. Crystal structure of b-quinacridone


Crystals of b-quinacridone were obtained from Prof. Lincke. A crystal with dimensions 0.45 6 0.17 6 0.05 mm3 was fixed in a Mark tube using a tiny amount of grease and placed on a 4-circle diffractometer (Nicolet) equipped with a scintillation counter. We used the omega scan to improve the resolution of neighboured reflections (very high mosaicity, relatively large c axis). The speed of measuring was varied between 2 and 20u min21, depending on the weakness of a reflection. Every 68 12), the mutual deviations reflections the standard reflection (1 of which were less than 1.7%, was measured again. The phase problem was solved by direct methods using the program SHELXTL.33 The non-hydrogen atoms could be found in the phased Fourier map, whereas the hydrogen atoms had to be included in calculated positions. Despite the low crystal quality, it was possible to refine the non-hydrogen atoms anisotropically with reasonable results. Hydrogen atom posi. tions were calculated with a CH distance of 0.93 A Crystallographic and refinement data are given in Tables 1 and 2. Atomic coordinates of the non-hydrogen atoms are shown in Table 3. CCDC reference numbers 620257620259. For crystallographic data in CIF or other electronic format see DOI: 10.1039/b613059c
Table 1 Crystal data for aI, b and c quinacridones (standard deviations in brackets) Crystal phase Space group, Z Unit cell dimensions a/A b/A c/A a/u b/u c/u 3 Volume V/A Temperature T/K aI , Z = 1 P1 3.802(2) 6.612(3) 14.485(6) 100.68(8) 94.40(6) 102.11(5) 346.7(1) 293(2) b P21/c, Z = 2 5.692(1) 3.975(1) 30.02(4) 90. 96.76(6)u 90. 674.5(9) 293(2) c P21/c, Z = 2 13.697(9) 3.881(3) 13.4020(10) 90. 100.44(1)u 90. 700.6(7) 293(2)
32

O(1) N(1) C(01) C(02) C(03) C(04) C(05) C(06) C(07) C(08) C(09) C(10)

0.2030(8) 0.7900(9) 0.4916(13) 0.6420(13) 0.8501(12) 0.9001(12) 0.7430(10) 0.5353(11) 0.6497(10) 0.4406(11) 0.3805(12) 0.2955(11)

The molecular structure of quinacridone in the b-phase is shown in Fig. 3. The angles between the different rings of the molecule are smaller than 1.7u, i.e. the molecule is planar. The . crystallographic site symmetry is 1 In all quinacridone polymorphs, the molecules are connected with their neighbours by 4 hydrogen bonds of the type NHOLC. In b-quinacridone, each molecule is bonded to two neighbouring molecules via two hydrogen bonds each (Fig. 4, SCHAKAL plot34). The resulting chains are not parallel, but half of the chains run in the [110] direction, the 0] direction (Fig. 5). Nevertheless all chains other half in the [11 are symmetrically equivalent. In the b direction, the molecules form stacks. The normal vector of the molecular plane is tilted by 32.0u with respect to the b axis. These stacked chains form layers parallel to (001), which are held together by van der Waals interactions only. The chains themselves are not exactly planar, but exhibit small steps between the molecules. The height of the steps is . In order to investigate if these steps are caused by a 0.35 A packing effect, we performed energy minimisations on the crystal structure of b-quinacridone, and on a single molecular chain. We used the Dreiding force field35 with atomic charges calculated by the Gasteiger method.36 Upon optimisation, the isolated molecular chain is exactly planar, whereas in the optimised b-quinacridone packing the chains continue to
CrystEngComm, 2007, 9, 131143 | 133

This journal is The Royal Society of Chemistry 2007

Fig. 3

Molecular structure of quinacridone in the b-phase. Ellipsoids are drawn with 50% probability.

easily observe whether the applied calculation methods are suitable to describe the various intermolecular interactions. Furthermore, one can try to reproduce the UV/vis solid state spectra37 of b and c quinacridones.

5. Crystal structure of c-quinacridone


The crystal structure of c-quinacridone was first investigated by Koyama et al. in 1966.38 They determined the correct unit cell and space group, however, due to limitations in their data they found that the molecule is not planar, but adopts an S shape: the pyridone ring was strongly bent along the NCLO axis, resulting in interplanar angles of about 40u between the terminal phenyl and the central benzene rings. Later this turned out to be wrong; in fact the molecule is planar. Crystal data and figures of this structure analysis are also included in a Japanese paper by Nagai and Nishi in 1968.39 In 1971, the crystal structure was investigated by Chung and Scott, but the structure could not be solved.40 For our X-ray structure determination we used crystals of c-quinacridone, which were grown by sublimation in vacuum at 300 uC.41 The best crystal was a thin, bent plate with dimensions of 0.65 6 0.3 6 0.01 mm3. Details of the crystal structure solution and refinement are given in the ESI.{ The crystal structures of b- and c-quinacridone were published at conferences by Paulus in 1989 and Dietz in 1991.4,5 The structure of c-quinacridone was confirmed (with better R values) by Potts et al. in 1994 and Mizuguchi et al. in 2002.3 Potts et al. grew single crystals by sublimation at 420 uC at about 1023 mbar, yielding small, red plate-like crystals. One of these crystals, with dimensions 0.35 6 0.075 6 0.015 mm3, was measured using synchrotron rays and an area detector. Mizuguchi et al. achieved to grow a single crystal with dimensions 0.33 6 0.10 6 0.04 mm3 from DMF solution after having purified the compound twice by sublimation at about 430 uC. The crystal packing of the b and c phases are remarkably different: In b-quinacridone, each molecule is connected to two neighbouring molecules by two hydrogen bonds each, but in c-quinacridone, each molecule is connected by single hydrogen bonds to four neighbouring molecules, see Fig. 6. Consequently, in the c-phase, the molecules are not arranged in chains, but they form a criss-cross pattern, see Fig. 7 and 8. Along the b axis, the molecules are stacked; the normal vector
This journal is The Royal Society of Chemistry 2007

0]. Fig. 4 b-Quinacridone, view direction [11

5]. The chains Fig. 5 b-Quinacridone, view direction approx. [73 0] direction are drawn darker than the chains running running in the [11 in the [110] direction.

exhibit steps. This finding shows that the steps are not caused by interactions within the chain, but by the stacking of the chains: the formation of the steps decreases the empty volume between the edges of the molecules in neighbouring stacks. b-Quinacridone is an ideal test case for quantum mechanical calculations in the solid state: in the a direction, the molecules are connected by hydrogen bridges, in the b direction the molecules are held together by van der Waals and electrostatic interactions (dispersion, induction, polarisation and Coulomb energies etc.), whereas in the c direction, there are only van der Waals interactions between C and H atoms. Hence from the lattice parameters a, b, and c of an optimised structure one can
134 | CrystEngComm, 2007, 9, 131143

b and c quinacridone crystallise in the same space group: P21/c, Z = 2, with molecules on inversion centres. But in the c-phase, the c axis is doubled and the a axis is halved. Despite the considerably different lattice parameters and the completely different packings, the X-ray powder diagrams show some common features (see Fig. 2). The b phase is about 3 percent more dense than the c phase, but according to experimental observations the c phase seems to be the more stable one, at least at high temperatures. Also in the lattice energy calculations (see below) the c phase is calculated to be thermodynamically more stable than the b phase.

6. Crystal structure of aI-quinacridone


Fig. 6 c-Quinacridone; one molecule with 4 neighbours. View 0]. c axis horizontal, a and b axes vertical. direction [13

6.1. X-Ray powder diagrams and solid state NMR Single crystals of this phase cannot be grown. The powder diagrams typically show 89 broad peaks; hence indexing is not possible (later the compound turned out to be triclinic; but with 6 lattice parameters every set of 89 broad lines could be indexed). Solid state 13C- and 15N-NMR measurements were carried out under cross-polarisation magic-angle-spinning conditions. Under these conditions crystallographically equivalent atoms are magnetically equivalent; if the compound contains more than one molecule per asymmetric unit, the NMR peaks start to split. The spectra of aI, b, and c quinacridone are different from each other, but in all cases it is clearly visible that the crystals contain only half a molecule per asymmetric unit. 6.2. Crystal structure prediction of quinacridone polymorphs The crystal structures of quinacridone were predicted by one of the authors (FJJL) in 1995.6 The b and c phases were reproduced well, and the unknown structure of the aI-phase was solved, and subsequently refined by Rietveld methods. The results were presented42,43 on various occasions in order to demonstrate the power of the newly developed polymorph predictor software package, but the atomic coordinates have not yet been published. Recently, Panina repeated the calculations, without distinguishing between aI and aII phases.44 Here we report the original work of Leusen and Paulus,45 and add a careful Rietveld refinement of the aI-phase; additionally we made calculations on the possible disorder in aI-quinacridone. A crystal structure prediction is the determination of the energetically favourable packings of a molecule with a given molecular geometry.46 The molecular structure of quinacridone was optimised with the 631 G** basis set in the ab initio quantum mechanics package Gaussian92,47 and atomic charges were fitted to the electrostatic potential.48 These charges were used in combination with the Dreiding 2.21 force field,35 and Ewald summation.49 For the prediction of possible polymorphs, the program polymorph predictor within the Cerius2 molecular modelling environment50 was used. The algorithm is based on the method of Gdanitz and Karfunkel.51 Firstly, a rigid body Monte Carlo search procedure ran in the top 17 crystallographic space groups in order of occurrence (together accounting for more than 95% of the known molecular crystals52). A total of 29 365 trial structures were generated. Secondly, a clustering algorithm was applied to
CrystEngComm, 2007, 9, 131143 | 135

Fig. 7 c-Quinacridone, view direction [010].

Fig. 8 c-Quinacridone, view direction [001]. a axis horizontal, b axis vertical.

of the molecules is tilted with respect to the b axis by 37.1u. The packing differences are the reason for the different colours of b- and c-quinacridone (see section 8).
This journal is The Royal Society of Chemistry 2007

remove duplicate structures. Finally, the 408 most promising and distinct structures were subjected to a lattice energy minimisation with respect to all degrees of freedom (lattice parameters, position and orientation of molecules, intramolecular flexibility). The resulting 13 lowest energy crystal structures are listed in Table 4. The calculations were made with one molecule per asymmetric unit, situated on a general position. Since the molecule has 2/m symmetry, higher crystal symmetries could occur during the minimisations. In order to verify the polymorph prediction results, simulated powder X-ray patterns of the predicted crystal structures were compared to the experimental powder patterns. The most stable predicted structure (No. 1) was found in calculations made in space group P21, Z = 2. The final structure has P21/c symmetry. Its simulated powder pattern was in good agreement with the experimental pattern of the c polymorph. Rietveld refinement53 was applied to refine the structure to a Rp -factor of 9.6%. The result is in excellent agreement with the c-quinacridone structure determined from single crystal data. Structure 8 (as numbered in Table 4) showed a good fit with the powder pattern of the b polymorph (Rp -factor of 12.4% after Rietveld refinement), and indeed resembles the structure of the b polymorph. Also here the space group changed from P21 to P21/c upon optimisation. Structures 12 and 13 are in fact identical. In both cases the optimised structures have additional inversion centres, and the , Z = 1 with the molecule on resulting crystal symmetry is P1 the inversion centre. The simulated X-ray powder diagram was similar to the experimental diagram of the aI-phase. A preliminary Rietveld refinement with DBWS converged with a Rp -factor of 9.7%. These Rp factors for the aI, b, and c phases prove beyond any doubt that the crystal structures of the three quinacridone polymorphs, including the previously undetermined aI form, have been successfully predictedin the correct stability order. What are the other structures listed in Table 4? Structure 2 is similar to c (structure 1), but in the wrong space group (Pbca, Z = 8). Apparently, the c packing is most favourable in this space group, although the symmetry does not
Table 4 Predicted polymorphs of quinacridone Lattice parameters No. Space group
a

allow reproduction of the c structure. Also 4,11-dichloroquinacridone crystallises in Pbca with a criss-cross lattice, but with Z = 4.4,40 Structures 3, 5 and 7 are termed pseudo-c1; their packing within the layers of hydrogen bonded molecules is identical to c, but the space group symmetry does not allow the efficient inter-lacing packing at one side of each layer. The van der Waals energy penalty is 1.2 kcal mol21 in comparison to c, and the density decrease is 0.05 g cm23. For structures 9 and 11, termed pseudo c2, the space group symmetry prohibits the inter-lacing packing at both sides of each layer. Consequently, the van der Waals energy penalty and density decrease are about twice the values observed for pseudo-c1: 2.3 kcal mol21 and 0.10 g cm23, respectively. A similar analysis applies to structure 10 (pseudo-b), when compared to b (structure 8). These eight structures can therefore be discarded. Finally, structure 4, which is reproduced by structure 6 in a different space group, shows a hydrogen bonding pattern identical to c. However, the orientation of the molecules with respect to each other is different, as if the stacks of molecules are squashed. Despite a higher density than c, the Coulomb contribution to the lattice energy is about 1 kcal mol21 less favourable. Since this polymorph is not observed experimentally, a crystal dynamics simulation was performed to assess its thermodynamic stability at room temperature (using exactly the same force field and charges as applied in the prediction sequence). During the simulation, which was performed with a constant number of molecules in the lattice, constant pressure and temperature (300 K), both the unit cell and its contents were fully flexible. After about 30 ps, the structure decayed to the c polymorph, which shows that the energetic path leading from this polymorph (predicted at 0 K but not stable at room temperature) to the stable c form has a low barrier due to the identical hydrogen bonding pattern. 6.3. Rietveld refinement of the aI-phase In order to determine the crystal structure of aI-quinacridone as accurately as possible, the powder diagram was carefully measured in transmission geometry on a STOE-Stadi-P diffractometer equipped with a curved Ge[111] monochromator and a linear position sensitive detector. Cu Ka1

Lattice energy/kcal mol21 a/u b/u c/u Density/g cm


23

a/A

b/A

c/A

Total 2366.5 2366.1 2365.4 2365.2 2365.1 2365.0 2364.8 2364.5 2364.0 2364.0 2363.9 2363.8 2363.8

v. d. Waals 24.7 25.4 25.9 24.6 26.1 26.0 26.4 22.2 27.0 22.3 27.1 22.3 22.0

Coulomb 2392.3 2392.2 2392.1 2391.4 2391.9 2392.3 2392.0 2389.1 2391.8 2388.9 2391.8 2389.3 2389.3

H-bond 27.2 27.2 27.2 26.9 27.2 27.0 27.2 25.8 27.2 25.8 27.2 25.6 25.4

Phase c pseudo-c pseudo-c1 new pseudo-c1 new pseudo-c1 b pseudo-c2 pseudo-b pseudo-c2 a a

2 13.86 3.96 1 P21 2 Pbca 8 13.50 55.25 3 C2/c 8 56.52 3.98 4 P21 2 14.93 6.38 5 Pnma 8 13.55 57.22 6 Pbca 8 13.04 35.04 7 Pbcn 8 58.61 3.82 8 P21 2 4.12 5.62 4 13.58 3.88 9 Pna21 10 Pca21 4 61.44 5.73 11 C2 4 29.32 3.91 12 P1 2 4.03 32.19 13 P1 1 3.94 16.15 a Space group used in the prediction (Z9

13.45 90.0 78.2 90.0 1.436 3.95 90.0 90.0 90.0 1.408 13.47 90.0 80.4 90.0 1.390 8.51 90.0 60.6 90.0 1.468 3.90 90.0 90.0 90.0 1.371 6.34 90.0 90.0 90.0 1.432 13.62 90.0 90.0 90.0 1.361 33.50 90.0 117.4 90.0 1.507 29.36 90.0 90.0 90.0 1.341 3.96 90.0 90.0 90.0 1.487 13.54 90.0 89.8 90.0 1.337 6.91 85.6 64.9 60.8 1.484 6.90 80.4 64.7 62.4 1.477 = 1, molecule on general position).

136 | CrystEngComm, 2007, 9, 131143

This journal is The Royal Society of Chemistry 2007

Fig. 9 Rietveld plot of aI-quinacridone. Measured diagram black; simulated diagram red; background green; difference plot blue.

radiation was used. The sample was spinning during the measurement. The powder diagram was of very low quality, which was caused by the low crystallinity, not by the measurement conditions. The Rietveld refinement was carried out with the program GSAS.54 Since the powder diagram showed just some humps in the range 2h . 35u, only the 2h range 335u was taken into account for the refinement. The profile was described by a pseudo-Voigt function55 with the asymmetry correction of Finger, Cox and Jephcoat.56 In the first step only the scaling factor was refined. Subsequently restraints for bond distances, bond angles and planar groups were introduced. The crystallographic inversion centre of the molecule was used: the Rietveld refinement was carried out with half a molecule, which was fixed to the inversion centre using a dummy atom at (0,K,0). Hydrogen atoms were included throughout the whole refinement. Generally a Le Bail fit is carried out before starting the Rietveld refinement. For aI-quinacridone, the Le Bail fit looked promising (R = 2.3%, Rwp = 3.0%, red. x2 = 5.356), but using the profile parameters from the Le Bail fit in subsequent Rietveld refinement did not result in a reliable refinement. This may be caused by the peak broadening, which did not allow a reliable determination of peak profile parameters in the Le Bail step. Therefore we started directly with the Rietveld refinement; atomic coordinates, peak profile parameters, and lattice parameters were refined alternately. The Rietveld refinement converged with R = 3.7%, Rwp = 4.8%, (Rexp = 1.4%, RF2 = 4.5%), red. x2 = 11.90 for 92 reflections in the 2h range 3.0 to 35.0u. Although the applied restraints were weak, the resulting molecular structure was close to the molecular structure found in the single crystal structure determinations. The final Rietveld plot is shown in Fig. 9. Crystallographic data are included in Table 1. Atomic coordinates are given in Table 5. 6.4. Description of the crystal structure of aI-quinacridone In aI-quinacridone, the molecules show the same hydrogen bonding pattern as in the b polymorph, i.e. each molecule is connected to two neighbouring molecules, thus forming a molecular chain (Fig. 10). The main difference between the
This journal is The Royal Society of Chemistry 2007

structures of the aI- and the b-phases is that all chains are parallel in the aI-phase, whereas there are two different chain orientations in the b-phase. In aI- as well as in b-quinacridone, the chains are not completely planar, but there are small steps between the molecules (Fig. 11). Lattice energy calculations again show that these steps must be considered as a packing effect caused by the stacking of the molecules in the a direction. aI-Quinacridone is isostructural to 2,9-dimethylquinacridone; both compounds form a continuous series of mixed crystals (solid solutions). In principle, aI-quinacridone can also be used as a test structure for quantum mechanical calculations in the solid state. The calculations may even be easier than for b-quinacridone, since in aI-quinacridone there is only one molecule per unit cell. On the other hand, the accuracy of the structural data for aI-quinacridone is limited; thus, the isostructural compound 2,9-dimethylquinacridone would probably be a better choice. 6.5. Calculation of disorder in aI-quinacridone The molecular packing of aI-quinacridone would not change, if all quinacridone molecules were rotated by 180u around the
Table 5 Atomic coordinates for aI-quinacridone from Rietveld refinement (standard deviations in brackets) x C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 H12 H13 H14 H15 H16 H17 N18 O19 20.1801(9) 20.0950(6) 20.2214(7) 20.4174(7) 20.4314(6) 20.2718(7) 20.0973(5) 0.0479(9) 0.0258(9) 20.1523(8) 0.0137(9) 20.2016(22) 20.5250(19) 20.5273(9) 20.4150(9) 20.3004(13) 20.3104(6) 0.2160(13) y 0.2899(8) 0.9384(3) 0.8813(3) 0.6726(3) 0.5204(3) 0.5766(3) 0.7822(3) 0.8326(3) 0.6712(7) 0.4601(3) 1.080(2) 0.9878(6) 0.6310(4) 0.3772(13) 0.2832(13) 0.1535(15) 0.4163(3) 1.0388(14) z 0.0039(2) 0.3294(1) 0.4013(1) 0.3955(1) 0.3149(1) 0.2449(1) 0.2491(1) 0.1582(1) 0.0767(2) 0.0806(2) 0.3320(2) 0.4616(3) 0.4523(8) 0.3166(2) 0.1618(2) 0.0079(3) 0.1616(1) 0.1601(3)

CrystEngComm, 2007, 9, 131143 | 137

Fig. 10 aI-Quinacridone, view direction [100].

long axis of the molecule, i.e. by exchanging the NH and CLO groups. The hydrogen bond pattern would also be maintained. Lattice energy minimisations show that this alternative structure is worse in energy by only 0.21 kcal mol21, which is not a significant value and it depends on the charge model

used. Hence, from lattice energy calculations we cannot decide which orientation is the correct one. We also calculated the energy when only one chain is rotated (using a larger superstructure with 9 molecules per unit cell): After optimisation the energy is only 0.24 kcal mol21 higher than the original

11]. Fig. 11 aI-Quinacridone, view direction approx. [1

138 | CrystEngComm, 2007, 9, 131143

This journal is The Royal Society of Chemistry 2007

structure. The small energy differences between all these models indicate that the real structure of aI-quinacridone may be disordered, i.e. it may contain chains, which are rotated by 180u along their chain axis. This disorder of a few single chains would also be possible for the b-phase. If only a single molecule is rotated, the energy gets very high (increase of at least 20 kcal mol21), and the structure in the neighbourhood of the rotated molecule gets distorted, because energetically unfavourable CLOOLC and NHHN contacts are formed. This indicates that although the whole chains may be disordered, the local ordering of molecules within the chains is very high. The solid state NMR experiments did not indicate disorders, but it is questionable whether a rotation of a whole chain would be detectable. In the Rietveld refinement, the rotation of all quinacridone molecules would lead to a good fit with R values similar to the R values from the fit with the original orientation. The reason is that a carbon atom has almost the same diffracting power as a NH group, thus it is only the oxygen atom which makes the difference. But the location of a single oxygen atom is not reliable with the present data. Hence it cannot be ruled out that in the true structure of aI-quinacridone, the positions of the CLO and NH groups have to be exchanged. In any case, the packing motif (parallel chains of molecules) would be maintained. 6.6. Electron diffraction on aI-quinacridone The aI-phase was investigated by electron diffraction by Ogawa et al:20 Purified quinacridone was vacuum-deposited on alkali halide single crystals at 140170 uC, and quinacridone single crystals with sizes up to 700 6 100 6 20 nm were grown. By tilting the sample in the transmission electron microscope, the authors found the same d-spacings as Labana et al.,19 which confirms that the sample contained the aI-phase. The pattern was indexed and the intensities of 120 h0l reflections were measured. Ogawa et al. also succeeded in getting high-resolution TEM images showing the lattice of quinacridone in the (010) plane (Fig. 12). The crystal structure of aI-quinacridone was solved from a Patterson map and the HRTEM images, and refined against the observed intensities. The lattice parameters , c = 6.37 A , b = 103u, assuming a = c = were a = 14.5 A 90u. The parameter b was estimated to be about 4 A leading to Z = 1. The molecule was found to be inclined against the (010) plane. The space group was not given explicitly, but crystallographic considerations lead to P1 as the only possibility: from the lattice parameters, the crystal system seems to be monoclinic. There are only three monoclinic space groups which allow for Z = 1, namely P2, Pm, and P2/m. But all these space groups require the molecule to be exactly parallel to the (010) plane; thus monoclinic space groups can be ruled out and the crystal lattice must be triclinic. Consequently, the angles a and c (which cannot be measured from h0l reflections), may be different from 90u. In a triclinic system with Z = 1, the space , since the molecule has an inversion centre. group must be P1
This journal is The Royal Society of Chemistry 2007

Fig. 12 High resolution transmission electron micrograph of aI-quinacridone. Image kindly provided by T. Ogawa.

To the best of our knowledge the atomic coordinates have not been published yet. Since a = c = 90u was assumed in the electron diffraction, the lattice parameters in real space are different from those determined from X-ray powder data. However a comparison of the reciprocal lattice parameters shows that the electron diffraction data correspond quite well to our result, which confirms our structure solution (Table 6).

7. On the structure of the aII-phase


In 1996, Lincke and Finzel tried to solve the crystal structure of aII-quinacridone.2 They dissolved c-quinacridone from Ciba-Geigy in 96% sulfuric acid, and added this solution to a beaker with water and ice under agitation, causing the quinacridone to precipitate immediately. The crystal quality of the resulting powder was subsequently improved by heating in 3-nitrotoluene at 7080 uC for 28 d. X-ray powder diagrams were measured in reflection mode on a Philips PW1730 diffractometer, giving an X-ray powder diagram with 14 lines. with two The crystal structure was constructed in P1
Table 6 aI-Quinacridone. Comparison of the crystal structure determined by X-ray powder diffraction and by electron diffraction Parameter 1/a*/A 1/b*/A 1/c*/A a*/u b*/u c*/u X-Ray powder diffractiona 14.131 3.690 6.325 103.22 101.92 83.09 Electron diffraction 14.1 ca. 4 6.21 90 (assumed) 103 90 (assumed)

a Structure from crystal structure prediction with subsequent Rietveld refinement; unit cell transformed with a9 = 2c, b9 = 2a, c9 = b.

CrystEngComm, 2007, 9, 131143 | 139

independent molecules per unit cell, both located on inversion centres. The molecules were assumed to form a criss-cross pattern like in c-quinacridone. This packing was manually fitted to the powder diagram. The final lattice parameters were , a = 107.13, b = given as a = 14.934, b = 3.622, c = 12.935 A 2 92.84, c = 91.39u; but from the hkl values given and from the figures shown in the paper, it becomes clear that the angles have to be exchanged, and the correct angles are b = 107.12, c = 92.84, a = 91.39u. Lincke noticed the differences between his structure and the structure of Leusen and Paulus, and wrote in a letter to Paulus on May 27, 1997, that there are obviously two different alpha-phases: the pigmentary a-phase and the crude a-phase (i.e. aII and aI, respectively). In the polymorph prediction, the structure proposed by Lincke and Finzel could not have been found since it contains two symmetrically independent molecules, which was outside the search range of the predictions and could not be reached by any groupsupergroup transition. The proposed aII structure looks chemically sensible and the simulated powder diagram has some similarities with the experimental powder diagram. But on the other hand: (i) a powder diagram with 14 lines can always be fitted by a triclinic unit cell (6 parameters) with two independent molecules (2 6 3 , Z = 2, with orientational parameters); (ii) structures in P1 both molecules on inversion centres are quite rare, and have not been observed for any quinacridone derivative or any other organic pigment;57 (iii) upon optimisation with the Dreiding force field, the structure transforms to the c structure (which is, strictly speaking, not definite proof against this metastable polymorph, since it has been observed in other cases that two experimentally observed polymorphs collapse into the same minimum upon optimisation, e.g. terephthalic acid43). In our opinion the structure of the aII-phase remains questionable. The red colour, and some similarities between the X-ray powder diagrams of the aII and c phases (especially of cIV) suggest that aII may exhibit a criss-cross pattern; but much more detailed analysis is required.

Scheme 2 Explanation for the enhanced conjugation of the p system of quinacridone in the solid state

8. On the colours of quinacridones


Why are quinacridones reddish to violet in the solid state, but yellow in solution? Quantum mechanical calculations show that the isolated quinacridone molecule should be yellow or orange. Also very diluted solutions of quinacridone show a yellow colour. (If one tries to dissolve larger amounts of quinacridone, one gets an orange or red clear liquid, which contains colloids of solid quinacridone; within a few weeks the colloids aggregate forming an orangered precipitation, and the remaining solution becomes yellow.) The red or violet colours of solid quinacridones are a solid state effect, which is caused by two factors: (i) The formation of intermolecular hydrogen bonds increases the conjugation within the p-system of the molecule (see Scheme 2). In the isolated molecule, the p-systems of the benzene rings are only weakly conjugated via CLO and NH groups. Hence the colour is similar to the yellow colour
140 | CrystEngComm, 2007, 9, 131143

observed for other substituted benzene compounds. In the solid state, hydrogen bridges are formed, and consequently the CLO and NH bonds are weakened and the conjugation between the benzene rings is increased. In principle, the hydrogen atoms could completely move to the neighbouring molecules (Scheme 2, right), resulting in hydroxy-pyridine instead of pyridone moieties; but in the solid state the pyridone tautomer is preferred. Nevertheless the increased conjugation within the p system results in a smaller HOMOLUMO separation. The pp* absorption band shifts from violet (for an isolated molecule) to green (for the crystal). Correspondingly, the observed colour (which is always the complimentary colour of the absorbed light wavelength) shifts from yellow to red. Differences in the strength of the hydrogen bonds and in the hydrogen bond patterns can result in different colours of the quinacridone polymorphs. (ii) The solid-state colour of an organic compound depends not only on the pp* transition energy of a given molecule, but also on the exciton coupling, i.e., on the interaction of the transition dipole moments. This coupling can be positive or negative, depending on the position and spatial orientation of the neighbouring molecules. The coupling works through space to all neighbouring molecules, not only to those connected by hydrogen bonds. In quinacridone the transition dipole is along the short molecular axis. Because of the intermolecular hydrogen bonds, the transition dipoles of neighbouring molecules are aligned in a head-to-tail arrangement causing a large bathochromic shift.58 Since position and orientation of the neighbouring molecules depend on the polymorphic form, the colours of the quinacridone polymorphs are different. A rough estimation is that the effect of the exciton coupling on the colour shift from yellow to red/violet is
This journal is The Royal Society of Chemistry 2007

Scheme 3 Crystal engineering on substituted quinacridones: the chain structure (b-phase) is only stable for X1 = X4 = H. For any other substituent X, the compound must adopt a criss-cross packing like in the c-phase. Correspondingly the colour switches from violet to red shades. Substituents on the positions Y allow for both packing motifs.

N 4,11-Dichloro-quinacridone shows a criss-cross lattice as expected. The structure is not fully isostructural to c-quinacridone, the crystal symmetry being Pbca instead of P21/c.4,40 As a result of the criss-cross pattern, 4,11-dichloro-quinacridone exhibits a bright orange-red shade, even as solid solution with unsubstituted quinacridone (P.R. 207).1 On the other hand substituents at the positions 2, 3, 9, or 10 allow for both the chain and the criss-cross packing motifs. If the chain motif is formed, the colour will be considerably more violet than in the case of a criss-cross packing, for example: N 2,9-Dimethyl-quinacridone (P.R. 122) forms molecular chains4,63 and is isostructural to aI-quinacridone; consequently its colour is considerably more violet than c-quinacridone.64 In addition there exists a second phase of 2,9-dimethyl-quinacridone, which is isostructural to c-quinacridone.65 N 2,9-Dichloro-quinacridone (P.R. 202)4,66 is also isostructural to aI-quinacridone and exhibits a bluish red to violet shade.1,19 Like most quinacridone compounds, also 2,9dichloro-quinacridone is polymorphic. There is a second phase crystallising in P21/c (like b- and c-quinacridone) which, surprisingly, does not exhibit any NHOLC hydrogen bond.67

probably even larger than the effect caused by the formation of hydrogen bonds.59

Conclusions
This work is another example that the combination of crystal structure prediction and Rietveld refinement is a valuable tool to determine crystal structures from low-resolution X-ray powder data.68 The knowledge of the crystal structures is used to perform crystal engineering, i.e., to design new molecular materials having targeted propertiesin the case of quinacridones e.g. to synthesise new, red pigments of industrial importance.

9. Crystal engineering
The knowledge of the crystal structures of aI, b and c quinacridone can be used for crystal engineering:60 As shown in Scheme 3, the chain motif of b-quinacridone does not allow for substituents (except H) at the positions 1, 4, 6, 8, 11, or 13. Any substituent at one or more of these positions would result in negative steric interactions with the neighbouring molecules, the chain motif becomes energetically unfavourable and the compound forms a criss-cross pattern like in c-quinacridone. This was proven experimentally on a series of substituted quinacridones,4 as well as on quinacridone-quinone having CLO groups in positions 6 and 13. c-Quinacridone, forming a criss-cross pattern, is red, whereas aI- and b-quinacridone, having a chain motif, exhibit a dark reddish violet colour. Hence, adding a substituent at the positions 1, 4, 6, 8, 11, or 13 is the best way to find new, red pigments. This is especially true if the electronic effect of the substituent is small in comparison to the effect caused by the packing, e.g., for alkyl or chloro substituents. Even if the H atoms are only partially substituted by other groups, the resulting mixed crystal (solid solution) will form a crisscross pattern. This can be proven experimentally. Examples include: N Pigment Red 209, which is a mixture of 1,10-dichloroquinacridone with its 1,8- and 3,10-isomers, shows a bright red shade (see photo of commercial 1Hostaperm Red EG, Fig. 1). The powder diagram of this pigment is similar to that of c-quinacridone. Also a solid solution containing 10% of this mixture and 90% of unsubstituted quinacridone is isostructural to c-quinacridone.61 In contrast, pure 3,10-dichloro-quinacridone,62 having no substituents at the relevant positions, forms a chain structure, which is isostructural to aI-quinacridone.
This journal is The Royal Society of Chemistry 2007

Acknowledgements
The authors thank C. Buchsbaum (Univ. Frankfurt am Main) for the Rietveld refinement of aI quinacridone. We are grateful to T. Ogawa (Kyoto University) for the TEM micrograph of aI-quinacridone. D. Schnaitmann, W. Schwab and T. Schmiermund (all Clariant, Frankfurt am Main) are acknowledged for their cooperation. Powder diagrams were measured by U. Conrad (Hoechst AG, Frankfurt am Main) in cooperation with B. Mu ller (Hoechst AG, now Sanofi Aventis, Frankfurt am Main), M. Ermrich (X-ray laboratory Dr. Ermrich, Reinheim), and E. Alig (Univ. Frankfurt am Main). Solid state NMR experiments were performed by N. Egger (Hoechst AG, Frankfurt am Main, now Sanofi-Aventis, India). We thank G. Lincke (FH Niederrhein, Krefeld), F. Prokschy (Hoechst AG, now Clariant, Frankfurt am Main) and A. Kroh (Hoechst AG, Frankfurt am Main) for crystallisations. Single crystal diffraction measurements were made by H. Schweitzer, and Rietveld refinements were performed by W. Heyse (both Hoechst AG, now Sanofi-Aventis, Frankfurt am Main)we thank both for their kind cooperation. The authors thank M. R. S. Pinches, N. E. Austin, S. J. Maginn, R. Lovell (all Molecular Simulations Ltd, Cambridge), and H. R. Karfunkel (Ciba-Geigy, Basel) for their contributions to the original crystal structure prediction. Photo images of
CrystEngComm, 2007, 9, 131143 | 141

quinacridone samples were made by L. Fink and E. Alig (both Univ. Frankfurt am Main). Financial support of Clariant GmbH is gratefully acknowledged.

33

References
1 W. Herbst and K. Hunger, Industrial Organic Pigments, 3rd edn, Wiley-VCH, Weinheim, 2004. 2 G. Lincke and H.-U. Finzel, Cryst. Res. Technol., 1996, 31, 441452. 3 G. D. Potts, W. Jones, J. F. Bullock, S. J. Andrewa and S. J. Maginn, J. Chem. Soc., Chem. Commun., 1994, 25652566; J. Mizuguchi, T. Sasaki and K. Tojo, Z. Kristallogr. - New Cryst. Struct., 2002, 217, 249250. 4 E. F. Paulus, E. Dietz, A. Kroh and F. Prokschy, Presentation at the 12th European Crystallographic Meeting, Moscow, 2029 August 1989, Collected Abstracts Vol. 2, pp. 2324. 5 E. Dietz, A. Kroh, E. F. Paulus, F. Prokschy and G. Lincke, 11. Internationales Farbensymposium, Montreux, 23.26. September 1991, published in Chimia 1991, 45, (10), p. 13. 6 F. J. J. Leusen, J. Cryst. Growth, 1996, 166, 900903. 7 A. D. Reidinger and W. S. Struve, US Pat., 2844484, 1955; C. W. Manger and W. S. Struve, US Pat., 2844581, 1955. 8 W. S. Struve, US Pat., 2844485, 1955. 9 E. E. Jaffe, Eur. Pat., 305328, 1989. 10 Toyo, Jap. Pat., Sho5111372, 1976. 11 F. Ba bler and E. E. Jaffe, Eur. Pat., 530143, 1993. 12 E. E. Jaffe, Eur. Pat., 267877, 1988. 13 W. Deuschel, F. Gundel, H. Wuest and E. Daubach, US Pat., 3074950, 1963. 14 K. Hashizume, M. Miyatake, M. Shigemitsu, I. Kumano, H. Katsura and M. Oshima, Ger. Pat., 1805266, 1970. 15 F. Ba bler, Eur. Pat., 1074587, 2001. 16 G. Lincke, J. Mater. Sci., 1997, 32, 64476451. 17 H. R. Schweizer, US Pat., 3272821, 1962. 18 R. S. Tyson and L. Schapiro, Ger. Pat., 2435219, 1975. 19 S. S. Labana and L. L. Labana, Chem. Rev., 1967, 67, 118. 20 T. Ogawa, C. Furukawa, S. Moriguchi, T. Nemoto, S. Isoda and T. Kobayashi, Presentation at the 18th Congress of the International Union of Crystallography, Glasgow, 4th13th August 1999, Book of Abstracts, p. 400401. 21 The samples shown in Fig. 1 have stood on the windowsill in the office of one of the authors (MUS) for many years without any change in their colours or their diffraction patterns. 22 H. Liebermann, H. Kirchhoff, W. Gliksman, L. Loewy, A. Gruhn, T. Hammerich, N. Anitschkoff and B. Schulze, Justus Liebigs Ann. Chem., 1935, 518, 245259. 23 T. B. Reeve and E. C. Botti, Du Pont Official Digest, 1958, 31, 9911002. 24 M. U. Schmidt, T. Schmiermund and M. Bolte, Acta Crystallogr., Sect. C, 2006, 62, m37m40; M. U. Schmidt, T. Schmiermund and M. Bolte, Acta Crystallogr., Sect. C, 2006, DOI: 10.1107/ S1600536806050781; M. U. Schmidt, T. Schmiermund and M. Bolte, Acta Crystallogr., Sect. E, submitted. 25 B. Stengel-Rutkowski and M. Bo hmer, Farbe und Lack, 3/2002, 108, 3035. 26 M. Urban, D. Schnaitmann and M. Bo hmer, Eur. Pat., 799862, 1996. 27 O. Fuchs and A. Kirsch, Ger. Pat., 115046, 1959. 28 M. Urban, M. Bo hmer, D. Schnaitmann and J. Weber, Eur. Pat., 971001, 1998. 29 D. Thetford, J. Cherryman, A. P. Chorlton and R. Docherty, Dyes Pigm., 2004, 63, 259276. 30 This nice experiment was repeatedly shown on open days at Hoechst and Clariant, using a microscope equipped with a video camera. 31 G. Lincke, Dyes Pigm., 1999 (Volume Date 2000), 44, 101122. 32 Prof. Dr. G. Lincke, FH Niederrhein, Krefeld, Germany (retired). The crystals were reported to be grown by sublimation. Note added in proof: Recently, Mizuguchi et al. reported, that crystals of b-quinacridone can be grown by sublimation at 200250 uC, whereas at 400 uC the c phase is formed. They also determined the crystal structure of the b phase, coming to very similar results as we

34 35 36 37

38 39 40 41 42

43

44

45

46

47

48

49 50

did: ; N. Nishimura, T. Senju and J. Mizuguchi, Acta Crystallogr., Sect E, 2006, 62, o4683o4685. G. M. Sheldrick, SHELXTL, an integrated system for solving, refining and displaying crystal structures from diffraction data. Universita t Go ttingen, 1983. E. Keller, SCHAKAL99. Kristallographisches Institut der Universita t, Freiburg, 1999. S. L. Mayo, B. D. Olafson and W. A. Goddard, III, J. Phys. Chem., 1990, 94, 88978909. J. Gasteiger and M. Marsili, Tetrahedron, 1980, 36, 321922. Absorption spectra of quinacridone powders finely dispersed in PVC are available from the corresponding author (MUS) upon request. H. Koyama, H. J. Scheel and F. Laves, Naturwissenschaften, 1966, 53, 700. Y. Nagai and H. Nishi, Senryo to Yakuhin, 1968, 13, 81108, p. 90 (in Japanese). F. H. Chung and R. W. Scott, J. Appl. Crystallogr., 1971, 4, 506511. G. Lincke, personal communication, 1985. E.g.: F. J. J. Leusen, Presentation at the 213th ACS National Meeting, San Francisco, April 1317, 1997. Abstract in the Book of Abstracts, COMP-024; M. C. Wahle, O. Ko nig, G. E. Engel and F. J. J. Leusen, Presentation at the 216th ACS National Meeting, Boston, August 2327, 1998, Abstract in the Book of Abstracts, COMP-180. F. J. J. Leusen, Presentation at the 15th European Crystallographic Meeting ECM-15, Dresden, 28 August2 September 1994, Abstract in Z. Kristallogr., 1994, 8, 161; F. J. J. Leusen, Presentation at the Conference Kristallisation fu r die Strukturanalyse und Polymorphie, Bern, 1720 September 1995. N. Panina, F. Leusen, R. van de Ven, F. Janssen, P. Verwer, H. Meekes, E. Vlieg and G. Deroover, Presentation at the 23rd European Crystallographic Meeting ECM-23, Leuven, 611 August 2006, published in Acta Crystallogr., 2006, A62, p. s79. F. J. J. Leusen, M. R. S. Pinches, N. E. Austin, S. J. Maginn, R. Lovell, H. R. Karfunkel and E. F. Paulus, unpublished manuscript (Rejected by Science, 1995). M. U. Schmidt, Kristallstrukturberechnungen metallorganischer Moleku lverbindungen (Crystal structure calculations of organometallic molecular compounds), Verlag Shaker, Aachen, 1995; P. Verwer and F. J. J. Leusen, in: Reviews in Computational Chemistry, ed. Lipkowitz and Boyd, Wiley-VCH, New York, vol. 12, 327365, 1998; J. P. M. Lommerse, W. D. S. Motherwell, H. L. Ammon, J. D. Dunitz, A. Gavezzotti, D. W. M. Hofmann, F. J. J. Leusen, W. T. M. Mooij, S. L. Price, B. Schweizer, M. U. Schmidt, B. P. van Eijck, P. Verwer and D. E. Williams, Acta Crystallogr., Sect. B, 2000, 56, 697714; W. D. S. Motherwell, H. L. Ammon, J. D. Dunitz, A. Dzyabchenko, P. Erk, A. Gavezzotti, D. W. M. Hofmann, F. J. J. Leusen, J. P. M. Lommerse, W. T. M. Mooij, S. L. Price, H. Scheraga, B. Schweizer, M. U. Schmidt, B. P. van Eijck, P. Verwer and D. E. Williams, Acta Crystallogr., Sect. B, 2002, 58, 647661; G. M. Day, W. D. S. Motherwell, H. L. Ammon, S. X. M. Boerrigter, R. G. Della Valle, E. Venuti, A. Dzyabchenko, J. D. Dunitz, B. Schweizer, B. P. van Eijck, P. Erk, J. C. Facelli, V. E. Bazterra, M. B. Ferraro, D. W. M. Hofmann, F. J. J. Leusen, C. Liang, C. C. Pantelides, P. G. Karamertzanis, S. L. Price, T. C. Lewis, H. Nowell, A. Torrisi, H. A. Scheraga, Y. A. Arnautova, M. U. Schmidt and P. Verwer, Acta Crystallogr., Sect. B, 2005, 61, 511527. M. J. Frisch, G. W. Trucks, M. Head-Gordon, P. M. W. Gill, M. W. Wong, J. B. Foresman, B. G. Johnson, H. B. Schlegel, M. A. Robb, E. S. Replogle, R. Gomperts, J. L. Andres, K. Raghavachari, J. S. Binkley, C. Gonzalez, R. L. Martin, D. J. Fox, D. J. Defrees, J. Baker, J. J. P. Stewart and J. A. Pople, Gaussian92, Gaussian, Inc., Pittsburgh, 1992. D. E. Williams, in Reviews in Computational Chemistry, ed. K. B. Lipkowitz and D. B. Boyd, Verlag Chemie, New York, 1991, vol. 2, p. 219. P. P. Ewald, Ann. Physik, 1921, 64, 253; D. E. Williams, Top. Curr. Phys., 1981, 26, 3. Program package Cerius,2 Molecular Simulations, Inc., now Accelrys Ltd., 334 Cambridge Science Park, Cambridge CB4 0WN, UK.

142 | CrystEngComm, 2007, 9, 131143

This journal is The Royal Society of Chemistry 2007

51 R. J. Gdanitz, Chem. Phys. Lett., 1992, 190, 391; H. R. Karfunkel and R. J. Gdanitz, J. Comput. Chem., 1992, 13, 1171; H. R. Karfunkel and F. J. J. Leusen, Speedup, 1993, 6(Nr. 2), 4350; H. R. Karfunkel, B. Rohde, F. J. J. Leusen, R. J. Gdanitz and G. Rihs, J. Comput. Chem., 1993, 14, 1125; H. R. Karfunkel, F. J. J. Leusen and R. J. Gdanitz, J. Comput. Aided Mat. Des., 1993, 1, 177185. 52 W. H. Baur and D. Kassner, Acta Crystallogr., Sect. B, 1992, 48, 356. 53 H. M. Rietveld, Acta Cryst., 1967, 22, 151152; H. M. Rietveld, J. Appl. Crystallogr., 1969, 2, 6571. 54 A. C. Larson and R. B. von Dreele, General Structure Analysis System (GSAS), Los Alamos National Laboratory Report LAUR 1994, 86748. 55 P. Thompson, D. E. Cox and J. B. Hastings, J. Appl. Crystallogr., 1987, 20, 7983. 56 L. W. Finger, D. E. Cox and A. P. Jephcoat, J. Appl. Crystallogr., 1994, 27, 892900. 57 Cambridge Structural Database CSD, Cambridge Crystallographic Data Centre, Cambridge, UK, 2006. 58 J. Mizuguchi, A. Endo and S. Matsumoto, Nippon Gazo Gakkaishi, 2000, 39, 94102. 59 P. Erk, personal communication. 60 M. U. Schmidt, Adv. Colour Sci. Technol., 2003, 6, 5961.

61 M. Urban, M. Bo hmer, J. Weber, D. Schnaitmann and M. Haberlick, Eur. Pat., 1020497, 2000. 62 T. Senju, T. Hoki and J. Mizuguchi, Acta Crystallogr., Sect. E, 2006, 62, o261263. 63 G. Lincke, Chem. Zeit., 1985, 109, 8996; J. Mizuguchi, T. Senju and M. Sakai, Z. Kristallogr., 2002, 217, 525526. 64 Clariant GmbH, Organic pigments for the paint industry, Technical information brochure, Frankfurt am Main, 2001. 65 Y. Otaka, Nippon Kagaku Kaishi, 1975, 1838. 66 T. Senju, N. Nishimura, T. Hoki and J. Mizuguchi, Acta Crystallogr., Sect. E, 2005, 61, o2596o2698. 67 T. Senju, T. Hoki and J. Mizuguchi, Acta Crystallogr., Sect. E, 2005, 61, o1061o1063. 68 For other examples see, e.g.: M. U. Schmidt and R. E. Dinnebier, J. Appl. Crystallogr., 1999, 32, 178186; M. U. Schmidt, in Crystal Engineering: From Molecules and Crystals to Materials, ed. D. Braga, F. Grepioni and A. G. Orpen, Kluwer Academic Publishers, Dordrecht, 1999, 331348; M. U. Schmidt, M. Ermrich and R. E. Dinnebier, Acta Crystallogr., Sect. B, 2005, 61, 3745; M. U. Schmidt, D. W. M. Hofmann, C. Buchsbaum and H. J. Metz, Angew. Chem. , 2006, 118 , 13351340; M. U. Schmidt, D. W. M. Hofmann, C. Buchsbaum and H. J. Metz, Angew. Chem., Int. Ed., 2006, 45, 13131317.

This journal is The Royal Society of Chemistry 2007

CrystEngComm, 2007, 9, 131143 | 143

You might also like