31 Growth from melt II

Prof. P. N. Kotru

epgp books

 

 

31.1   Introduction

 

The techniques were originated by Bridgman (1925) and Stockbarger (1938) and so are named after them. In these techniques a crucible containing the material to be grown is lowered through a furnace in such a way that the lowest point in the crucible and the solidification surface rises slowly up the crucible. It means that the melt contained in the crucible is progressively frozen to yield a single crystal. The rate of lowering the crucible may range from about 0.1 to 200 mmh─1 but in most of the cases it may range somewhere in between 1-30 mmh─1. There are situations where the movement of the crucible is reversed. In other words, the crucible is raised up through the furnace and so is advantageously applicable for materials which are volatile; the interface with the vapour being

 

the coolest part of the charge. In this case, it is important that the material does not have different densities when in the solid and liquid phases. Suppose the solid material has lower density than its melt, the crucible containing the charge is likely to crack. However, if the solid has a greater density than its melt, it will lead to differential dimensions of parts of the growing crystal with respect to the crucible. These are the problems with such materials and can be overcome by some specialized techniques which may, however, take the technique far from being a simple one.

 

There is one important requirement of any crucible based technique and so applies to Bridgeman technique also. It is that neither the liquid nor its vapour attacks the crucible. If there is any possibility of crucible getting attacked, its slow dissolution will contaminate the growing crystal. Also, it is important that the crystal and crucible have the same thermal expansion coefficients. If it is not so, the adhering crystal will get strained which will adversely affect its utility.

 

There have been several modifications in the basic technique of growth right from its very origin e.g., Stober (1925); Kapitza (1928); Pfann (1958) and several others. We shall, however, discuss here only the Bridgman technique.

 

31.2    Basic Principle.

 

The basic principle behind this technique is normal freezing, in which the ingot of the concerned composition other as shown in a schematic diagram of figure 31.1 freezing , also known as  directional is  gradually frozen from one  end to the

Figure 31.1: Schematic diagram revealing the principle of directional freezing or normal freezing.

 

In this technique of growth a crucible is used and the directional solidification is carried out in order to achieve single crystal growth from melt. There are a number of ways in which this method is applied and are known as Bridgman technique, Stockbarger technique, Tammann technique, Stober technique and Obreimov and Schubnikov technique. The basic principle used in all these techniques is however, the same as applied by Bridgman for the growth of fluorite crystals.

 

31.3     Apparatus Used.

 

The apparatus used in Bridgman technique is shown in figure 31.2

Figure 31.2: Schematic diagram showing the experimental arrangement of Bridgman technique of crystal growth

 

The furnace used here consists of two halves, the lower half and the upper half. The temperature maintained at the lower half is just below the melting point of fluorite and the temperature of the upper half is above the melting point of fluorite. The crucible is made up of platinum and has a pointed lower end as is shown in the diagram. The crucible is filled with high quality natural fluorite, or purified powder if required, to which desired impurities may also be added. A platinum wire hangs the crucible in the upper furnace until the contents are completely melted. This crucible is then lowered from the upper into the lower part of furnace over a period of several days. The lowering is done with the help of an electric motor and reduction gearing system. On lowering, since the tip of the crucible enters the lower furnace first the fluorite will start to crystallize there. It is how the nucleation problem is overcome. With the continuous lowering of the crucible, the crystallization proceeds until all the contents get solidified. With the temperature kept constant, crucible shaped correctly and with the perfectly uniform lowering, the crucible contents will end up as single crystals of fluorite.

 

In general, the material to be crystallized is contained in a cylindrical crucible which is lowered through a two zone vertical furnace (shown in figure 31.3). The temperature profile of the furnace is arranged to have the form as shown in figure 31.3.This method, however, cannot be used for materials which expand on solidification.

 

31.4      Examples of Some Crystals Grown By Bridgman Technique

 

There are several materials which have been successfully grown by Bridgman technique. Only a few may be given here as examples, besides those which have been mentioned at different places in this discussion. These are:

 

1.  Al2O3 crystals having melting point of 2037°C have been grown using Molybdenum crucible.

2.  FeAl2O3 crystals having melting point of 1790°C have been grown using iridium crucible.

3.  Ge crystals having melting point of 937°C have been grown using carbon coated silica.

4.  Cu crystals with melting point1083°C have been grown using crucible made of graphite powder.

Figure 31.3: Schematic diagram showing two zone vertical furnace alongwith temperature profile used in Bridgman technique of crystal growth

 

 

The technique is best suited for low melting point materials. Many of the materials grown by this technique are NaCl (halite), KCl (sylvite), CaF2 (Fluorite), CaWO4 (Scheelite), AgCl (Cerargyrite), AgBr (Bromyrite) and several other metals and semiconductors. This was the method used by Bridgman and others for the growth of large metal single crystals and later by Stockbarger for the growth of optical quality alkali halide crystals for prisms and lenses. It was developed independently in Europe by Tammann and Obreimov and Schubnikov.

 

31.5 Crystal Growth by Zone Melting.

 

31.5.1    Introduction.

 

This technique was developed by Pfann of Bell Telephone Laboratory in 1952. Initially this technique was applied for the purification of semiconductor materials. It can be classified as a solid-liquid-solid process and is used both as crystal growth and a purification technique.

 

31.5.2   Basic Apparatus.

 

The apparatus used for the growth of crystals by this method is shown in figure 31.4.The material to be grown as crystal is placed in a long boat with dimensions of 1 foot in length and 1 inch in diameter, so that it is completely filled. Then a section of the material is melted with the help of narrow heating coils. The molten material is now traversed along the boat by moving either the boat or the heating coil. As heating coil proceeds the material melts leaving material behind it to solidify. The process of travelling of the heating coil continues with the melting of new material in front and solidifying to form crystalline material at the back. After several hours of this process and under suitable conditions the contents of the boat end up as one single crystal. For purification many more runs are performed repeatedly till the desired results are obtained.

Figure 31.4: Schematic diagram showing arrangement for zone melting technique of crystal growth

 

31.5.3     Floating-Zone Process.

 

A modification of the above said technique was given by Keck and Golay in 1953. The modified version is known as “floating zone technique”. In this modified technique the material to be purified or grown as a single crystal is arranged in a vertical compact rod as shown in figure 31.5.

Figure 31.5: Schematic diagram showing arrangement of floating zone technique using vertical ingot

 

This technique does not require any crucible. The molten zone floats below two solid parts of the rod held in place by surface tension. Since it does not require retaining crucible, the technique has the advantage that possible contamination from the container is avoided. It is used for the growth of silicon which is the only material grown on a very large scale around 1000 tonnes per year. It is also used for the growth of very pure crystals, mostly metals but on a small scale just for research purposes. Because of radio-frequency heating, the surface of the molten zone is heated and the liquid near to the melt- vapour surface is hotter than elsewhere. As a result, fluid flow is produced. The shapes of the solid-liquid surfaces are controlled by these flows and the rate movement of the zone.

 

31.5.4     Apparatus for Materials Involving Volatile Components.

 

Figure 31.6 shows an arrangement of zone melting technique in which there is involvement of a volatile component such as As in InAs or GaAs. In this arrangement, a tube is used which contains the boat filled with the material for zone refining or zone melting. The material is maintained at the required temperature and a movable heater melts a molten zone which is passed down the ingot. A seed crystal may be introduced at the starting end itself if one wants to have single crystal growth. There is a second boat which contains the volatile component and is placed at just the required temperature so as to make sure that there is no net loss of that component from the molten zone. The heating is done by radio frequency induction. There is a second boat which contains a volatile component and is placed separately at a lower temperature. The temperatures are so adjusted as to prevent any net transport to or from the molten zone.

Figure 31.6: Schematic diagram showing arrangement for Zone melting / refining involving volatile components.

 

Zone melting is an extremely important addition to crystal growth and purification techniques with immense potential for producing materials which are of great use in modern technology. Keeping that in mind, Zone melting and its applications will be discussed separately in greater detail.

 

31.6 Other Processes of Growth From Melt.

 

31.6.1 Introduction.

 

This category of crystal growth from melt distinguishes itself from the ones already described. In fact, eighty percent of melt grown crystals are produced by Czochralski and Bridgman techniques, but there are a variety of crystals which cannot be grown by these techniques where the melt is in contact with a crucible. If there are no suitable crucibles available, then there is no option but to use other methods. So, other processes of growth were devised, which are five in number, where melt is in contact with its own solid or with a seed crystal. Obviously, methods of this category are advantageous in the sense that the contamination is avoided. In two of these techniques there are two melt-solid surfaces (i.e., one with the crystal and one with the retaining solid) which raise two sets of problems. One set of problem arises on account of the fact that the melt volume can change. This change in volume can be very significant. It creates problems in achieving uniform doping which can be detrimental to the applications of these crystals. The other set of problems is on account of strange temperature distributions which, except for Verneuil method, creates convective flow which add to the difficulties in the growth of homogeneous crystals. The temperature distributions can also become a cause for radial heat flow. There was a time when it was felt that the radial flow was unavoidable and that high dislocation densities in crystals were expected. The technique of floatingzone refining which succeeds in producing dislocation-free crystals proved it wrong. The details of this technique will be taken further in the text.

 

31.6.2    Flame Fusion Technique .

 

This is the most widely used crystal growth technique as far as gem stones are concerned. This method originated before the turn of the century. It was discovered by Verneuil somewhere in 1890’s and had been used without any change upto 1950’s.This process has been used for a large number of materials of high melting points , viz. , ZrO2 (m.p. 2700⁰C ) , Y2O3 (m.p.2400⁰C) , MgAl2O4 (m.p.2130 ⁰C) and other high melting point materials. However, the largest use of this technique has been for the growth of alumina (Al2O3) with chromium as a dopant (popularly known as Ruby). Several types of heat sources have been used to produce temperatures as high as 2800⁰C and more as for example Solar Furnace, Glow Discharge, Plasma Torch, Arc Image and Radio-Frequency.

 

31.6.3     Apparatus for the Crystal Growth of Ruby.

 

 

Figure 31.7: Schematic diagram illustrating Flame Fusion growth.

 

 

For growth of crystals by this method the first step to go ahead is the preparation of feed material. For the growth of ruby, for example, the feed material consists of finely divided free -flowing alumina containing chromium oxide as colouring agent. The feed material is prepared from ammonium alum NH4Al (SO4)2.12H2O. Ammonium alum is crystallized from water until it is quite pure. This crystallized ammonium alum is then heated in a furnace at about 1100⁰C. It decomposes with the formation of ammonia (NH3), sulphuric anhydride (SO3), and water (H2O) as gases. Alumina is left behind in a fluffy form. After careful grinding and screening of alumina, it is used in the growth process. The colouring agent is normally added before the heating step. Verneuil furnace for the growth of ruby is shown in figure 31.7.

 

The main part of the system is a blow torch, using hydrogen and oxygen set up with the flame pointing downwards. The central tube carrying oxygen has in it a part of salt shaker containing the feed powder. A small hammer automatically hits the feed hopper several times a second so as to produce a constant flow of powder down the centre of the flame. The flame is made to impinge on a pedestal where a small pile of partly fused alumina quickly builds up. With the rise in the size of the pile, it touches the hotter part of the flame (over 2050⁰C) with the result that the tip becomes completely molten. This molten region increases in size and starts to solidify at the lower end. As more powder arrives, the solidifying region broadens into a ruby crystal with a molten cap on it. At this point, lowering of the pedestal is begun and the crystal continues to grow in length only. Such a crystal is called “boule”.

 

Constant growth conditions are extremely important for the prevention of “boule” from cracking. Care is required to be taken in controlling the flow of gas and in lowering of the pedestal. This lowering of the pedestal is done mechanically. To prevent colour banding in the crystal, the tapping mechanics, which gives rise to intermittent feed flow , is replaced by a small vibrator or screw feed.

 

31.6.4       Modified Technique .

 

An interesting modification of this technique, as developed by Linde Company, is to grow crystals in the form of discs. Here, the verneuil torch is made to impinge on a horizontal ruby rod which rotates as shown in figure 31.8. As the crystal grows, the rotation mechanism is lowered, producing ultimately a flat disc about 3/4// thick and up to 5// in diameter.

Figure 31.8: Schematic diagram illustrating verneuil growth in the form of disc

 

 

A    large number of different heating techniques have been used in place of hydrogen oxygen torch, either to produce temperatures or to enable the atmosphere, in which the crystal grows, 275 to be controlled. Various forms of  radio frequency ( r.f.), plasma heating , electron beam heating or microwave energy have been used , as well as image furnaces in which the heat from sun or from a carbon arc is focussed by mirrors to melt the powder. The general growth technique, however, remains the same as with the original verneuil method.

 

A number of crystals have been grown with the modified Verneuil technique. These are rutile  (TiO2), strontium titanate  (SrTiO3), fabulite, ruby, sapphire and semiconductor crystals.  In general, Verneuil growth tend to produce crystals which are imperfect and can be put to use only for applications  where imperfections  do not  matter, say for example  , jewellery , or  where they are required for  increasing the  hardness of crystals.  Such hard materials  find applications  in highly strained components like record player styli which get improved by high dislocation densities.

 

 

31.7     Arc Fusion Method.

 

This method is used to grow crystals of certain oxides like MgO, ZrO2, SrO, CaO and BaO. Let us describe the basic process involved in this method through schematic diagram shown in figure 31.9

Figure 31.9: Schematic illustration of sequential description of submerged arc fusion technique of crystal growth

 

The sequence of various stages of this process is given as follows:

  1. The raw material (of which the crystals are to be grown) in the powder form is packed either in a firebrick or in a water-cooled steel enclosure as shown in figure 31.9(a). Along with the powder, graphite electrodes (typically of diameter 5mm) are put in place inside the packed powder charge.
  2. There is a vent hole which extends almost right up to the region of the arc.
  3. The density of the powder charge is kept at more than 35% of the density of the crystal to be grown by this method. The critical density of the powder is achieved either by pressing or sintering it.
  4. An arc is struck in this powdered charge through the electrodes which melts the powder around the electrodes and thus a molten pool of the material is produced as shown in figure 31.9(b).
  5. The arc is turned down slowly and the charge is left as such so as to get solidified. The poor ther-mal conductivity of the powder charge further helps in keeping the rate of cooling slow. The crys-talline material so produced as shown in figure 31.9 (c) is then collected on breaking open the fused shell.
  6. The empty space provided by taking out the crystalline mass from the firebrick is filled with the fresh raw material again and the process repeated for the second run. The left out empty shell after harvesting the crystals is shown in figure 31.9 (d).
  7. The process can be repeated multiple times depending on the quantity of crystalline material required.

 

The method described above is due to Rabenau (1964), Schupp (1968) and Butler et al (1970, 1971). The crystals so produced are in the dimensions of 20-80 mm and around 500 g by weight. However, the crystals produced are of poor quality with high density of dislocations and may also contain gas inclusions and/or particles of powder that could not be fused. This method requires a large power supply of the order of 35 kw at 10 to 100 V. The technique is utilised for obtaining low-cost crystals of materials with high melting points and difficult to produce by other techniques. Like verneuil growth here also melt is created in contact with its own solid.

 

 

31.8   Skull Melting.

 

It is another technique where melt is created in contact with its own solid. In this method watercooled crucible is used. This method is used for the growth of a number of materials in the crystalline form. It has been used extensively for the growth of cubic Zirconia. The method is basically limited to electrically conducting materials as the radio frequency fields couple only to conducting materials. It is well known that most materials become conducting on heating. This property is utilised here by heating the charge and then inserting a conducting cylinder or pipe which is withdrawn when the charge gets suitably heated and gets coupled to the field. The method has an advantage over arc fusion method as there is far better control over the power input. It uses radio-frequency systems for which power supply can be controlled to one part in several thousands. As against this, the power supply in an arc is difficult to control. The method makes use of specially designed crucible which is split into segments as shown in figure 31.11. These segments are required to prevent the crucible from getting coupled to the radio-frequency field. Use of segmented crucible is, therefore, essential.

Figure 31.10: Schematic illustration of various stages in the process of growth by Skull melting.

 

Figure 31.11: Schematic diagram of segmented crucible in the process of growth by Skull melting

 

The process of skull melting may be explained through schematic diagrams of figure 31.10 in which each stage is described separately as follows:

  1. The crucible is filled with the raw material of which crystals are to be grown as shown in figure 31.10 (a).
  2. The charge is melted by heating the crucible with the help of radio-frequency as shown in figure 31.10 (b).
  3. The supply of power is reduced to allow the melt to get progressively frozen as depicted in figure 31.10 (c).

 

you can view video on Growth from melt II

References.

 

1.Brice, J.C. : “ Crystal Growth Processes”, Blackie & Son Ltd., Glasgow,1986.

2. Pamplin, B.R.: “ Crystal Growth”,II edition, Pergamon Press,1980.

3. Gilman, J.J.: “ The Art & Science of Growing Crystals”,Wiley, N.Y.,1963.

4. Brice,J.C.: “ The Growth of Crystals from Liquids”,North-Holland-Elsevier,N.Y.,1973.

 

Suggested Reading For More Information.

 

In addition to what is already suggested in module-XXX, the following references are given as sources of more information on melt growth in general .

  1. Aleksandrov,L.N.: “ Growth of Crystalline Semiconductor Materials On Crystal Surfaces,Elsevier, Amsterdam,1983.
  2. Keller,W and Muhlbauer,A: Floating-Zone Silicon,Dekker,N.Y.1981.
  3. Vergnoux,A.M.,Giordano,J and Foux,M: “ Crystal growth”ed. H.S.Peiser.Pergamon,Oxford,1967