(IUCr) Reversible displacive phase transition in [Ni(en)3]2+(NO3-)2: a potential temperature calibrant for area-detector diffractometers

Volume 36
Part 1
Pages 141-145
February 2003

Received 25 October 2002
Accepted 26 November 2002

Reversible displacive phase transition in [Ni(en)₃]²⁺(NO₃^-)₂: a potential temperature calibrant for area-detector diffractometers

L. J. Farrugia,^a ^* P. Macchi^b and A. Sironi^b

^aDepartment of Chemistry, University of Glasgow, Glasgow G12 8QQ, Scotland, and ^bDipartimento di Chimica Strutturale e Stereochimica Inorganica, Università degli Studi di Milano, Via Venezian 21, 20133 Milano, Italy
Correspondence e-mail: louis@chem.gla.ac.uk

The coordination complex [Ni(en)₃]²⁺(NO $[{}_{3}^{- }]$ )₂ (en = 1,2-diaminoethane) undergoes a sharp reversible displacive phase transition at 109 K, changing space group from P6₃22 above the transition temperature to P6₅22 below. The phase change is accompanied by a tripling of the c axis on cooling, resulting in an easy detection of the transition in images from area-detector diffractometers. The transition has been followed using a Nonius KappaCCD and a Bruker SMART APEX CCD. Data sets were collected over the temperature range 100-113 K and integrated using the low-temperature orientation matrix. Reflections with l [not equal to] 3n show a smooth and rapid decrease in intensity to zero on warming from 106.5 to 111 K. The results are reproducible to within ±2 K in two laboratories and suggest that this compound may be useful as a liquid-nitrogen cryo-calibrant for diffraction instruments equipped with area detectors.

Keywords: phase transition; temperature calibrant; area detectors.

1. Introduction

The use of area-detector diffractometers with liquid-nitrogen cryo-cooled samples is becoming the norm, both for routine structure determinations and for charge density studies using laboratory X-ray sources. The accurate temperature calibration of such devices is time-consuming and often relies on observing phase transitions in compounds that have known transition temperatures. For reliability of calibration, it is most convenient to have a calibrant compound that undergoes the phase transition near the working temperature, at around 100-120 K. One such compound is KH₂PO₄, which undergoes a paraelectric to ferroelectric phase transition at 122.5 K (Baur, 1973; Kobayashi et al., 1970). The space group changes from tetragonal I $[\bar{4}]$ 2d to orthorhombic Fdd2 and the transition may be observed through the splitting of the tetragonal 200 (or 020) reflection. This is a relatively straightforward procedure to undertake on a serial-detector diffractometer, but less simple with an area detector. Moreover, since the reduced cell undergoes only a subtle change, the phase transition is not immediately obvious from the images obtained from area-detector diffractometers.

The racemic labile coordination complex [Ni(en)₃]²⁺(NO₃^-)₂ (en = 1,2-diaminoethane), (1), is unusual in that it spontaneously resolves on crystallization from saturated aqueous solutions in the chiral space group P6₃22. The room-temperature X-ray structure of compound (1) was determined most recently by Korp et al. (1980) in order to correlate the absolute configuration of (1) with the sign of the optical rotary strength of the nickel(II) d-d transitions. In view of our continuing interest in the electron distribution in chiral nickel(II) complexes (Smith et al., 1997), we have assessed the suitability of (1) for an accurate charge density analysis. During the course of this assessment, we observed a sharp and reversible phase transition at 109 K; we now propose that this compound may be a suitable calibrant for liquid-nitrogen cryo-devices.

2. Experimental

2.1. Sample preparation

Although compound (1) is not commercially available, the synthesis is trivial and requires no specialized apparatus. An aqueous solution of commercial Ni(NO₃)₂ is treated with a slight stoichiometric excess of ethylenediamine, and the resultant deep-purple solution is allowed to evaporate in air. Large purple hexagonal needles grow rapidly as the solution becomes concentrated. For the diffraction studies, samples were cleaved from larger well formed crystals. Since the racemic complex spontaneously resolves on crystallization, the homochiral crystals obtained contain either the [Delta] or [Lambda] isomer of the tris-chelate cation. To facilitate a direct comparison of the structures, crystal specimens of the same [Delta] isomer were used in this study.

2.2. Data collection

Single-crystal diffraction data were collected on a Nonius KappaCCD fitted with an Oxford Cryosystems Series 700 Cryostream low-temperature device (in Glasgow) and on a Bruker SMART APEX CCD fitted with a Bruker Kryoflex liquid-nitrogen device (in Milano). The Bruker Kryoflex was calibrated by means of a pair of connected copper-constantan thermocouples. One was mounted on a goniometer head and placed in the goniometer centre, in the same position occupied by a crystal sample. The measured voltage was converted using standard conversion factors. In the range 90-120 K, the measured sample temperature was consistently 1.0-2.0 K higher than the temperature measured by the thermocouple inside the nozzle. For the Oxford Cryosystems Series 700 instrument, the temperature at the sample was taken from the instrumental digital readout. According to the manufacturers calibration curve, the sample temperature is given to an accuracy better than ±0.5 K when samples are placed at a standard 5 mm distance from the Cryostream tip (Oxford Cryosystems, 2002). The probe temperature is measured by a platinum resistance thermometer situated in the heat-exchanger region and the digital readout is factory-calibrated according to the flow rate and evaporator settings. During all the data collection experiments, the temperature was stable within ±0.1 K (monitored every 30 s).

Accurate atomic parameters for (1) above and below the transition temperature were obtained from high-resolution data sets. These were collected at 123 and 100 K on separate crystal specimens using a KappaCCD diffractometer, with either [omega] or [varphi] scans. At 123 K, 1664 frames were collected (34 oscillation sets, 1.8° oscillation angle, 2-160 s per frame), affording a mean redundancy of 27.7. At 100 K, 5523 frames were collected (16 oscillation sets, 0.4° oscillation angle, 1-6 s per frame), affording a mean redundancy of 13.3. The data were integrated using DENZO SMN (Otwinowski & Minor, 1997 $[Otwinowski, Z. & Minor, W. (1997). Processing of X-ray Diffraction Data Collected in Oscillation Mode, in Methods in Enzymology, Vol. 276, Macromolecular Crystallography, part A, edited by C. W. Carter Jr & R. M. Sweet, pp. 307-326. New York: Academic Press.]$ ) and processed using locally modified versions of DENZOX and SORTAV (Blessing, 1997). Absorption effects were corrected by Gaussian quadrature using the local program GAUSSIAN (Mallinson & Muir, 1985). Full-matrix least-squares refinements on F² were carried out using SHELXL97 (Sheldrick, 1997), with no constraints applied. All calculations were carried out using the WinGX software package (Farrugia, 1999). The results are summarized in Table 1.

Table 1
Crystallographic data for the complex [Ni(en)₃]²⁺(NO $[{}_{3}^{- }]$ )₂

	100 K	123 K
Formula	C₆H₂₄N₈NiO₆	C₆H₂₄N₈NiO₆
M_r	363.04	363.04
Crystal system	Hexagonal	Hexagonal
Space group	P6₅22	P6₃22
a (Å)	8.82020 (10)	8.83550 (10)
c (Å)	33.1447 (4)	11.08960 (10)
V (Å³)	2233.07 (4)	749.736 (14)
Z	6	2
Density (g cm^-3)	1.62	1.608
Linear absorption coefficient (mm^-1)	1.345	1.335
Wavelength (Å)	0.71073	0.71073
Crystal dimensions (mm)	0.44 × 0.41 × 0.33	0.50 × 0.39 × 0.30
Number of data collected	119347	92250
Number of independent data	7801	3253
R_int	0.0463	0.0313
Number of refined parameters	97	38
_max (°)	50.01	55.24
R(F), wR(F²) all data	0.0299, 0.0568	0.0217, 0.0552
_max, _min (e Å^-3)	0.415, -0.514	0.423, -0.877

Less extensive data sets were also recorded at 0.5 or 1.0 K intervals, in order to characterize the transition temperature. A typical data collection run on the KappaCCD consisted of two [omega] oscillations of 134° and 100° (262 frames, 5 s per frame). Data were reduced as described above. A typical data collection run on the SMART APEX CCD consisted of one [omega] oscillation of 180° (600 frames, 10 s per frame). At T = 90, 109 and 112 K, four oscillations (with [varphi] = 0, 90, 180, 270°) were collected. Data were reduced with SAINT (Bruker, 1999) and corrected for anisotropy effects with SADABS (Sheldrick, 1996).

In order to follow the phase transition in a semi-quantitative fashion, the sets of oscillation images in the temperature range 100-113 K were integrated using the orientation matrix determined at the lowest temperature, so that all data sets were indexed on the low-temperature P6₅22 cell, which is a super-cell of the high-temperature phase. No integration problems were experienced, though surprisingly the integration statistics were poorer in the intermediate region, where both phases are likely to coexist in separate domains.

3. Results and discussion

The chiral crystals of the coordination complex [Ni(en)₃]²⁺(NO₃^-)₂ undergo a structural phase transition. On cooling, this phenomenon is characterized by a tripling of the c axis and a change of the space group. Depending on the chirality of the constituent molecules, the high-temperature P6₃22 space group transforms into one of its subgroups: P6₅22 for those crystals containing the [Delta] isomer of the chelate, and P6₁22 for crystals containing the [Lambda] isomer.

The phase transition is also accompanied by a slight contraction of the a axis on cooling. Fig. 1 shows the temperature dependence of the cell length a and cell volume through the transition. A discontinuity in slopes is clearly visible for both parameters. The large change in the length of the c axis is immediately obvious from a visual inspection of the small-angle oscillation images typically used in data collection with area-detector diffractometers. Fig. 2 shows a typical single oscillation image, obtained with identical diffractometer setting angles, at temperatures of 106.5 and 111 K. The non-axial layer line 4 $[\bar{1}]$ l (l = 17-24), visible in the bottom right corner, clearly shows those reflections with l [not equal to] 3n (P6₅22 indexing) vanishing at 111 K. The intensities for all reflections with l 3n (P6₅22 indexing) drops to nearly zero through the transition temperature, as is shown in Fig. 3(a). The entire phase transition is very sharp and takes place between 106.5 and 111 K. The analogous curve for the SMART APEX data is shown in Fig. 3(b).

Figure 1
The temperature dependence of (a) the unit-cell a axis, (b) the unit-cell c axis (P6₅22 indexing) and (c) the unit-cell volume.

Figure 2
Oscillation images obtained from identical angular settings on the KappaCCD at (a) 106.5 K and (b) 111 K.

Figure 3
Overall diffracted intensities of reflections hkl for l = 3n, 3n + 1 and 3n + 2 as a function of temperature. Indexing is based on the low-temperature super-cell matrix and temperatures given are the instrumental setting temperatures. The sum of the averaged intensities within the three classes has been normalized to 1000. (a) KappaCCD data; (b) SMART APEX data. Note that a calibration of the Bruker Kryoflex showed a 1.0-2.0 K gap, compared with the actual temperature at the crystal (see §2.2

The phase transition is of the displacive type, because it involves a shift of the atomic position from higher- to lower-symmetry sites and no order-disorder mechanism can be proposed. In fact, the Ni atom lies on a site of 32 symmetry in P6₃22 (123 K) and on a site of twofold symmetry in P6₅22 (100 K), while the nitrate nitrogen atom N(100) lies on a threefold axis in P6₃22 and in a general position in P6₅22. The displacement is 0.4 Å for the cation and 0.442 Å for the anion from the threefold-axis sites in P6₃22. The projection of the unit cell (Fig. 4), viewed down the c axis, shows how the tris-chelate cations are arranged about the 3₁ axes. The displacement of the Ni atoms away from the 3₁ axis (or the 3 axis in P6₃22) is evident from this view.

Figure 4
Parallel projection view of the packing of the [Delta]

-[Ni(en)₃]²⁺ cations in the P6₅22 phase.

The molecular structure of the cation, shown in Fig. 5, remains essentially identical in both phases. The crystallographic D₃ molecular symmetry of the tris-chelate cation at 123 K is reduced to C₂ at 100 K, but nevertheless remains extremely close to D₃ as shown from calculations using the SYMMOL program (Pilati & Forni, 2000). The continuous symmetry measure CSM (Zabrodsky et al., 1993) is 0.112 and the root mean square (r.m.s.) atomic deviation from exact D₃ symmetry is only 0.033 Å. Thus the Ni-N distances at 100 K are in the range 2.1345 (5)-2.1389 (5) Å and at 123 K they are 2.1362 (4) Å. The two independent N-C-C-N torsion angles of the en ligand at 100 K are -54.5° and -55.5°, while at 123 K the corresponding (unique) torsion is -54.7°. The symmetry of the nitrate counterion is similarly reduced from crystallographic C₃ to C1, but barely deviates from idealized D_3h symmetry at either temperature. The r.m.s. atomic deviations from strict D_3h symmetry are 0.0083 Å at 100 K and 0.0075 Å at 123 K, while the corresponding CSM values are 0.0068 and 0.0056.

Figure 5
ORTEP view (Farrugia, 1997

[Farrugia, L. J. (1997). J. Appl. Cryst. 30, 565.]

) of the

-[Ni(en)₃]²⁺ cation at 100 K, along the molecular threefold axis. Displacement ellipsoids are shown at the 70% probability level, and primed atoms are related to unprimed atoms by the symmetry operation -x + y, y, 1/2 - z.

There is no sign of disorder in the high-temperature phase. The atomic displacement parameters (a.d.p.'s) show a virtually linear increase with temperature from 100 to 270 K (Fig. 6), consistent with being true thermal parameters. Moreover, the difference between the mean square displacement amplitude of atoms along the covalent bonds ( [Delta] _A,B) indicates that the derived a.d.p.'s are reasonable. For the Ni-N bonds, [Delta] _A,B is around 0.002 Å², and for the bonds involving the light atoms it is 0.0005 Å² or less, in the structure at 100 K.

Figure 6
Temperature variation of the largest principal eigenvalue of the U^ij tensor. Averaged values are used for atoms N(1), C(1) and O(100) at 100 K.

The reported phase transition appears to be of the first-order type for the following reasons: (a) the S-shape plots of the unit-cell dimensions (a, c and V) against temperature, which suggest a molar volume discontinuity at the phase transition point; (b) the large shift of the molecular centres of mass (away from the broken threefold axes).

The apparent continuity of the V/T plot (Fig. 1c) is probably due to the co-presence of the two phases. Thus the observed unit-cell parameters result from the weighted averages of those corresponding to the two lattices. The same reasoning holds for the observed intensities within the narrow (3-4 K) range of the phase transition. On the other hand, only minor hysteresis was observed on cooling and warming the crystal samples tested, at least judging from the unit-cell parameters. Further support for this classification and the temperature calibration could come from accurate differential scanning calorimetry measurements, but unfortunately the equipment installed in our laboratories cannot achieve such low temperatures.

4. Conclusions

We have shown that the displacive phase transition in [Ni(en)₃]²⁺(NO $[{}_{3}^{- }]$ )₂, which occurs around 109 K, is very sharp and reversible, and the measured transition temperature is reproducible to within ±2 K in two laboratories. The transition is immediately obvious from visual inspection of oscillation images, and so can serve as a quick indicator for area-detector diffractometers. We suggest that this compound may be useful as a liquid-nitrogen cryo-calibrant for this type of diffractometer.

References

Baur, W. K. (1973). Acta Cryst. B29, 2726-2731.
Blessing, R. H. (1997). J. Appl. Cryst. 30, 421-426.
Bruker (1999). SAINT Integration Software. Version 6.02. Bruker AXS Inc., Madison, WI, USA.
Farrugia, L. J. (1997). J. Appl. Cryst. 30, 565.
Farrugia, L. J. (1999). J. Appl. Cryst. 32, 837-838.
Kobayashi, J., Uesu, Y., Mizutani, I. & Enomoto, Y. (1970). Phys. Status Solidi A, 3, 63-69.
Korp, J. D., Bernal, I., Palmer, R. A. & Robinson, J. C. (1980). Acta Cryst. B36, 560-564.
Mallinson, P. R. & Muir, K. W. (1985). J. Appl. Cryst. 18, 51-53.
Otwinowski, Z. & Minor, W. (1997). Processing of X-ray Diffraction Data Collected in Oscillation Mode, in Methods in Enzymology, Vol. 276, Macromolecular Crystallography, part A, edited by C. W. Carter Jr & R. M. Sweet, pp. 307-326. New York: Academic Press.
Oxford Cryosystems (2002). http://www.oxfordcryosystems.co.uk .
Pilati, T. & Forni, A. (2000). J. Appl. Cryst. 33, 417-418.
Sheldrick, G. M. (1996). SADABS. Program for Absorption Corrections with the SMART System. University of Göttingen, Germany.
Sheldrick, G. M. (1997). SHELXL97. Release 97-2. University of Göttingen, Germany.
Smith, G. T., Mallinson, P. R., Frampton, C. S., Farrugia, L. J., Peacock, R. D. & Howard, J. A. K. (1997). J. Am. Chem. Soc. 119, 5028-5034.
Zabrodsky, H., Peleg, S. & Avnir, D. (1993). J. Am. Chem. Soc. 115, 8278-8298.

research papers

Reversible displacive phase transition in [Ni(en)3]2+(NO3-)2: a potential temperature calibrant for area-detector diffractometers