To link to this article:

Size: px

Start display at page:

Download "To link to this article:"

Sarah Bates
5 years ago
Views:

This article was downloaded by: [Fudan University] On: 16 January 2014, At: 19:11 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:

1 This article was downloaded by: [Fudan University] On: 16 January 2014, At: 19:11 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: Registered office: Mortimer House, Mortimer Street, London W1T 3JH, UK Philosophical Magazine Publication details, including instructions for authors and subscription information: Flux separation methods for flux-grown single crystals Thomas Wolf a a Karlsruher Institut fürtechnologie, Institut für Festkörperphysik, Hermann-von-Helmholtz-Platz 1, D Karlsruhe, Germany Published online: 14 May To cite this article: Thomas Wolf (2012) Flux separation methods for flux-grown single crystals, Philosophical Magazine, 92:19-21, , DOI: / To link to this article: PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content ) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at

2 Philosophical Magazine Vol. 92, Nos , 1 21 July 2012, Flux separation methods for flux-grown single crystals Thomas Wolf* Karlsruher Institut fu rtechnologie, Institut fu r Festko rperphysik, Hermann-von-Helmholtz-Platz 1, D Karlsruhe, Germany (Received 19 March 2012; final version received 20 March 2012) After the growth of crystals an important step is their separation from the flux. Here the temperature of separation, the mechanical robustness and the chemical reactivity of the crystals constitute the basic conditions which have to be respected by the method. In this contribution several different techniques of flux separation at the end of a crystal growth are reviewed. Both processes which operate as long as the flux is in its liquid state and processes which can be applied when the flux has already solidified are discussed. Several rarely used methods will be illustrated in more detail. This contribution is intended as a guideline for crystal growers in basic research laboratories. Keywords: crystal growth; flux separation method 1. Introduction Flux growth is a powerful method to grow single crystals. Here they usually grow after spontaneous nucleation during slow cooling. In most cases the flux is solidified at room temperature and the crystals are entrapped in the solid eutectic matrix which, in metallurgy, is called regulus. Then due to different thermal expansivities between the crystals and the matrix, the crystals may experience a large stress which can lead to defects in the crystals or, in extreme cases, cracks in or destruction of the crystals. Therefore, in the past a series of methods have been developed, where grown crystals and solvent are separated while the flux is still in the liquid state. As this process often has to take place at high temperatures, it is often preferred to just let the filled crucible cool down to room temperature. Then the crystals are surrounded by a solid matrix and have to be extracted by dissolving the matrix with acids or bases. The problem here is to find a solvent which preferentially dissolves the eutectic and not also the crystals. This is especially difficult in cases where crystals are grown from self-flux, because both flux and crystals possess similar chemical properties. In the following, a series of separation techniques, both in the liquid and solid state, are reviewed in order to assist crystal growers. For a given material, in principle, more than one of these methods is possible. The decision as to which process is suitable for the system in question is determined by various parameters. Here the laboratory equipment available, the temperature of * thomas.wolf2@kit.edu ISSN print/issn online ß 2012 Taylor & Francis

3 Philosophical Magazine 2459 separation, the mechanical robustness and the chemical reactivity of the crystals constitute the basic conditions which have to be compatible with the method. 2. Flux separation in the liquid state When crystals suffer from stresses by solidified flux or when there is no suitable solvent to dissolve the flux, separation should be carried out while the flux is liquid. Such separation methods have to operate at high temperatures and often require extra equipment such as a centrifuge or a crystal puller. Furthermore, in some cases, care has to be taken during cooling down to room temperature to avoid thermal cracking of the crystals. This route is utilized by the nowadays most common growth techniques like Czochralski and top-seeded solution growth (TSSG) [1]. In these methods, a seed crystal is fixed at the lower end of a pulling rod. This rod is lowered until the seed crystal dips into the liquid solution inside a crucible making the beginning of the growth. During the subsequent period of slow cooling the seed crystal grows further and finally is pulled out of the flux at the end of the growth process. These techniques can operate at highest temperatures, which are only limited by crucible attack or severe evaporation losses of the flux. For separation temperatures below 1200 C the method of centrifugation, or spinning is a useful tool. It is often applied when metals, for example Sn or Al, are used for flux growth. Here the starting mixture is placed in a crucible which is covered by a second inverted crucible filled with silica wool. This whole assembly is sealed into a SiO 2 ampoule and subsequently transferred to the furnace to start the crystal growth process. At the end of the growth process, the ampoule, which is still at high temperature, is quickly taken out of the furnace, inverted and inserted into a centrifuge to spin off the remaining liquid flux. In this setup, the silica wool acts as a sieve for separating the flux leaving the crystals in the lower crucible [2]. Spinning temperatures above 1200 C, in principle, are possible. Then, instead of silica, other refractory materials have to be used for ampoule and sieve. However, as the radiation losses increase strongly with temperature, the ampoule will cool down rapidly during transfer to the centrifuge. This in turn may lead to a sudden dendritic growth of the crystals or even to a partial solidification of the flux. When crystals are grown from self-flux, in particular oxidic fluxes, instead of metallic solvents, the flux separation has to generally be carried out at higher temperatures. Here the separation by pouring out or decanting is appropriate. For this purpose a porcelain capsule is placed in the furnace next to the crucible wherein the crystals have to grow. The capsule contains fragments of a fire brick and is covered with a ceramic sieve. To carry out the separation process at the end of growth, the furnace is opened, the crucible is grasped with tongs and by careful tilting the remaining flux is poured out through the ceramic sieve. The flux percolates through the sieve and is sucked up by the fire brick, leaving behind almost freestanding crystals. Figure 1 shows the result of such a separation using GdBa 2 Cu 3 O 7 crystals grown in air as an example. In this system the self-flux consists of BaO and CuO which form rather reactive melts. It is therefore important to select inert materials for sieve and fire brick. Otherwise the reaction heat might lead to a

2460 T. Wolf Figure 1. GdBa 2 Cu 3 O 7 single crystals after pouring out through a ZrO 2 tissue. Figure 2. YBaCo 4 O 7 in a Pt crucible after isothermal decanting in a closed furnace.

The whole pouring out process should not take longer than a few seconds to avoid a too strong cooling, which might result in a solidification of the flux or crystal cracking.

4 2460 T. Wolf Figure 1. GdBa 2 Cu 3 O 7 single crystals after pouring out through a ZrO 2 tissue. Figure 2. YBaCo 4 O 7 in a Pt crucible after isothermal decanting in a closed furnace. backmelting of the grown crystals. In this experiment a ZrO 2 cloth, which acts as a sieve, has been placed onto pieces of a porous ZrO 2 fire brick. The whole pouring out process should not take longer than a few seconds to avoid a too strong cooling, which might result in a solidification of the flux or crystal cracking. To avoid detrimental cooling of the furnace during pouring out this procedure has to be carried out inside a closed furnace. An air-tight furnace even allows this process without changing the composition of the growth atmosphere. For the growth of YBaCo 4 O 7 crystals in Figure 2 the Pt crucible therefore has been put directly onto the ceramic cloth. After growth the crucible has been overturned using a rod, without opening the furnace. As in both cases the flux flows off the crystals within s, which is rather quick, there s the risk that the liquid surface will break. This leads to an unwanted decoration of the crystal surface with solidified flux droplets.

Philosophical Magazine 2461 Figure 3. (a) Schematic crystal growth setup which permits flux soaking. (b) Porous ZrO 2 fire brick after flux soaking and withdrawing from the crucible.

5 Philosophical Magazine 2461 Figure 3. (a) Schematic crystal growth setup which permits flux soaking. (b) Porous ZrO 2 fire brick after flux soaking and withdrawing from the crucible. (c) Oriented grown YBa 2 Cu 3 O 7 single crystals with very clean surfaces after flux soaking. In an alternate, soaking-up, method this drawback is circumvented by prolonging the separation process to 1 3 min. For this a porous fire brick is suspended from a wire which allows for the dipping of the brick into the flux at the end of growth (see Figure 3a). The brick then will suck up the melt like a sponge and, after the process is complete, can be withdrawn again from the crucible (see Figure 3b). In the now empty crucible crystals with particular clean surfaces remain (see Figure 3c). The difficulty in this method is to find a suitable sponge that can be wetted by the flux and at the same time is chemically inert to the flux. 3. Flux separation in the solid state In all the cases where a crystal growth experiment is cooled down below the solidus temperature of the flux, the crystals are entrapped in a solid eutectic matrix.

2462 T. Wolf Figure 4. MgB 2 single crystal platelet after evaporation of the Mg flux. Before use they have to be extracted either mechanically, e.g. with a dentist s drill or chemically by using a suitable solvent.

6 2462 T. Wolf Figure 4. MgB 2 single crystal platelet after evaporation of the Mg flux. Before use they have to be extracted either mechanically, e.g. with a dentist s drill or chemically by using a suitable solvent. The easiest way to remove solidified flux is to dissolve it with acids or bases. When salts are used as a flux, extraction often can be achieved by immersing the regulus in water. In many cases, however, the crystals are attacked by the solvent, too. Then it is necessary to optimize the dissolution rates of the crystals and flux by adjusting etching time, concentration and temperature of the solvent to keep the losses of the crystals as low as possible. A very gentle method to eliminate adhering flux is use evaporation. This works with metal fluxes, like Mg, Yb, Hg, Zn, etc., which exhibit vapor pressures in the range mbar at relatively low temperatures. Of course, the vapor pressure of the crystal phase has to be much lower. Here the solidified regulus is put into a silica tube which has to be pumped continuously. After heating the tube to an appropriate temperature the flux starts to evaporate and then condenses at a colder part of the tube leaving behind crystals with extremely clean surfaces. Figure 4 shows a MgB 2 platelet crystal grown from a Mg flux in a closed Mo ampoule. In this separation process a temperature of 600 C was sufficient to completely evaporate the flux within 3 h. In the following, a little-known separation method for fluxes like Mg, Al, Y or lanthanides will be discussed which works without using acids or bases. It is based on the reaction of the flux metals with isopropanol. The underlying reaction, M þ 3C 3 H 7 OH! HgCl 2 MOC ð 3 H 7 Þþ1:5H 2 þ Hg ðm ¼ Mg, Al, Y, rare earth elementþ, proceeds at temperatures of C within hours up to days [3]. HgCl 2 facilitates the reaction like a catalyst, but most likely, the formed Hg acts as a surfactant by removing oxide layers from the metal surface. To start the separation process the regulus is placed into a round bottom flask of an evaporator together with isopropanol and 10 4 mol HgCl 2 per g-atom of flux metal and then heated up. After a few hours the solution has to be decanted. One has to bear in mind that the formed isopropoxide in the solution freezes quickly and may block a filter paper. Furthermore, it hydrolyzes rapidly in humid air, precipitating a compound which

7 Philosophical Magazine 2463 Figure 5. Mg 1 x Al x B 2 single crystals uncovered by the isopropanol route. stays solid throughout the process. The cycle has to be repeated with fresh isopropanol and HgCl 2 until the crystals are free of flux. Figure 5 shows the result of the growth of Mg x Al 1 x B 2 single crystals out of an Al Mg flux which have been uncovered by the isopropanol method. To end, a separation method for the frequently used Sn flux will be presented. It operates, however, also for those flux elements which form a liquid alloy with Hg at rather low temperatures (5100 C). Here the regulus with the entrapped crystals is immersed in a dip of liquid Hg. At temperatures of C the bath metal dissolves the flux within 1 2 days. After decanting the liquid alloy the crystals left are contaminated with Hg. In a second step Hg can be distilled easily as described before. To circumvent the risks with toxic Hg, Ga can be used as solvent instead. Figure 6a shows the result of growth experiment where the grown YbNiSi 3 single crystals are still entrapped in the solid Sn flux. After immersing the regulus in liquid Ga and decanting the liquid alloy the now free-standing crystals are covered with a liquid layer of Ga (see Figure 6b). In contrast to the process where Hg is used as a solvent, the Ga route does not allow to clean the contaminated crystals by evaporation, because the vapor pressure of Ga is too low. Therefore, Ga has to be removed by chemical means, for example by reaction with I 2. To achieve reasonable dissolution rates a high concentration of I 2 in a solution is required. Salvador et al. [4] report on a 5-molar solution of I 2 in dimethyl formamide (DMF), which transforms Ga into GaI 3 already at room temperature. The contaminated crystals have to be immersed in this solution for minutes to hours. Afterwards they can be rinsed with water or ethanol. Figure 6c shows the crystals of Figure 6b after removal of Ga by I 2 -DMF. 4. Discussion and summary After the growth of crystals their separation from the flux is an important step. This can be achieved at high temperatures where the flux is still liquid, or at room temperature where the grown crystals are entrapped in a solid matrix. In the former

2464 T. Wolf Figure 6. (a) SiO 2 growth ampoule with YbNiSi 3 inside the Sn regulus on a mm grid. (b) Contaminated YbNiSi 3 single crystals after dissolving the Sn regulus with liquid Ga.

8 2464 T. Wolf Figure 6. (a) SiO 2 growth ampoule with YbNiSi 3 inside the Sn regulus on a mm grid. (b) Contaminated YbNiSi 3 single crystals after dissolving the Sn regulus with liquid Ga. (c) Clean YbNiSi 3 crystals after dissolving adhering Ga with an I 2 -DMF solution. Table 1. Flux separation techniques. Flux separation in the liquid state Flux separation in the solid state Czochralski method Top-seeded solution growth (TSSG) Centrifugation, spinning off Pouring out, decanting Soaking up, ceramic sponge Drilling Etching (acids, bases, water,...) Evaporation (Mg, Yb, Hg, Zn,...) Isopropanol route Alloying (with Hg, Ga) I 2 -DMF route case, often specialized equipment is needed to separate the crystals, whereas in the latter case a chemical dissolution or a commercial dentist s drill is already sufficient. The most current separation techniques which have been discussed in this contribution are listed again in Table 1. The decision as to which process is suitable for the system in question is dependent on various parameters. Here available laboratory equipment, the temperature of separation, the mechanical robustness, the

9 Philosophical Magazine 2465 chemical reactivity and the requested degree of perfection of the crystals constitute the basic conditions which have to be compatible with the method. Acknowledgements I would like to thank Prof. Paul C. Canfield, Dr Kai Grube and Dr Christoph Meingast for initiating and critically reading the manuscript. References [1] J.J. Gilman (ed.), The Art and Science of Growing Crystals, John Wiley, New York and London, [2] P.C. Canfield and Z. Fisk, Phil. Mag. B 65 (1992) p [3] L.M. Brown and K.S. Mazdiyasni, Inorg. Chem. 9 (1970) p [4] J.R. Salvador, C. Malliakas, J.R. Gour and M.G. Kanatzidis, Chem. Mater. 17 (2005) p.1636.

Hristo Kyuchukov a b a St. Elizabet University, Bratislava, Slovakia

Hristo Kyuchukov a b a St. Elizabet University, Bratislava, Slovakia This article was downloaded by: [Hristo Kyuchukov] On: 11 October 2012, At: 22:49 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer