A versatile high-temperature furnace for neutron four-circle diffractometers

HAL Id: jpa-00245248

https://hal.archives-ouvertes.fr/jpa-00245248

Submitted on 1 Jan 1984

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

A versatile high-temperature furnace for neutron four-circle diffractometers

G. Heger, W.F. Kuhs, S. Massing

To cite this version:

G. Heger, W.F. Kuhs, S. Massing. A versatile high-temperature furnace for neutron four-circle diffrac- tometers. Revue de Physique Appliquée, Société française de physique / EDP, 1984, 19 (9), pp.735-738.

�10.1051/rphysap:01984001909073500�. �jpa-00245248�

A versatile high-temperature ^{furnace for neutron} four-circle diffractometers

G. Heger, W. F. Kuhs (+) ^{and S.} Massing

Kernforschungszentrum, Karlsruhe, ^F.R.G.

Institut Laue-Langevin, 156X, 38042 Grenoble Cedex, ^France

Résumé.

Ce rapport ^{décrit un} four à haute température, ^facile d’emploi, léger (0,9 kg) ^et adapté ^aux besoins d’un diffractomètre de neutrons à quatre cercles. Il est adapté ^{à des} températures ^{entre 330 et 1 200 K avec une stabilité} supérieure ^{à 0,05 K. Les} mouvements des cercles sont très peu limités : seule la rotation 03A6 ^est limitée à 200°. ^Le four fonctionne sous vide (typiquement ^{de 10-5} mbar) ^{et est} refroidi à l’eau. La température ^{est contrôlée par trois} thermocouples chromel-alumel, l’un flexible est en contact avec l’échantillon. La puissance électrique ^{est de 60 W}

à la température maximale. Seulement deux réflecteurs de vanadium à paroi ^mince (0,1 mm) ^{et une} enveloppe

extérieure sphérique ^{en aluminium} (enveloppe cylindrique ^{en Al dans une version} précédente) ^{se trouvent dans le}

faisceau de neutrons. Le four a été utilisé avec succès, pour collecter des données ^sur la phase incommensurable située entre les phases ^03B1 et 03B2 ^du quartz.

Abstract.

The design ^and performance ^{of a versatile} closed, light (0,9 kg), high temperature ^furnace adapted ^to

the needs of a neutron four-circle diffractometer is reported. ^It operates between 330 and 1 200 K with a long

term stability of better than 0.05 K. There are few restrictions on the movement of the circles; only ^the 03A6-range

is limited to 200°. The furnace operates ^under high ^vacuum (typically ~ 10-5 mbar) ^{with a} water-cooled base.

The temperature is controlled by 3 chromel-alumel thermocouples, ^one of which is flexible to allow it to be fixed

directly ^{at the} sample. The maximum electrical power requirement ^is

^{60 W.} Only ² thin-walled (0.1 mm) ^vana-

dium reflectors and an outer spherical ^{aluminium can} (cylindrical ^{Al can in a former} version) ^{are in the neutron}

beam. The furnace was used successfully ^{e.g. to} collect data of the incommensurate phase ^{at the} 03B1-03B2 ^transition

of quartz.

1. Technical specifcations.

During the last few years the demand for neutron diffraction measurements at high temperatures ^has been increased considerably. Especially ^the tempera-

ture range up ^to about 1 200 K is of interest for _expe- riments on e.g. structural phase transitions, ^ionic conductors, disorder and anharmonicity. ^{For suffi-} ciently precise diffraction data in most cases single crystal measurements of high quality ^are required.

We have adopted ^a water-cooled ^vacuum furnace to the needs of a neutron four-circle diffractometer with little restrictions of the angular ranges of crystal

orientation. The versatility ^{of the} device results from its low weight (0.9 kg) combined with highly ^flexible supply ^{tubes for vacuum and water} cooling. ^{In this}

way extended data sets of reflection intensities can be

collected, comparable ^to the situation without fur-

nace. In the neutron beam there are only ^two cylin-

drical vanadium reflectors (0.1 ^{mm thick} each) ^{and a}

thin-walled (0.5 mm) ^{aluminium can} (spherical ^or cylindrical). Therefore, ^the background ^{is rather low}

and can be corrected accurately ⁱⁿ particular by using

the w-scan technique. ^A sectional view of the furnace is shown in figure 1 ; its characteristics are summerized in table I. The resistivity heating is realized with a

thermocoaxial wire within a flexible steel tube. In this way there is no danger ^{of a} short circuit. The life time of the heater unit is not restricted by poor ^vacuum

operation. The maximum temperature of the heater is limited to 1400-1500 K. In figure ^{2 there are} plotted

the temperatures ^{of the two} thermocouples ^{fixed at the}

middle of the heater unit and at the sample position, respectively, ^{as a} function of electric _power. These

curves are taken at a vacuum of about 10- 6 mbar and with clean, ^new V-reflectors. The temperature gradient

increase with increasing temperature. It is recommend- ed to operate the furnace only ^{in the} range up ^to 60 W

corresponding ^to temperatures ^{at the} sample ^{of about}

1 200 K and at the heater of about 1 400 K. For the

higher temperature region > 900 K a good reflectivity

of the metal shields is essential. Nb-foils _may be used instead of V to reduce the incoherent scattering ^back- ground (if ^{the much} larger ^coherent scattering ^cross-

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/rphysap:01984001909073500

736

Table 1.

Characteristics of the furnace for ^neutron four-circle diffractometers.

water cooling of the furnace base

Fig. 1. - Vertical section through ^{the furnace.}

section of Nb is not troublesome). The fumace was

operated successfully ^{also at lower} temperatures (330-700 K) ^with Al-shields.

A sample crystal ^{can be either} directly glued ^on top of the W-sample ^{holder or} wrapped ^{in a} metallic foil for a better homogeneity ^of temperature. ^{In case of} high vapour pressure of the sample ^{material a closed}

metal or quartz ^{container can} be inserted. Samples

with dimensions _up to 10 mm can be used For the

preparation ^{of an} experiment ^the furnace with the mounted crystal ^visible (without reflectors and vacuum

can) ^can be centred within the Eulerian cradle of ^a four-circle diffractometer at room temperature (Fig. 3).

Fig. ^2.

Temperatures ^{measured at the} middle of the heater unit ( e) ^{and at the} top ^{of the} sample ^holder (in ^{direct contact}

with the pin) ( x ) ^{as a} function of electric _power. The diffe-

rence between the two curves gives ^the temperature gra- dient between heater and sample. These data were obtained with clean, highly reflecting vanadium shields. The broken line indicates the upper limit for routine operation ^{of the}

furnace.

2. Experimental verification.

The performance ^{of the furnace was tested} by ^neutron

diffraction experiments. ^{The main} points of interest

Fig. ^3.

^{View of} the open furnace mounted on the Eule- rian cradle of the four-circle diffractometer P32 at thé

Kernforschungszentrum ^Karlsruhe.

were the stability ^and homogeneity ^of temperature.

Due to the small dimensions of the furnace important temperature gradients between the sample ^{and the}

heater have to be taken into account (see Fig. 2). ^A

test was performed ^on the four-circle diffractometer D9 of the Institut Laue-Langevin using ^{a 3 x 3 x}

6 mm3 quartz single crystal mounted with ceramic

glue ^on top ^{of the} sample pin ^{with its} largest ^axis

parallel ^{to the} pin. Apart ^{from the two V-shields no}

further precautions ^were taken in order to homoge-

nize the sample temperature. ^{In a} first run the inten-

sity ^{of the} (113) Bragg reflection was recorded near the

ce-INC-fl-phase transition ; this reflection is reasonably

strong ^{in the} a-phase, ^{while the} intensity ^{is almost}

zero in the p-phase. ^The temperature ^was changed

in steps ^{of 1 K} approximately every 15 min and con-

trolled at the central heater thermocouple. ^{The results}

of these measurements are shown in figure ^{4. From the}

observed x-INC coexistence _range a temperature

gradient ^{of about 10} K/cm ^{across the} sample ^{is dedu-}

ced. The différence between the sample and heater tem-

peratures ^{was measured as 60 K} (probably ^{with an} important gradient in the ceramic glue). ^{A measure-}

ment at fixed temperature in the (x-INC coexistence range gave ^no significant intensity changes ^{over a} period of several hours. A second test run was perform-

ed controlling ^the temperature directly ^{at the} sample position. ^The long ^term stability (two days) ^{was found}

to be better than 0.05 K, ^while controlling ^{at the heater} always gave slightly drifting temperatures (typically

1 K/day) probably ^{due to} environmental changes ^i.e.

improving ^vacuum.

In a following experiment [1] performed ^{on the}

four-circle diffractometer D 10 of the Institut Laue-

Langevin further information on the performance ^of

Fig. ^4.

Temperature variation of integrated peak intensities of Bragg and satellite reflections at the 03B1-INC-03B2-phase ^tran-

sition in quartz : a) (113) Bragg peak measured with a freely ^{mounted 3 x 3 x 6 mm’} sample ^on heating ^and cooling, b) (- ^{0.03 2} 2) ^satellite peak measured with a 7 ^x 7 x 7 MM3 sample wrapped in aluminium foil on cooling [1]. Ti denotes the

INC-p phase transition, Tr the K-INC transition. In the temperature range between Tc ^and T’c or T’ ^the a-phase is coexist-

ing with the INC phase ^{due to thermal} inhomogeneities ^{of the} sample.

738

the furnace was obtained. This time the 7 ^x 7 x

7 mm3 sample ^of quartz ^was wrapped ^{in Al-foil in}

order to reduce the thermal gradient ^{over the} sample.

The temperature ^was controlled with a thermocouple

in direct contact with the sample. ^The intensity ^of

several satellite reflections of the incommensurate

phase (INC), ^{which covers the range} of approximately

1.3 K between the oc- and the p-phase, ^{was measured}

as a function of temperature. ^A typical series of inten-

sity measurements is shown in figure ^{4. The} range of coexistence of 1 K observed for the ^oc- and the INC-

phase ^{set an} upper limit on the thermal gradient ^over

the whole sample ^{of a few 0.1} K/cm ; the strain induc- ed by ^{the cement between} pin ^and sample may have contributed to widen the coexistence _range. This

was further confirmed by comparing ^this expérimental

results with results from a bigger furnace for triple-

axis instruments [2] ^{with a} gradient ^{of 0.1} K/cm [1].

The long ^term stability achieved with the four-circle set up ^{was +} 0.02 K over a period ^{of a few} days.

3. Routine opération.

The furnace is routinely operating between 330 and 1 200 K for samples up ^to 500 mm3. For the lowest temperatures ^{it is} preferable ^to work under _poor

vacuum to compensate partially ^{for the lack of}

radiation heating, especially ^for samples, ^{which are}

not in direct contact with the pin. ^The furnace allows for great freedom in all movements of a four-circle diffractometer. Good long ^term stability ^{is achieved} by controlling ^the temperature ^{at the} sample position

and small gradients ^are obtainable by wrapping ^the sample into metal foil. Some care has to be taken in adjusting ^{the PID} parameters for every given setup (a ^tunable microprocessor-based controller is

preferable ^{in this} case). Obviously controlling ^{at the}

heater unit gives ^less problems, ^but drifting vacuum,

changes ^{in the} reflectivity of the shields or in the thermal contact (e.g. ^{due to} chemical reaction of

sample ^and cement) normally ^{do not} allow for a

stability better than 1 K/day ^at higher temperatures.

One major problem in routine operation is the fact that the high ^vacuum prevents ^the measurement of

freely ^mounted samples with noticeable vapour pres-

sure. Such samples ^{have to} be sealed in a vacuum

tight ^{metal of} quartz container and then inserted

on a special ^holder.

Furnaces of this type ^are operational since almost 4 years ^at different laboratories (Kernforschungszen-

trum Karlsruhe, ^ILL Grenoble, ^C.E.N. Grenoble,

C.E.N. Saclay) ^without any major ^incident. Quite ^a

number of experiments have shown their high ^reliabi- lity ^and simple handling.

References

[1] ^DOLINO, G., BACHHEIMER, ^J. P., BERGE, B. and ZEYEN, C. M. E., ^J. Physique ^{45 (1984)} in press.

[2] DOLINO, G., BACHHEIMER, ^J. P., BERGE, B., ZEYEN, C. M. E., ^VAN TENDELOO, G., ^VAN LANDUYT, ^S.

and AMELINCKX, S., ^J. Physique ⁴⁵ (1984) ^sub-

mitted.