Aluminyum fluorür üç hidrat (AIF3.3H2O) kristalizasyonu

Abstract

Ü1 ÖZET Alüminyum fluorür elektrolitik raf inasyonla saf alüminyum metali üretiminde ve alüminyum indirgeme hücrelerinin sıvı elektro litinde kullanılan ana katkı maddesidir. Alüminyum fluorür üretiminin en önemli birim işlemlerinden biri AlF3.3HzO'm kristal izasyonudur. Endüstriyel kristalizasyonda kristal büyüme hızının belirlenmesi, gerek kristalizör tasarımı ge rekse istenileiı verimde kontrollü bir kristalizasyon işleminin yapı labilmesi için elzemdir. Bu çalışmanın ilk aşamasında alüminyum fluorür sistemi teorik olarak incelenmiş daha sonra ön hazırlık deneyleri yapılarak araştır manın teorik ve pratik bazı oluşturulmuştur. Ayrıca A1F3.3H20 ve AlF3.3,5H20'ın kararlılığı konusundaki belirsizlikler aydınlatılmıştır. İkinci aşamada alüminyum fluorür üç hidratın kristal büyüme ve çözünme kinetiği ile ilgili olarak %5-10'luk A1F3 çözeltileriyle 30- 85°C sıcakLık aralığında yapılan deneyler tamamlanarak kristal büyüme ve çözünme hızı çözeltideki aşırı doygunluğun fonksiyonu oİarak belir lenmiştir. Buna paralel olarak kristal büyüme ve' çözünme kinetiği ile ilgili mekanizmalar ortaya çıkarılmış geliştirilen teorik modellerin deney sonuçlarına uygunluğu doğrulanmıştır. Böylece AlF3.3H20'ın kris tal büyüme ve çözünme kinetiğinin `Parabolik Hız Kanunu `na uyduğu an laşılmıştır. Bunun yanısıra endüstriyel alüminyum fluorür üretimine benzer şartlarda %20-25'lik A1F3 çözeltileriyle çalışılarak bu şart larda AlF3.3H20,ın ortalama kristal büyüme hızı belirlenmiştir. Gene bu çalışmanın kapsamı içinde özellikle endüstride A1F3.3H20 kristalizasyoriu sırasında ortamda bulunan saf sızlıkların kristal büyüme hızına etkileri incelenmiştir. Ayrıca değişik sıcaklıklardaki çözünür lük değerleri saptanarak literatürdeki verilerle karşılaştırılmıştır. Bu çalışmada aynı zamanda tek ve iri A1F3.3H20 kristalleri üretilmeye çalışılmış değişik ortam ve şartlarda bu kristallerin nasıl ve ne şe kilde oluştuğu incelenmiştir» Son aşamada ise yapılan teorik ve deneysel çalışmanın sonuçla-, rina dayanarak optimum proses koşulları belirlenmiş ve AIF3.3H2Ö için bir pilot tesis tasarımı yapılmıştır.

IV SUMMARY Aluminum fluoride is used as a major make-up ingredient in the fused electrolyte of the aluminum reduction cell and in the electroly tic process for refining aluminum. The electrolyte, which ^contains aluminum fluoride, sodium aluminum fluoride and calcium fluoride, into which aluminum oxide is repeatedly dissolved as the ultimate source of aluminum, slowly loses the elements of aluminum fluoride. About 30 kg of aluminum fluoride need to be replaced for each ton of aluminum pro duced in the cell [1]. Minor uses of aluminum fluoride include additions to ceramics an,d glass, glazes and enamels, welding rod coatings and` catalysts [1,2]. The anhydrous aluminum fluoride is prepared by calcination of the trihydrate. This is formed by precipitation from an aqueous solution which may be prepared for instance by dissolving the hydroxide in hydro fluoric acid or fluosilisic acid., In aluminum fluoride' production processes which are widely used in industry one of the most important unit operation is crystallization. For successful operation of the process in, industrial applications and. for1 the design of a crystallize!: it is necessary to determine; 1) the conditions in which A1F3.3H20 can be crystallized, 2) the data for the induction period of aluminum fluoride solutions 3) the crystal growth rate as a function of temperature and super- saturation of the solution. A particular feature öf the aluminum fluoride crystallization process is that, as far as the stable sparingly soluble AlF3.3H20 is concerned, it requires prolonged heating of supersaturated solutions, preferably near to the boiling temperature instead of cooling, as is usually the case in other typical processes of this kind. Cooling the aluminum fluoride solutions provides different hydrates of aluminum fluoride all öf which are unstable and progressively transform to the A1F3.3HZ0. The different phases of AlF3-hydrates were investigated earlier by several authors using various methods fl3-'].5, 20-27], From the literature survey on aluminum fluoride it appears that there are some uncertainties about the stability of A1F.3H20 and A1FS.3,5H20.On the other hand, the studies on the crystallization of aluminum fluoride is concentrated on the clarification of the induction period which can be shortened by increased temperature and initial concentration in the presence of seed crystals. Additionally, eventhough the excess of HF shortens the induction period, in industrial applications the reactants are used in stochiometric amounts in order to prevent the corrosive and toxic effects. However in the production of aluminum fluoride from H SiF, it is suggested to use 1-10% excess acid to precipitate SiO in an easily filterable form [39}. Eventhough the knowledge of the dependence of crystal growth rate on concentration and temperature (See Equation (1.9) and (1.10)) is very important for the crystallizer designer and producer, the explicit relation for A1F3 crystallization has not been given in the literature. The reason for this may be due to certain difficulties in establishing a well defined homogeneous solution of sufficiently high super saturation, for the crystals to grow. If one tries to make a supersaturated solution by mixing a solu tion of an aluminum slat with a solution of an alkali fluoride, cryolite (Na3AlF6) or similar double salts precipitate instead of aluminum fluoride. In the case of using HF as fluorine compound for the reaction with A1(0H), high temperatures are required for the dissolution of A1(0H)` and it seems also difficult to prepare stochiometric A1F3 solution. The excess of HF or Al(0H)g in the solution causes problems in the investigation of crystal growth kinetics., In the literature it is stated that A1F3 crystallization is a slow process i compared to other processes and requires high supersatura- tion and temperatures, but the reasons and mechanisms of this fact are not indicated clearly. The main purpose of this thesis is to produce data for crystallizer designer and to establish practical information to the producer of aluminum fluoride, from an engineering point of view. It is also interesting to investigate A1F3-H20 system under certain conditions and determine crystal growth and dissolution kinetics. in order to obtain a closer view into the mechanisms involved in the crystallization and dissolution of A1F3 and the formation of particular solid phases. EXPERIMENTAL PART In the kinetic dissolution experiments, A1F3.3H20 (Riedel de Haen) after recrystallization was used. This was done by rotating 150 g of this product in 500 ml water at 75°C for 48 hours. Through the use an optical microscope it is ascertained that all the original particles (having the shape of intergrown spheres) had been transformed into A1F,3H 0 crystals, which were usually agglomerated.VI Experiments on the dissolution kinetics and for determination of solubility were performed by adding' 5 or, 10 grams of recrystallized Â1F3.3H2Q to 500 mL of pure water in a rotating. conductivity cell (See Fig. 3.'3). The; cell was placed in a thermostated water bath at the chosen temperature. (30-85°C). Electric conductivities were measured by means of a Radiometer conductivity meter CDM3 and recorded on a Servbgor recorder. The conductivity cell was calibrated directly in concentration units by performing dissolution experiments with less solid than corres ponding to solubility equilibrium, so that a known final concentration was obtained within a reasonable time (See Pig. 3.6). '.. Experiments on precipitation kinetics with pure supersaturated solutions of A1F3 were performed by pure AIF3.3H2O (Rie'del de Hâen) in boiling water at 100°C until saturation Was obtained. The solution was filtered and then evaporated `in 1.5 to 2 hours to between 1/8 and 1/4 of the original volume. The resulting supersaturated solution was then transferred to a stirred conductivity cell and placed in a thermostated water bath ât the experimental temperatures. in the range of 30-85°C (See Fig. 3.4). The conductivity was recorded as a function of time, and the cell was siliconized in order to avoid deposition of precipitate on the walls. At the highest concentrations (0.8-1.0 m) this was insufficient and the deposition was constantly counteracted by having a few `rolling stones` (plexiglass cylinder, 5x5 mm) circling with the solution through the cell. The conductivity was calibrated directly into aluminum fluoride concentration by conducting experiments with solution of known concentra tions (See Fig. 3.8). After each run the liquid was analysed quantitative^ for aluminum and flourine [49,50] and X-Ray powder diffraction method was used repeatedly to identify the precipitates. In the kinetic experiments reported here the precipitate was found to consist of A1F3.3H20 crystals only. The specific area determinations of the crystals were- made by a Quantasorb specific area measuring apparatus. To see the effect of addi tives on the rate of crystallization of aluminum fluoride trihydrate some, preliminary experiments were performed. For this purpose different com pounds in various amounts were added to saturated aluminum fluoride solu tions prepared as described above and through the analysis of final solu tions and the conductivity curves a conclusion about the effect of addi- tivities were tried to arrive at.. ^ Within the scope of this study the effects of the composition, temperature and the way of mixing of the solutions to xthe shape and the size of the A1F..3H 0 crystals were also examined.