K₂[HCr₂AsO₁₀]: redetermination of phase II and the predicted structure of phase I

Our prediction that phase II of dipotassium hydrogen chromatoarsenate, K₂[HCr₂AsO₁₀], is ferroelectric, based on the analysis of the atomic coordinates by Averbuch-Pouchot, Durif & Guitel [Acta Cryst. (1978), B34, 3725-3727], led to an independent redetermination of the structure using two separate crystals. The resulting improved accuracy allows the inference that the H atom is located in the hydrogen bonds of length 2.555 (5) Å which form between the terminal O atoms of shared AsO₃OH tetrahedra in adjacent HCr₂AsO₁₀^2- ions. The largest atomic displacement of 0.586 Å between phase II and the predicted paraelectric phase I is by these two O atoms. The H atoms form helices of radius ~0.60 Å about the 3₁ or 3₂ axes. Normal probability analysis reveals systematic error in seven or more of the earlier atomic coordinates.

S1. Comment

The possibility that the potential barrier between the structure of K₂HCr₂AsO₁₀ phase II (II—K₂Cr₂O₇; Averbuch-Pouchot et al., 1978; hereinafter A—P) and that of paraelectric phase I is surmountable below the Curie temperature led to the prediction that phase II is ferroelectric (Abrahams, 2003). Crystals grown for associated physical property measurements were also used for an independent structural redetermination on two different crystals. Unsurprisingly, crystal 1 in space group P3₁ has a small enantiomorphic component, whereas crystal 2 in space group P3₂ is single-component (see the Refinement section for further discussion of this). \sch

The asymmetric unit of phase II contains one independent AsO₃OH and two CrO₄ tetrahedra, sharing O atoms to form individual HAsCr₂O₁₀²⁻ ions (Fig. 1a). The average Cr—O distance for the terminal O atoms in both crystals is 1.600 (4) Å (Table 1) and this does not differ significantly from the value of 1.612 (7) Å in II—K₂Cr₂O₇ (Weakley et al., 2004), whereas the mean bridging Cr—O distance of 1.844 (4) Å in K₂HCr₂AsO₁₀ very significantly exceeds the value of 1.785 (4) Å in II—K₂Cr₂O₇.

Unlike the bond-length distribution in the CrO₄ tetrahedra, the mean As—O distance of 1.632 (3) Å for the terminal O5 atom is significantly less than the mean As—O distance of 1.699 (9) Å for the two bridging atoms and the terminal O6 atom in the AsO₃OH tetrahedron. The short inter-anionic contact O5···O6ⁱ [2.550 (11) and 2.555 (5) Å for crystals 1 and 2, respectively; symmetry code: (i) y-x, −x, z − 1/3 for crystal 1, −y, x-y, z − 1/2 for crystal 2] results from the formation of an O5···H···O6ⁱ bond between HAsCr₂O₁₀²⁻ ions. The arsenate distances given in Supplementary Table S1 show that the mean bridging and terminal As—O distances generally do not differ significantly, at 1.708 (20) and 1.682 (12) Å, and so the mean As—O distance may be taken as 1.69 (2) Å.

The bond-valence (BV) sums (Brown & Altermatt, 1985) for each atom in K₂HCr₂AsO₁₀ are 5.03 (3) for As, 5.98 (4) for Cr1 and 5.96 (4) for Cr2, 1.327 (8) for K1 and 1.157 (6) for K2, 1.612 (14) for O5 and 1.400 (14) for O6, and 1.94–2.20 (3) for all other O atoms in crystal 2, with comparable values in crystal 1. The connectivity within the asymmetric unit is shown in Fig. 1(b), and the labelling for all atoms in the unit cell is given in Supplementary Fig. S1. The BV sum for As agrees with the expectated value, while those for Cr, K and O atoms other than O5 and O6 agree with the values observed in II—K₂Cr₂O₇ (Weakley et al., 2004). The deficit of 0.99 in the joint valence of atoms O5 and O6 is clearly equivalent to an H atom which, while not located directly, results in the bond revealed in both crystals by the short O5···O6 distance. The H atoms form helices of radius ~0.60 Å about the 3₁ or 3₂ axes (Fig. 2), as they link HCr₂AsO₁₀²⁻ anions. H-atom location was not considered in the report by A—P; if taken as midway between O5 and O6ⁱ (Table 2), then the H atom in crystal 2 is at (0.0555, 0.0307, −0.0245). Since the BV sum for As—O6 is 1.23 (1), this bond is close to single, hence the H atom is most likely to be nearer O6.

Atomic coordinates (x',y',z') in the predicted supergroup P3₁21 or P3₂21 are derived from the coordinates of related pairs of atoms in P3₁ or P3₂ under the higher-symmetry constraint (Table 2). Maximum differences between the positions of independent atoms in phase II and those predicted in paraelectric phase I, following the necessary symmetry conversions, are 0.41 Å for H and 0.586 Å for O5 and O6, confirming the basis for the original prediction of ferroelectricity (Abrahams, 2003). [Atoms O9 and O10 in the designation of Averbuch-Puchot et al. (1978) are equivalent to atoms O5 and O6 in Table 2.] The predicted atomic arrangement in phase I is shown in Supplementary Figs. S2 and S3.

The formation of K₂[CrO₃AsO₃OHCrO₃] from its constituent ions (see Experimental) is possible only by fission of an O—Cr—O bond in the Cr₂O₇²⁻ anion and the presence of an AsO₃OH²⁻ ion. The two resulting As—O—Cr bonds form by elimination of O²⁻ as H₂O. The protonation of a terminal O atom on As results in chains of CrO₃AsO₃OHCrO₃²⁻ ions linked through hydrogen bonds, as shown in Figs. 1(a) and 2. The potential barrier to these rearrangements appears to be surmounted only at aqueous reaction solution temperatures close to boiling.

Comparison of the three independent sets of atomic coordinates is made by plotting the ordered experimental quantiles Q_exp = |ξ_i(1) - ξ_i(2)|/{σ²[ξ_i(1)] + σ²[ξ_i(2)]²}^1/2 against the corresponding normal quantiles Q_norm, where ξ_i(1), ξ_i(2) are the ith parameters from the first and second independent determinations with the same setting, and σ[ξ_i(1)], σ[ξ_i(2)] are the corresponding standard uncertainties (s.u.s) of each parameter. In the absence of error, a linear array of unit slope and zero intercept results (Abrahams & Keve, 1971). The Q_norm magnitudes are conveniently calculated by the program NORMPA (Ross, 2003).

The 45 atomic-coordinate magnitudes determined with crystals 1 and 2 are compared in Fig. 3, the straight line giving the fit obtained by linear regression. The absence of outliers and the small departure of the slope from unity or linearity is indicative of minor systematic error in either coordinate set, but with joint s.u.s (j.s.u.s) that are underestimated by a factor of ~1.3. The corresponding normal probability comparison of each set with the A—P atomic coordinates (Supplementary Figs. S4 and S5) contrasts strongly with that in Fig. 3. Seven Q_exp - Q_norm terms in each of the latter two cases depart strongly from linearity (Table 3), hence these may be rejected as outliers, since major departures from a normal distribution can be due only to major systematic error. The strong possibility of uncompensated enantiomorphous twinning giving rise to the strong parameter correlation in the least-squares refinement noted by A—P is one of several likely sources of error. The remaining 38 terms exhibit a slightly S-shaped distribution, with j.s.u.s underestimated by a factor of ~1.7 in the comparison of crystal 1 and A—P, and of ~2.0 for the comparison of crystal 2 and A—P. The correlation coefficient in all fits made by linear regression is 0.987–0.992. The experimental uncertainties should be corrected by factors not less than those derived from the magnitude of the slopes listed in Table 3; inclusion of outliers would clearly increase these factors.

The recommended phase-transition nomenclature (Tolédano et al., 1998, 2001) for phases I and II of K₂HCr₂AsO₁₀, using the thermal values of Ylvisaker et al. (2001), is I|~590–540 K| P3₂21 (152) |Z = 3| non-ferroic | decomposes above ~590 K, II| 540–270 K| P3₂ (145) |Z = 3| ferroelectric| 2 variants, lower thermal limit not known.

Table 1. Selected distances (Å) in K₂HCr₂AsO₁₀ phase II.

Table 2. Atomic coordinates of K₂HCr₂AsO₁₀ in phase II and predicted coordinates in phase I, with component Δ(x) and total Δ(xyz) differences between the phases in Å.

Table 3. Linear-regression indicators for the K₂HCr₂AsO₁₀ Q_exp - Q_norm plots.

S2. Experimental

Averbuch-Pouchot et al. (1978) prepared the title chromatoarsenate by the reaction K₂Cr₂O₇ + H₃AsO₄ → K₂HCr₂AsO₁₀ + H₂O. Substitution of either 1/2[As₂O₅.xH₂O] with x ≈ 3 or KH₂AsO₄ for H₃AsO₄ also yields K₂HCr₂AsO₁₀ (CAS Registry No. 69107–38-6). Initial crystal growth in either reaction using 99.9% pure starting products with 0.1 mol of each reactant dissolved in about 10 mol H₂O, on heating to boiling for several minutes and then cooling naturally to room temperature, commonly gives two crops of strong reddish-orange [ISCC-NBS Color (Supplement to NBS Circular 553). A complete list with RGB values is given at https://swiss.csail.mit.edu/jaffer/Color/nbs-iscc-rgb.pdf Link broken and also at https://tx4.us/nbs-iscc.htm] crystals (Weakley et al., 2004). Two crystals from first-growth crops were selected for the present study. The solubility at 298 K was determined as 207 g l⁻¹.

S3. Refinement

The general structure solution made use of a SIR92 electron-density map (Altomare et al., 1994). The Flack (1983) parameter shows that the space group for crystal 1 is primarily P3₁ but with a small enantiomorphic component, whereas the space group for crystal 2 does not differ significantly from P3₂. The H atom could not be located directly with confidence in either crystal but was later inferred (see Comment).

Computing details

For both compounds, data collection: CAD-4-PC Software (Enraf-Nonius, 1993); cell refinement: CAD-4-PC Software; data reduction: TEXSAN (Molecular Structure Corporation, 1997); program(s) used to solve structure: TEXSAN; program(s) used to refine structure: SHELXL97 (Sheldrick, 1997); molecular graphics: ATOMS (Dowty, 2003).

Figures

Fig. 1(a). The K2HCr2AsO10 structure at 298 K. The O6—H···O5 hydrogen bonds are shown as dashed lines. The labelling of all atoms in the unit cell is given in Supplementary Fig. S1. Fig. 1(b). The HCr2AsO10^2− ion of crystal 1, showing the interionic hydrogen bond, with displacement ellipsoids drawn at the 50% probability level. The H atom is shown as a sphere and the hydrogen bond as single lines. [Symmetry code: (i) y-x, −x, z − 1/3]. Fig. 2. The K2HCr2AsO10 structure at 298 K, with H atoms shown as small spheres and K atoms as large spheres. The AsO4 tetrahedra are plus-hatched and the CrO4 tetrahedra are cross-hatched. Fig. 3. A normal probability Qexp - Qnorm plot for the atomic coordinates determined for K2HCr2AsO10, with data for crystal 1 plotted versus those for crystal 2.

(I) dipotassium hydrogen chromatoarsenate

Crystal data

K2[HCr2AsO10]	Dx = 2.779 Mg m^−3
Mr = 418.13	Mo Kα radiation, λ = 0.71073 Å
Trigonal, P31	Cell parameters from 25 reflections
Hall symbol: P 31	θ = 14.1–16.4°
a = 7.6931 (8) Å	µ = 6.33 mm^−1
c = 14.623 (3) Å	T = 296 K
V = 749.50 (19) Å^3	Hexagonal prism, intense red-orange
Z = 3	0.40 × 0.22 × 0.18 mm
F(000) = 600

Data collection

Enraf-Nonius CAD-4 diffractometer	2040 reflections with I > 2σ(I)
Radiation source: fine-focus sealed tube	Rint = 0.019
Graphite monochromator	θmax = 30.0°, θmin = 3.1°
ω/2θ scans	h = −8→10
Absorption correction: analytical (de Meulenaer & Tompa, 1965)	k = 0→10
Tmin = 0.205, Tmax = 0.320	l = −20→20
2063 measured reflections	3 standard reflections every 300 reflections
2051 independent reflections	intensity decay: 0.0%

Refinement

Refinement on F^2	Hydrogen site location: inferred from neighbouring sites
Least-squares matrix: full	H atoms treated by a mixture of independent and constrained refinement
R[F^2 > 2σ(F^2)] = 0.050	w = 1/[σ^2(Fo^2) + (0.0086P)^2 + 7.6027P] where P = (Fo^2 + 2Fc^2)/3
wR(F^2) = 0.129	(Δ/σ)max = 0.005
S = 1.29	Δρmax = 1.31 e Å^−3
2051 reflections	Δρmin = −1.08 e Å^−3
137 parameters	Extinction correction: Zachariasen (1967), Acta Cryst. (1968). A24, p. 213, eq. (3)
1 restraint	Extinction coefficient: 0.0000072 (11)
Primary atom site location: structure-invariant direct methods	Absolute structure: Flack (1983), with how many Friedel pairs
Secondary atom site location: difference Fourier map	Absolute structure parameter: 0.17 (3)

Special details

Experimental. The correct polarity in Crystal 1 was ascertained by refinement in space-groups P3₁ and P3₂. It was confirmed by refinement of a Flack parameter; the significantly non-zero value is ascribed to a minor twin component.

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 e.s.d.'s involving l.s. planes.

Refinement. Refinement of F² against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F², conventional R-factors R are based on F, with F set to zero for negative F². The threshold expression of F² > 2σ(F²) is used only for calculating R-factors (gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F² are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
As1	0.10826 (13)	−0.02094 (14)	0.49989 (8)	0.01945 (18)
Cr1	0.0041 (2)	−0.4249 (2)	0.60617 (10)	0.0200 (3)
Cr2	0.4201 (2)	0.4191 (2)	0.43023 (10)	0.0201 (3)
O1	−0.1383 (15)	−0.3891 (14)	0.6749 (7)	0.039 (2)
O2	−0.1214 (19)	−0.5908 (15)	0.5301 (7)	0.055 (3)
O3	0.1440 (15)	−0.4780 (16)	0.6664 (8)	0.046 (2)
O4	0.1691 (11)	−0.1884 (12)	0.5449 (6)	0.0289 (16)
O5	−0.0079 (14)	−0.0971 (13)	0.4020 (5)	0.0361 (19)
O6	−0.0392 (14)	0.0180 (15)	0.5724 (5)	0.0349 (18)
O7	0.3389 (12)	0.1870 (11)	0.4962 (6)	0.0310 (17)
O8	0.2329 (16)	0.4022 (16)	0.3750 (8)	0.046 (2)
O9	0.5089 (19)	0.5990 (14)	0.5027 (6)	0.047 (3)
O10	0.5860 (14)	0.4415 (15)	0.3572 (6)	0.0375 (19)
K1	−0.0199 (4)	−0.4513 (3)	0.34633 (17)	0.0288 (4)
K2	0.6027 (4)	−0.0061 (4)	0.51916 (17)	0.0292 (4)
H	0.0256	−0.0279	0.3206	0.050*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
As1	0.0201 (4)	0.0202 (4)	0.0171 (3)	0.0094 (4)	−0.0003 (3)	0.0027 (3)
Cr1	0.0197 (7)	0.0170 (6)	0.0210 (7)	0.0075 (6)	0.0006 (5)	0.0047 (5)
Cr2	0.0223 (7)	0.0167 (6)	0.0211 (6)	0.0096 (6)	0.0005 (5)	0.0011 (5)
O1	0.040 (5)	0.035 (5)	0.048 (5)	0.023 (4)	0.016 (4)	0.012 (4)
O2	0.072 (8)	0.030 (5)	0.031 (4)	0.002 (5)	−0.004 (4)	−0.005 (3)
O3	0.034 (5)	0.046 (6)	0.060 (6)	0.022 (4)	−0.002 (4)	0.020 (5)
O4	0.022 (3)	0.030 (4)	0.038 (4)	0.015 (3)	0.007 (3)	0.016 (3)
O5	0.043 (5)	0.039 (5)	0.013 (3)	0.010 (4)	−0.007 (3)	−0.002 (3)
O6	0.050 (5)	0.049 (5)	0.021 (4)	0.036 (5)	−0.004 (3)	−0.007 (3)
O7	0.020 (3)	0.017 (3)	0.042 (4)	−0.002 (3)	−0.006 (3)	0.011 (3)
O8	0.045 (5)	0.048 (6)	0.052 (6)	0.029 (5)	−0.003 (4)	0.008 (4)
O9	0.082 (8)	0.028 (4)	0.025 (4)	0.024 (5)	0.002 (4)	−0.007 (3)
O10	0.037 (5)	0.047 (5)	0.030 (4)	0.023 (4)	0.005 (3)	0.005 (3)
K1	0.0296 (11)	0.0286 (11)	0.0291 (10)	0.0153 (9)	0.0029 (8)	0.0010 (8)
K2	0.0298 (11)	0.0278 (11)	0.0278 (10)	0.0128 (9)	−0.0017 (9)	−0.0010 (8)

Geometric parameters (Å, º)

As1—O5	1.633 (7)	K1—O3^ii	2.786 (14)
As1—O6	1.685 (8)	K1—O5	2.801 (10)
As1—O7	1.695 (7)	K1—O7^iii	3.139 (10)
As1—O4	1.707 (7)	K1—O8^iv	2.724 (14)
Cr1—O3	1.594 (9)	K1—O9^iii	2.826 (11)
Cr1—O2	1.602 (10)	K1—O10^v	2.720 (12)
Cr1—O1	1.609 (9)	K2—O1^ii	2.828 (11)
Cr1—O4	1.847 (7)	K2—O2^vi	2.821 (11)
Cr2—O8	1.599 (10)	K2—O3^ii	3.112 (13)
Cr2—O9	1.600 (9)	K2—O4	2.925 (10)
Cr2—O10	1.606 (9)	K2—O6^vii	2.778 (12)
Cr2—O7	1.842 (7)	K2—O7	3.075 (10)
O5—H	1.28	K2—O8^viii	3.030 (13)
K1—O1^i	2.742 (10)	K2—O9^iv	2.760 (10)
K1—O2	2.854 (11)	K2—O10^viii	2.852 (9)
O5···O6^iii	2.550 (11)
O5—As1—O6	108.6 (5)	O1—Cr1—O4	109.0 (4)
O5—As1—O7	115.9 (4)	O8—Cr2—O9	112.5 (6)
O6—As1—O7	109.4 (5)	O8—Cr2—O10	107.9 (5)
O5—As1—O4	112.0 (5)	O9—Cr2—O10	112.0 (6)
O6—As1—O4	110.7 (4)	O8—Cr2—O7	109.3 (5)
O7—As1—O4	99.9 (4)	O9—Cr2—O7	106.4 (4)
O3—Cr1—O2	113.2 (7)	O10—Cr2—O7	108.6 (5)
O3—Cr1—O1	107.7 (6)	As1—O4—Cr1	128.1 (4)
O2—Cr1—O1	112.3 (6)	As1—O5—H	135.1
O3—Cr1—O4	107.5 (4)	As1—O7—Cr2	128.3 (5)
O2—Cr1—O4	107.0 (5)

Symmetry codes: (i) −x+y, −x−1, z−1/3; (ii) −x+y+1, −x, z−1/3; (iii) −x+y, −x, z−1/3; (iv) x, y−1, z; (v) x−1, y−1, z; (vi) x+1, y+1, z; (vii) x+1, y, z; (viii) −y+1, x−y, z+1/3.

(II) dipotassium hydrogen chromatoarsenate

Crystal data

K2[HCr2AsO10]	Dx = 2.778 Mg m^−3
Mr = 418.13	Mo Kα radiation, λ = 0.71073 Å
Trigonal, P32	Cell parameters from 25 reflections
Hall symbol: P 32	θ = 13.5–14.9°
a = 7.6963 (9) Å	µ = 6.33 mm^−1
c = 14.6171 (11) Å	T = 294 K
V = 749.82 (14) Å^3	Hexagonal prism, intense-orange
Z = 3	0.29 × 0.13 × 0.11 mm
F(000) = 600

Data collection

Enraf-Nonius CAD-4 diffractometer	Rint = 0.011
Radiation source: fine-focus sealed tube	θmax = 27.5°, θmin = 3.1°
Graphite monochromator	h = −8→8
ω/2θ scans	k = 0→9
1229 measured reflections	l = −18→18
1224 independent reflections	3 standard reflections every 300 reflections
1182 reflections with I > 2σ(I)	intensity decay: 0.5%

Refinement

Refinement on F^2	Hydrogen site location: inferred from neighbouring sites
Least-squares matrix: full	H atoms treated by a mixture of independent and constrained refinement
R[F^2 > 2σ(F^2)] = 0.021	w = 1/[σ^2(Fo^2) + (0.0181P)^2 + 0.0502P] where P = (Fo^2 + 2Fc^2)/3
wR(F^2) = 0.053	(Δ/σ)max = 0.008
S = 1.08	Δρmax = 0.46 e Å^−3
1224 reflections	Δρmin = −0.33 e Å^−3
137 parameters	Extinction correction: Zachariasen (1967), Acta Cryst. (1968). A24, p. 213, eq. (3)
1 restraint	Extinction coefficient: 0.0000096 (12)
Primary atom site location: structure-invariant direct methods	Absolute structure: Flack (1983), with how many Friedel pairs
Secondary atom site location: difference Fourier map	Absolute structure parameter: 0.030 (16)

Special details

Experimental. The correct polarity in Crystal 2 was ascertained by refinement in both space-group P3₁ and P3₂. It was confirmed by refinement of a Flack parameter; the significantly-zero value validates the absence of any enantiomorphic component.

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 e.s.d.'s involving l.s. planes.

Refinement. Refinement of F² against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F², conventional R-factors R are based on F, with F set to zero for negative F². The threshold expression of F² > 2σ(F²) is used only for calculating R-factors (gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F² are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
As1	0.12958 (6)	0.02179 (6)	0.50004 (5)	0.01795 (11)
Cr1	0.00117 (11)	−0.41910 (11)	0.43038 (5)	0.01938 (16)
Cr2	0.42899 (11)	0.42481 (11)	0.60655 (5)	0.01996 (16)
O1	0.1424 (6)	−0.4423 (7)	0.3570 (3)	0.0379 (10)
O2	−0.0883 (8)	−0.5988 (6)	0.5024 (3)	0.0491 (12)
O3	−0.1692 (6)	−0.4012 (7)	0.3758 (3)	0.0421 (10)
O4	0.1520 (6)	−0.1879 (6)	0.4966 (3)	0.0318 (8)
O5	0.0914 (6)	0.1001 (5)	0.4024 (2)	0.0309 (8)
O6	−0.0582 (5)	−0.0195 (7)	0.5728 (2)	0.0330 (9)
O7	0.3559 (5)	0.1869 (5)	0.5468 (3)	0.0258 (7)
O8	0.6242 (6)	0.4793 (7)	0.6657 (4)	0.0465 (12)
O9	0.4665 (9)	0.5909 (6)	0.5314 (3)	0.0515 (13)
O10	0.2526 (7)	0.3923 (7)	0.6749 (3)	0.0388 (9)
K1	0.43080 (16)	0.45187 (16)	0.34636 (8)	0.0280 (2)
K2	0.60914 (17)	0.00606 (17)	0.51907 (8)	0.0285 (2)
H	0.0256	−0.0279	0.3206	0.050*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
As1	0.0195 (2)	0.0160 (2)	0.01738 (19)	0.00821 (18)	−0.00247 (17)	−0.00145 (17)
Cr1	0.0200 (3)	0.0152 (3)	0.0205 (4)	0.0070 (3)	−0.0008 (3)	−0.0013 (3)
Cr2	0.0219 (3)	0.0160 (3)	0.0216 (4)	0.0092 (3)	−0.0049 (3)	−0.0050 (3)
O1	0.039 (2)	0.048 (2)	0.036 (2)	0.029 (2)	0.0047 (18)	−0.0059 (18)
O2	0.071 (3)	0.0216 (19)	0.030 (2)	0.004 (2)	0.001 (2)	0.0050 (17)
O3	0.031 (2)	0.046 (3)	0.051 (3)	0.021 (2)	−0.0144 (19)	−0.009 (2)
O4	0.036 (2)	0.0216 (16)	0.040 (2)	0.0164 (16)	−0.0169 (17)	−0.0068 (15)
O5	0.042 (2)	0.0284 (19)	0.0218 (17)	0.0175 (17)	−0.0046 (15)	0.0024 (14)
O6	0.0225 (17)	0.052 (3)	0.0227 (18)	0.0175 (17)	0.0042 (14)	0.0040 (17)
O7	0.0196 (16)	0.0216 (16)	0.0343 (19)	0.0089 (13)	−0.0048 (14)	−0.0153 (14)
O8	0.031 (2)	0.050 (3)	0.059 (3)	0.021 (2)	−0.023 (2)	−0.028 (2)
O9	0.086 (4)	0.026 (2)	0.038 (2)	0.024 (2)	−0.001 (2)	0.0030 (18)
O10	0.039 (2)	0.042 (2)	0.041 (2)	0.0248 (19)	0.0033 (18)	−0.0111 (18)
K1	0.0255 (5)	0.0269 (5)	0.0315 (6)	0.0131 (4)	0.0015 (4)	−0.0013 (5)
K2	0.0296 (5)	0.0258 (5)	0.0287 (5)	0.0130 (4)	−0.0002 (5)	0.0019 (4)

Geometric parameters (Å, º)

As1—O5	1.633 (4)	K1—O4^ii	3.129 (5)
As1—O6	1.691 (3)	K1—O5	2.784 (4)
As1—O7	1.703 (3)	K1—O7	3.450 (4)
As1—O4	1.708 (4)	K1—O8^iv	2.782 (6)
Cr1—O2	1.595 (4)	K1—O9	2.871 (5)
Cr1—O3	1.599 (4)	K1—O10^v	2.731 (7)
Cr1—O1	1.599 (4)	K2—O1^vi	2.845 (5)
Cr1—O4	1.840 (4)	K2—O2^iii	2.766 (4)
Cr2—O8	1.597 (4)	K2—O3^vi	3.042 (5)
Cr2—O9	1.599 (4)	K2—O4	3.076 (5)
Cr2—O10	1.601 (4)	K2—O6^vii	2.777 (5)
Cr2—O7	1.845 (3)	K2—O7	2.935 (4)
O5—H	1.4683	K2—O8^iv	3.127 (6)
K1—O1^i	2.724 (5)	K2—O9^viii	2.818 (4)
K1—O2^ii	2.830 (5)	K2—O10^iv	2.825 (5)
K1—O3^iii	2.731 (5)
O5···O6^ii	2.555 (5)
O5—As1—O6	108.9 (2)	O1—Cr1—O4	109.0 (2)
O5—As1—O7	112.40 (19)	O8—Cr2—O9	113.1 (3)
O6—As1—O7	110.30 (18)	O8—Cr2—O10	108.4 (2)
O5—As1—O4	116.5 (2)	O9—Cr2—O10	110.7 (3)
O6—As1—O4	108.8 (2)	O8—Cr2—O7	107.3 (2)
O7—As1—O4	99.66 (16)	O9—Cr2—O7	108.2 (2)
O2—Cr1—O3	112.8 (3)	O10—Cr2—O7	109.1 (2)
O2—Cr1—O1	111.6 (3)	As1—O4—Cr1	128.0 (2)
O3—Cr1—O1	107.9 (2)	As1—O5—H	122.0
O2—Cr1—O4	106.3 (2)	As1—O7—Cr2	128.4 (2)
O3—Cr1—O4	109.1 (2)

Symmetry codes: (i) x, y+1, z; (ii) −y, x−y, z−1/3; (iii) x+1, y+1, z; (iv) −y+1, x−y, z−1/3; (v) −y+1, x−y+1, z−1/3; (vi) −x+y+1, −x, z+1/3; (vii) x+1, y, z; (viii) x, y−1, z.