European Journal of Mineralogy; June 2001; v. 13; no. 3;
p. 557-564; DOI: 10.1127/0935-1221/2001/0013-0557
© 2001 E. Schweizerbart'sche Verlagsbuchhandlung Science Publishers
A new radiative microfurnace for X-ray single-crystal diffractometry
Gianmario MOLIN1,*,
Fedora MARTIGNAGO2 and
Alberto DAL NEGRO2
1 Centro di Studio per la Geodinamica Alpina $CNR, c/o Dipartimento di Mineralogia e Petrologia, Università degli Studi di Padova, Corso Garibaldi 37, I-35100 Padova, Italy
2 Dipartimento di Mineralogia e Petrologia, Università degli Studi di Padova, Corso Garibaldi 37, I-35100 Padova, Italy
* e-mail: molin{at}dmp.unipd.it
This paper was presented at the EMPG VIII meeting in Bergamo, Italy (April 2000)
 |
Abstract
|
|---|
Microfunaces available today for X-ray single-crystal investigation (radiative, gas-flame, gas-flow) are generally temperature-monitored by a thermocouple placed in a different site with respect to the crystal. The new F1 microfurnace aims at the following goals: fast crystal mounting; access to a large portion of reciprocal lattice; low absorption of direct beam; efficient in situ temperature measurement; fast power supply regulation; high temperature stability; controlled atmosphere. In the F1, the crystal is directly glued to a thermocouple with refractory cement. The heating body, fixed on the 
and radially shiftable along the 
axis circle, hosts the Pt winding and connections of both gas flow and power supply, and is enclosed in a thin pyrolitic boron nitride (PBN) shield. Temperature stability is within ± 1°C.
Calibration of the F1 was carried out by collecting X-ray data from a single-crystal synthetic periclase at temperatures of 28, 150, 300, 450, 600, 700, 900 and 1000°C, and yielded the equation: a = 0.0000625 (7) [T(K) - 273] + 4.2083 (4), based on the ratio between cell edge and temperature. Structural refinements gave structural parameters and thermal expansion values at the investigated temperature. The mean coefficient of linear expansion up to 1000°C was 14.3 x 10–6 °C.
Key-words: microfurnace, high-temperature, X-ray diffraction, structure modifications, periclase.
|
Introduction
|
|---|
Structure modifications and phase transitions with temperature are goals peculiar to experimental mineralogy. In order to reach these goals, several microfurnaces attached to single-crystal four-circle diffractometers have been developed, of which an exhaustive résumé is given by Swanson & Prewitt (1986). More recently, Böhm (1995) and Scheufler et al. (1997) developed a simple hot-gas-stream device aiming at ensuring wider circle movements and reducing X-ray absorption. The various microfurnaces have several techniques for heating (radiative, gas-flame, gas-flow) and are similar as regards temperature monitoring and control. Temperature is generally monitored by a thermocouple placed in a different site with respect to the crystal. In the former devices (Swanson & Prewitt 1986), the effective experimental temperature is generally based on the ratio between temperature and the cell-edge expansion of a standard crystal (usually NaCl or MgO), commonly sealed close to the crystal in a single or double evacuated glass capillary in order to avoid the oxidation of Fe2+ in iron-bearing crystals. These techniques have several disadvantages, mainly slow, difficult crystal mounting and centering, and temperature estimation is also affected by propagation of errors of both cell-edge measurements of the standard crystal and uncertainties in experimental calibrations. Although the devices of Böhm (1995) and Scheufler et al. (1997) solve most of the problems affecting the former apparatus, the lack of continuous, instantaneous monitoring of the effective temperature on the crystal during data collection does not ensure in situ temperature-control and regulation.
|
The new F1 microfurnace, objectives and design
|
|---|
The new microfurnace (Fig. 1a), hereafter called F1, allows operation with a crystal directly glued to a small rounded thermocouple mounted on a modified goniometer head. With the F1, installed on a SIEMENS AEDII 4-circle diffractometer, the following series of specific goals were planned: 1) fast crystal mounting; 2) access to a large portion of reciprocal lattice; 3) low absorption of direct beam; 4) efficient in situ temperature measurement, with direct regulation of power supply; 5) high temperature stability; 6) temperature range from room temperature to 1100°C; 7) controlled atmosphere around the crystal by laminar fluxing of pure or mixed gases.
View larger version (99K):
[in this window]
[in a new window]
|
Fig. 1a. Photograph of F1 microfurnace attached to circle of SIEMENS AEDII 4-circle diffractometer during data collection at high temperature.
|
|
The F1 furnace has two main parts: a modified goniometer head, which constitutes the "bottom" of the furnace (part A, Fig. 1b) screwed on the circle, and the heating body (part B, Fig. 1b), fixed on the circle opposite to part A and radially shiftable along the ; axis. Part B is composed of a thin shield of pyrolitic boron nitride (PBN), which is very transparent to X-ray beams and allows the interior atmosphere of the furnace to be insulated with respect to the external one. Part B hosts the Pt winding and connections of both gas flow and power supply.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 1b. Sketch of furnace: A, modified goniometer head; B, mobile heating body; PBN, pyrolitic boron nitride cylinder.
|
|
|
Microfurnace, part A
|
|---|
The modified goniometer head supports the Pt, Pt-Rh10% thermocouple in a ceramic holder. A series of three thermal shields in Cr-Ni steel protects the goniometer head from direct thermal radiation and is the base of the cylindrical heating chamber (Fig. 1c). The thermocouple, directly connected to the power supply, yields crystal temperature and also allows rapid temperature regulation. When part B is far from the goniometer head, crystal centering under the diffractometer microscope is easily achieved.
View larger version (148K):
[in this window]
[in a new window]
|
Fig. 1c. Detail of goniometer head with thermocouple, thermal shields and heating body with Pt winding on alumina ceramic support.
|
|
|
Crystal mounting
|
|---|
Single crystal mounting is easily done under a stereoscope, the crystal being glued to the thermocouple with a very small amount of refractory cement (M-B
G
-100) mainly composed of aluminum phosphate and silica.
|
Microfurnace, part B
|
|---|
Part B is the mobile radiating body of F1. The shielding material is a thin pyrolitic boron nitride (PBN) cylinder (0.7 mm thick, 33 mm in diameter, 40 mm long). The main features of PBN are: 1) high X-ray transparency (absorption of direct MoK
X-ray beam less than 20 %); 2) high thermal-shock resistance; 3) resistance to oxidation up to 900°C (a temperature not reached by the cylinder wall up to a crystal temperature of 1100°C). X-ray diffuse scattering at both strike points of the direct beam on the PBN shield, mainly due to the polycrystalline nature of PBN, causes background modulation with a single, weak, broad peak, detectable in the range of 2
angles of 12–15°.
The 0.6-mm Pt wire is supported by a U-shaped alumina ceramic support (Fig. 1c), designed to allow a maximum radiating rate toward the crystal and to provide access to a large portion (up to 2 angle of 75°) of its reciprocal space. The alumina support is coplanar to the 
circle. In the high-temperature areas of F1, the load-bearing metal structure is composed of Cr-Ni steel; molybdenum is used for electric connections. Purifier equipment for O2 and H2O is connected to the gas flow. The purified gas enters the F1 through a flexible pipe, flow rate being 3 liters per minute, to allow pressurization of the F1 heating chamber, protecting both the components of the microfurnace and the single crystal from oxidation.
|
Power supply and temperature control
|
|---|
F1 temperature is controlled by a PID thermoregulator which supplies a current generator operating up to 25 A and 15 V. At the operating temperature of 1100°C, a power of about 250 W is required. The increase in Pt-winding resistance, due to volatilization at high temperature, is monitored by comparison between required temperature and applied power. After repeated heatings at the highest temperatures (30 h at 800–900°C and 15 h at 1000°C), no significant increase in Pt-winding resistance was observed. During F1 testing at maximum power, 1200°C was reached in a few minutes and remained constant, indicating that the highest routine working temperature of 1000°C is to be considered a sufficient safeguard for F1 itself.
Up to 1100°C, the temperature stability of F1 in routine conditions and without circle rotation is $pL 1°C. Hazen & Finger (1982) observed that temperature varies widely, due to the "chimney effect" by rotation of the microfurnace around the circle. The maximum chimney effect, obtained in F1 by sudden rotation of circle from 0° to $pL 90°, is 4°C, which may be reduced to the standard stability value of $pL 1°C in less than 2 minutes. The low shift of temperature due to the chimney effect in F1 is obtained without any kind of T-T’- algorithm or programmable controller (Finger et al., 1973) and is comparable with the high stability obtained by Swanson & Prewitt (1986) in their M5 microfurnace.
|
Experimental procedure
|
|---|
A single crystal of transparent, colourless, synthetic periclase was employed for furnace and thermocouple calibration. MgO is a particularly good candidate because its changes in lattice parameters with temperature have been widely reported in the literature (Skinner, 1957; Hazen, 1976; Swanson & Prewitt, 1986).
Chemical analysis was performed on a thin polished section of the same crystal, using a C
-C
electron microprobe operating with a fine-focused beam (
1 µm) at 15 kV with 20 nA sample current in the wavelength-dispersive mode (WDS), with counting times of 20 s for peak and background. Bulk composition, obtained by averaging 27 microprobe spots, is: MgO 100.35 (29), FeO 0.09 (2), MnO: 0.02 (2), NiO 0.01 (2), ZnO 0.02 (3), Cr2O3 0.04 (3), TiO2 0.01 (1), Al2O3 0.01 (1), Sum 100.55 wt.%, with estimated standard deviations in parentheses.
The MgO crystal (radius of the equivalent sphere 0.22 mm) was directly glued to a thermocouple with refractory cement.
Data for eight different temperatures were collected on a SIEMENS AEDII four-circle diffractometer using graphite-monochromatized MoK radiation. Unit-cell dimensions and diffraction intensities were measured at room temperature (28°C) and then at 150, 300, 450, 600, 700, 900 and 1000°C. A set of 56 reflections (18 independent) was collected up to 2 max 70°. The time for each data collection was about 2 hours, including the time required for orientation matrix and cellparameter measurement. The cell-edge value was determined at each temperature from 43 reflections in the range 43° < 2 < 64°. Each reflection was centered on the positive and negative sides of 2.
All structural refinements were carried out in space group Fm3m using the S
-93 refinement program. In periclase, both magnesium and oxygen lie in special fixed positions, so that only three variables, a scale factor and the magnesium and oxygen isotropic temperature factors, need be refined in the least-squares procedure.
|
Results
|
|---|
The expansion data of periclase expressed as variations in cell edge as a function of temperature from 28 to 1000°C are plotted in Fig. 2a, together with previous data from Skinner (1957), Hazen (1976) and Swanson & Prewitt (1986). The results fit the above data, particularly those of the latter authors.
Unit-cell parameters and structural refinement results are listed in Table 1. Magnesium and oxygen isotropic temperature factors U (Å2) of periclase are plotted versus temperature in Fig. 2b. Both parameters reveal a linear trend with temperature.
The mean coefficient of linear expansion up to 1000°C, calculated following the equation of Hazen & Finger (1982) using the mean M-O bond length, is 14.3 x 10–6 °C. This value is in very good agreement with that of 14.2 x 10–6 °C calculated by Hazen & Prewitt (1977).
Calibration of F1, carried out by least-squares fit of cell-edge expansion data at the above temperatures, gave the following equation:
 | (4) |
|
Acknowledgements
|
|---|
This study was supported by grants from the CNR (Centro di Studio per la Geodinamica Alpina, Padova) and from MURST "Relations between structure and properties in minerals: analysis and applications"; 1997, to A. Dal Negro.
Received 20 May 2000
Modified version received 4 January 2001
Accepted 15 January 2001
|
References
|
|---|
Böhm, H.T. (1995): A heating device for four-circle diffractometers. J. Appl. Cryst., 28, 357.Finger, L.W., Hadidiacos, C.G., Ohashi, Y. (1973): A computer-automated, single-crystal, X-ray diffractometer. Carnegie Inst. Washington Year Book, 72, 694–699.
Hazen, R.M. (1976): Effects of temperature and pressure on the cell dimension and X-ray temperature factors of periclase. Am. Mineral., 61, 266–271.[Abstract][ISI][GeoRef]
Hazen, R.M. & Finger, L.W. (1982): Comparative crystal chemistry: temperature, pressure, composition and the variation of crystal structure. Carnegie Institution-Geophysical Lab. Washington, Ed. John Wiley & Sons, Chichester.
Hazen, R.M. & Prewitt, C.T. (1977): Effects of temperature and pressure on interatomic distances in oxygen-based minerals. Am. Mineral., 62, 309–315.[Abstract][ISI][GeoRef]
Scheufler C., Engel K.V., Kirfel A. (1997): An improved gas stream heating device for a single-crystal diffractometer. J. Appl. Cryst., 30, 411–412.[CrossRef][ISI]
Skinner, B.J. (1957): The thermal expansions of thoria, periclase and diamond. Am. Mineral., 42, 39–55.[ISI][GeoRef]
Swanson, D.K. & Prewitt, C.T. (1986): A new radiative single-crystal diffractometer microfurnace incorporating MgO as a high-temperature cement and internal temperature calibrant. J. Appl. Cryst., 19, 1–6.[CrossRef][ISI]
This article has been cited by other articles:
|
|
|
|
 
G. D. Gatta, F. Nestola, and J. M. Walter
On the thermo-elastic behaviour of kyanite: a neutron powder diffraction study up to 1200{degrees}C
Mineralogical Magazine,
June 1, 2006;
70(3):
309 - 317.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
F. Martignago, G.B. Andreozzi, and A. D. Negro
Thermodynamics and kinetics of cation ordering in natural and synthetic Mg(Al,Fe3+)2O4 spinels from in situ high-temperature X-ray diffraction
American Mineralogist,
February 1, 2006;
91(2-3):
306 - 312.
[Abstract]
[Full Text]
[PDF]
|
|
|
Copyright © 2008 by E. Schweizerbart'sche Verlagsbuchhandlung Science Publishers
Top
Abstract
Introduction
