HUNGAR~JOURNAL OF INDUSTRIAL CHEMISTRY VESZPREM Vol. 30. pp. 219-227 (2002) A NEW PROCEDURE OF PHOSPHOGYPSUM PURIFICATION IN ORDER TO DIMINISH THE CONTENT OF RADIONUCLIDES M. B. RAJKOVIC and D. TOSKOVIC1 (Institute of Food Technology and Bioschemistry, Faculty of Agriculture, University of Belgrade, P.O.Box 127, YU-11081 Belgrade-Zemun, YUGOSLAVIA 1Faculty of Technology, Zvornik, University of Serbian Saraevo, REPUBLIC OF SRPSKA) Received: October 7, 2002 Phosphogypsum obtained by "wet procedure" for phosphoric acid production is actually calcium-sulphate dihydrate, CaS04 • 2H20, which contains a high percent of impurities as well as radionuclides, originating from the basic raw material. Technology of phosphoric acid production by "wet process" intends for phosphogypsum to be a •·sponge" that absorbs all impurities from phosphoric acid (except uranium) and to be left in deposits as an unnecessary burden. As an enormous amount of phosphogypsum results from phosphoric acid production (the ratio is 5:1), phosphogypsum deposits are a burden per se, from the view of environmental protection, because they occupy large areas and may contaminate the air, ground water, but also surface water that receive phosphogypsum in the form of dilute suspension. By the proposed procedure, phospho gypsum is transformed into hemihydrate, CaS04 • Y2H20. Im the proces<> of removing radionuclides with barium-sulphate, 22~a isotope content decreases below the peimitted concentration, but this process is directly dependent on the type of the phosphate utilized in phosphoric acid production. Keywords: phosphogypsum, scanning electron microscopy (SEM), X-ray diffraction analysis, gammaspectrometric analysis, thermogravimetric analysis Introduction Phosphogypsum is the by-product of the phosphate fertilizer which is produced in large quantities world­ wide [1,2]. The "wet" prosphoric acid process is th.e most common one, used to produce phosphoric acid, it can be shown by the following reaction equation: [3] Ca10(P04)~2 + 10H2S04 + 20H20 -4 -4 6H3P04 + 10CaS04 • 2H20 + 2HF e.g. by scheme on Fig.l [8]. As it may be seen in Fig.l, phosphogypsum is separated on filters from phosphoric acid, which is then concentrated and subsequently utilized for fertilizer production. The phosphogypsum is disposed (Fig.2) [5], in the areas intended for this purpose (mines, opencasts) or is (temporarily) discarded as a very diluted (2 wt.%) slurry into natural water streems. Phosphogypsum represents a great problem for phosphoric acid producers. The problem of phosphogypsum arised at the moment as the frrst plant of phosphoric acid was established. working with .. wet process" technology, and the final solution of this problem could not be found yet. Namely, CaOIP20~ ratio in the phosphates used for phosphoric acid production, is 1,35-1,65, which means that the mass of dried phosphogypsum per 1 t of obtained P10~. ts between 4 and 5 t. This means that phosphogypsum, and not phosphoric acid, is the main product of the plant [ 6}. Phosphogypsum mainly consists of gypsum (CaS04 • 2H20), but it also contains small quantities of impurities, that may influence unfavourably the environment, as eg. heavy metals [7] and natural radionuclides [8}. The presence of impurities in phosphogypsum. especially of radionudides, may prevent its further utilization. Thus. its additional purification is necessary before any application (9.10). The components of phosphogypsum. originally present in raw phosphate, but are absent in na.tural gypsum, pollute the soil (as well as ground waters) and natural water streems [5]. Insoluble particles, sooner or later. are sedimented along the banks and shallows of the river bed, forming bigger accumulations. lessening the isolation of the bottom of the river and covering river flora and fauna with a thin layer of sediment Besides occupying large areas and preventing further development of flora and fauna, phosphogypsum. due to 220 Amlliloniated tripleSIIJ:l"r• ploOJpbafe * Phosphate rock" * Radioactive Fig.l The production of phosphate fertilizers Fig.2 Appearance of a phosphogypsum deposit - "white dead sea" the radionuclides which originate from phosphates, represents also a further source of radionuclides and their radioactive daughters, which may additionaly contaminate the environment [11]. For _that re?son, there are investigations in progress worldWide wtth the objective of exploitation of phosphogypsum as a raw material in construction industry This is represented schematically (optionally) in the Fig.J. All these were reasons to undertake investigations in order ~~ purify phosphogypsum and bring its compositiOn. closer to natural gypsum. To achieve the given objective. a new method was applied in phosphogypsum purification. and testing the quality. Chemical composition and physico-chemical analyses were performed using standard silicate analyses. as the method of alkaline melting, X-ray diffraction and thermo-gravimetric analysis. To compare the structure of natural gypsum with phosphogypsum, obtained immediately after filter separation in the plant of phosphoric acid production. as well _as after application of purification procedure. scannmg electron microscopy was applied. As the raw material ~ natural phosphates - contain radionudides (Fig./). they are also found in the phosphogypsum. For that reason. gamma spectrometric analysis of phosphogypsum ~vas performed. before and Fig.3 Scheme of application of phosphoric acid and optionally of phospho gypsum after the purification, and an equation is given for calculating the coefficient of radon liberation from phospho gypsum. Materials and Methods All investigations were performed with phosphogypsum, obtained from chemical plant in IHP Prahovo by a technological procedure in which raw ph~spha~es are treated with H2S04• Phosphogypsum PW:lficat~on w~s carried out by the new procedure, whtch will be dtscussed in further part of this paper. Gamma spectrometric analysis of phosphogypsum samples was performed as follows: phosphogypsum samples w~re homogenised, dried at 105°C (for 6 hours) and put. mto a container (marinelli) of appropriate geometnc shape and kept closed airtight (30 days) in order to achieve radioactive equilibrium. Gamma spectrometric measurements were performed by three pure germanium detectors manufactured by EG&G ··oRTEC'\ Germany~ with the efficiency of 25-30% and energy resolution 1,75-1,95 keV. The detectors were connected to a multi-channel analyser by the same ma~ufacturer and to corresponding computer eqmpment. Energy calibration, as well as calibration of detector efficiency were performed by radioactive standard supplied by Amersham. The measurement time for one sample was 60.000 to l 00.000 s and the basic radiation was measured after 250.000 s. • 100 .--------------- -o- NATURAL GYPSUM C ['/o] -+- PHOSPHOGYPSUM .95 90 c55 200 300 400 500 toe Fig.4 Thermogravimetric analysis of phosphogypsum samples Measurements of total activity were performed by ~;l3 anticoincidental proportional gas counter ( COUNTERMASTER") with basic radiation of 1 imp/min. Planchet radius was 2.3 em. Counter efficiency amounted to 24 % and was determined by a standard of 90Sr. Phosphogypsum and natural gypsum samples were analysed using scanning electron microscopy SEM, JSM-840A, JEOL, Japan. X-ray diffraction analysis was carried out using the diffractometer for powder SIEMENS D-500 with Ni­ filtered CuK... radiation. Identification of crystalline phases in recorded samples was carried out by position and intensity comparison of diffraction profiles with JC PDS data. Thermogravimetric analysis of phosphogypsum w~re performed on Dermatograph STANTON, England, With speed Of heating up tO 7 °C/min in air stream in mixture with Alz03. Samples were analysed in ceramic crucible, on temperature up to 500°C. Results and Discussion Thermogravimetric analysis of phosphogypsum samples were performed in air stream with speed of heating of 7°C/rnin, and the obtained results are ?resented at Fig.4, showing the change of sample mass m wt.% in dependence on temperature. The mass loss for phospho gypsum is 17.53 wt.% (gypsum loses 2 moi~ule of water), that would correspond to gypsmn portion of 84 wt.% in sample. For natural gypsum, which is hernihydrate (CaS04 • Y.ili20) the mass loss is 5.75 wt.% that would corresponds to gypsum portion of 93 wt.%. Temperature oc 80 60 40 20 40 611 Region Ill CaS04 Stable 221 CaSO 4 · t/2 H2o l!nstable Region! CaS04 ·H2S04 Stable The procedure of phosphogypsum purification is based on enhanced temperature. The conditions, temperature around 90°C and sulphuric acid concentration of 28 wt. %, were necessary for phospho gypsum (CaS04 • 2H20 - dihydrate) to be transformed into a hemihydrate - CaS04 • ~20, that corresponds to natural gypsum This is represented in Fig.5, the state diagram of calcium sulphate. From Fig.5 it can be seen that only under these conditions phosphogypsum is situated within region II, where the form CaS04 • %120 is dominant over CaS04 • 2H20. Phosphogypsum purification was carried out by solution of H2S04, with concentration of 28 wt. %. which was heated up to 90°C. Phosphogypsum was added to the solution and an emulsion was made with intensive stirring. After the addition of phosphogypsum, barium sulphate was added also with continuous stirring, until the temperature reached 85°C. Then the mixture was cooled. After cooling, it was filtered through a special Buchner funnel, separating the solid from the liquid phase. After drying and calcination. the grinding of the purified phosphogypsum was carried out by laboratory mixer. until the particle size reached 100 to 200 J.UU, only 12 % of the particles were bigger than 200 fltil. The results of analyss of chemical and mineralogic composition of natural gypsum. phosphogypsum taken from filter from the process of phosphoric acid production by "wet proceduren and previously processed and purified phosphogypsum by the new procedure, are presented in Table 1. From Table 1 it can be seen that phosphogypsum is in the form of dihydrate. immediately after separation from filters (74.16 wt.%). while the natural gypsum is hemihydrate in essence (81.34 wt.% CaS04·Y.zH20). By the proposed purifying procedure, phosphogypsum is tranformed into hemihydrate (77 .26 wt.% 222 Table 1 Chemical and mineralogical composition of natural gypsum, phosphogypsum taken from plant's pipe and purified phospho gypsum Natural gypsum Phosphogypsurn taken from Purified phosphogypsum plant's pipe Chemical composition CaS04·2H20 CaS04·V2H20 CaS04, anyhidride Bonded (fixed) water, H20 CaS04, total Free CaO so3 Free water, at 45°C Calcination loss Si02+insoluble residue MgC03 Na20 K20 Al20 3+Fe20 3 MgO CaC03 P20s Sum: Sum: Total: (in wt.%) 81.34 6.16 5.38 82.12 Excess: 0.70 88.20 Impurties: 1.30 0.98 4.15 0.06 5.09 11.58 99.78 Fig.6 SEM photograph of crystalline structure of natural gypsum (x500) CaS04·~ 0) and all the remaining parameter are far clo er to natural gyp urn than to phosphogypsum. To establish morphological compo ition and tru ·ture, their homogen ity and the pre ence of faults in the tru ture, an inspe 6on wa performed by anning electronic micros opy (SEM) of the natural gyp urn and arious pho phogypsum samples and the photograph are repre ented in Figs.6 to 9. SE photographs of natural gyp urn and phosphogyp urn show that th e two material regardle s of having the same chemical composition have different structures: natural gyp urn has poorly expres ed cry tal tructure and pho phogypsurn ha a marked cry tal tru ture mo tly of rhombi and he agonal forms hich indicate it more complex compo ition than that of the natura} gyp urn. Thi ha al o been confirmed by pre iou in estigation [1-,l ). (in wt.%) (in wt.%) 74.16 11 .54 77.26 18.02 7.21 56.14 81.59 0.77 0.29 74.93 89.09 0.34 1.03 2.18 0.60 21.75 4.55 0.04 0.17 4.48 0.84 0.09 25.15 10.92 100.08 100.01 Fig. 7 SEM photograph of crystalline structure of phosphogypsum taken from plant's pipe (xl,OOO) X-ray diffraction analysis of natural gypsum and phosphogypsum indicated the presence of impurities in phosphogypsum. The results confirm that the two materials are the same, but that phosphogypsum contains ingredients being absent in natural gypsum. The X-ray diffractograms obtained for natural gypsum and phosphogypsum are presented in Figs.JO and 11. X-ray diffraction analy is was carried out by comparing the position and intensity of diffraction profiles \-vith JC PDS data. The following crystalline phases were identified: for natural gypsum CaS04·~H20 (33-3 10) - identification number from standard· for phosphogypsum: CaS04·2H20 (33-311)· CaS04·YzH20 (33-310) spectra corresponded to natural gyp urn, CaHP04 (9- 0) or CaHP04·2H20 (9-77) pe tra al o corre ponded to natural gyp urn [14]. Fig.8 SEM photograph of crystalline structure of purified phosphogypsurn(x300) Fig.9 SEM photograph of crystalline structure of mixture of 75 wt.% natural gypsum and 25 wt.% phosphogypsurn (x500) The main difference between natural gypsum and phosphogypsum lies in their radionuclide content, although this content is variable, because it depends on the initial raw material- phosphate. Namely, phosphates formed from sediment phosphates (Florida, Marocco, Tunisia, Senegal) contain uranium in higher concentrations, while phosphates of volcanic origin (Cola, Caratau) either contain radionuclides in traces, or do not contain them at all. During technological processing of phosphates by wet procedure, due to degradation with H2S04, 14 wt.% of uranium from phosphates pass into phosphogypsum (and the residue into H3P04), while almost 80 wt.% od the present radium from the phosphate passes into phosphogypsum. Uranium in phosphate is in equilibrium with radium. Radium quantity may be calculated according to the equation: NJ4,u = NRJ!Ln,Ra where: N = number of atoms submitted to radioactive degradation; Ln = half-life of radioactive isotope. Oxidation state of uranium in phosphates and the mode of ore treatment determine uranium distribution between acid and phosphogypsum. Passage of uranium from phosphoric acid i in proportion with the use of P20 if ore dissolution is performed under oxidation condition [15]. In phosphates from Florida which contain U(IV) 223 Fig.JO Characteristic X-ray diffractogram of natural gypsum Fig.ll Characteristic X -ray diffractograrn of phospho gypsum fprm as uranium dominant form at normal conditions of processing by "wet process", between 60 and 80 wt.% uranium passes into the acid, and the residue into phospho gypsum, while for phosphate~ of African origin, where uranium is mostly in the form of U(VI . its participation in the acid may be almost 90 t.% [16). Process of uranium removal from phosphog psum In the available literature there are presented variou procedures for decreasing of radionuclide conte t in phosphogypsum [17-20], but no one could rem e completely that main "fault ' of phosphogyp urn in comparision with natural gypsum, because it is variable and is in direct relation with the type of pho ph ate u. ed as raw material [7]. The process for reducing radionuclide contamination in phosphogypsum, applied in this paper it is a primary object of our idea. It is a further object of this idea to provide a cour e purified gypsum product from phosphogyp urn containing radioactive contaminants It has now been discovered that the foregoing objects are accomplished for reducing the radioacti e contamination (radionuclide ) in phosphogyp urn in a process which comprise : • admixing pho phogyp urn ontaining radionu Jide with dilute ulphuric acid ~ontaining barium sulphate at an elevated temperat...l e t ~ rm an acidic lurry having a olid omponem compri ed of fin fra tion and a coar e fra tion and • separating the aid fine fracti n f ohd from the coar e fraction, 224 Table 2 Gamma-spectrometry results of samples unpurified and purified phosphogypsum by new process [21] Radioactivity (Bq/kg) Radionuclides Activity of unpurified phospho gypsum Activity of purified phospho gyPsum 439±20 379±20 8.7±0.5 7.1±0.5 Radionuclides artificial origin • whereby said fine fraction predominates in said barium sulphate and said radionuclides, and • whereby the coarse fraction predominates in phosphogypsum of reduced radioactive contamination. In the first step of purifying the phospho gypsum the termal decomposition of the phosphogypsum, from dihydrate to hemihydrate (in accordance with Fig.5), may generally be accomplished by dry kettle calcination of atmospheric pressure and temperatures about 115- 1600C or be dilute acid thermal dehydration such as in about an aqueous sulphuric acid solution having a concentration of 28 wt. %. at atmospheric pressure and temperature about 90°C, which is used to digest the phospho gypsum. In the second step of this process, sufficient barium sulphate is added to the slurry of phosphogypsum and sulphuric acid to provide a barium sulphate concentration in the resulting slurry. Excess barium sulphate may be employed, but it is generally unnecessary, since such an excess adds to the cost of carrying out the process without significantly improving the reduction of radioactive contamination (radionuclides). Barium sulphate is preferably added to a concentrated sulphuric acid solution containing from about 0.10 to 10 wt.% by weight of barium sulphate. Fuming sulphuric acid may be employed to dissolve the barium sulphate, if desired. Phosphogypsum was agitated with the solution of sulphuric acid and barium sulphate for a period of 12 minutes. while maintaining the temperature of the resulting slurry at about 85°C. After the desired elevated temperature is obtained. the slurry is agitated for a sufficient period to effect solubilization of the radioactive contaminations (radionuclides), followed by absorption and/or co-precipitation of radium sulphate crystals with barium sulphate crystals in fine divided fonn. This digestion period ranges from 5 to 250 minutes. After the reaction of the acid slurry was completed, the slurry is preferably cooled, at a temperature of 30- 350C and then filtered. If desired. the hot acid slurry can be washed without a separate cooling step. The acidic slurry with or without prior cooling is subjected to a solid-liquid separation step, such as filtration or cyclone separationt and the clarified acid is recovered. It may be recycled, after reconstruction. for use in reacting with additional imp¥fe phosphogypsum or used in other parts of the fertilizer process. The filter cake was washed with water and the solids were dried at a temperature of 60°C. The dried solids < 8.7 7.8±0.5 < 1.0 < 1.1 was slurried with water and then wet screened on a 100 micron screen. The solid phosphogypsum retained on the 100 micron was collected and dried at 60°C. The results of gamma spectrometric analysis of natural gypsum and phosphogypsum purified by the proposed procedure are represented in the Table 2. As it can be seen from the Table 2, activity of phosphogypsum purified by this procedure is decreased by approx. 10 wt.% in comparison with raw phosphogypsum. It is necessary to stress that the original raw material - phosphate determines much the content of radionuclides, which should be primarily taken into consideration when choosing the phosphate. The obtained results of radionuclide content analysis in raw phosphogypsum are in accordance with literature data obtained by investigation of 2~a isotope contents in various types of phosphogypsum, which range from 430 to 790 Bq/kg. The presence of thorium, as well as of the products of degradation of thorium sequence is not of any importance, because the activity originating from thorium is not high. . The presence of 226Ra isotope is dangerous because of liberation of (the only) radioactive gas radon (isotope-222), which is a daughter of radium degradation sequence (Table 3). The daughter product of 22~a, 222Rn, having a half-life of 3.82 days, and the extent to which this escapes from any biological system in which 22~a is deposited profoundly affects the dose. Radiation danger at the contact of human organism with radon is connected with its radioactive "short-lived radon· escendents": 218 Po 214 Pb 214 Bi and 214Po which mostly 84 , 82 ' 83 84 ' damage the lungs. A disintegration product later in series is RaD e10pb) with a half-life of 21 years; this element and those derived from it will not reach equilibrium amount in the life-time of a man but, for example, may be present in a person who has acquired an accidental body burden of 22~a. Determination of 222Rn emanition coefficient [22] After the 30 days equilibrium period, the total activity of 222Rn in secular equilibrium with 22~a. The activity of the total gaseous 222Rn in phosphogypsum (or sample} was calculated~ taking into consideration the volume of the vessel in which sample exist, the porosity of the phosphogypsum, and the volume removed during sampling. Porosity (P)~ fraction of total phosphogypsum was calculated by: 225 Table 3 Radioisotopes in the Radium Series Radioisotope Element Half-time Particle energiesa y-Ray energies (MeV) (historical name) (MeV) Radium Ra 1620 years a, 4.78 (94.3 %) 0.187 (5.7 %) a, 4.59 (5,7 %) Radon 222Rn 3.82 days a, 5.49 (99+%) 0.51 (0.075) a, 4.98 (<0,1 %) RaA 21sp0 3.05 min a, 6.00 (99+%) RaB 214pb 26.8 min ~-. 0,67-1,03 0.053-0.352 Rae 214Bi 19.7 min ~-. (99+%) 0.4-3.18 a, (0.04 %) ~c·(Ls)l 214p 0 160 ?sec a, 7.68 I RaC" (O.J %) 21171 1.32 min ~-. 1.96 0.32-2.36 Rk 210pb 21.4 years ~-. 0.017 (85 %) 0.047 (85 %) ~-, 0.064 (15 %) RaE 210Bi 5.0 days ~-. 1.16 (99+%) RaF zwp0 138.4 days a, 5.30 (99+%) RaG 206pb stable a Where the ~- or y- spectra contain many lines, only ranges of energy without abundances are given RaG - unradioactive lead (isotope 206Pb) or (radium's lead or RaG) where Db is bulk density (g!cm3 , phospho gypsum (solids+porosity) and Dp is particle densitl of phospho gypsum was assumed to be 2.32 g/cm , the value fos CaS04·2H20 [23]. The efficiencies of the scintillation cells were determined with a calibrated 222Rn source, which quantitatively delivered 3 Bq in an unspecified volume. The 222Rn emanation coeficient (c) was calculated by: total activity (Bq) of 222Rn in the gas phase £= ·100% total activity (Bq) of 226Ra in the sample Radioactivity is investigated as the contents of three radioactive isotopes: K--40, Ra-226 and Th-232. Biologically harmful evaluation of these three isotopes has the following relations: 40K: 22~a: 23Zrh = 1: 12,5: 16,50 which is not the case with phosphogypsum, because several times higher radioactivity originates from 22~a isotope, than from other isotopes {Table 2). From the Table 2, it rna~ be noticed that by the proposed procedure, isotope 2~a activity is decreased below tha maximal permitted concentration ( 400 Bq/kg) [24]. Upper permitted limits of radioactivity are calculated according to so-called Summary formula which has been formulated by National commission for protection from radiation in former USSR, and according to our currently valid regulations it has the following form [24]: 226 Ra (Bq/kg) Index for interiors = + 200 232 Th (Bq/kg) + .. 4(j K(Bq/kg) + V interiors (Bq/kg) = l 300 3000 4000 Index for exteriors = 226 Ra (Bq/kg) + 400 232 Th (Bqjkg) 40 K(Bqjkg) V~xteriors (Bq/kg) ___ __:::_-=.;.... + _ ___:_.........;:.:--=:..:... + = 1 300 5000 4000 Index for roads 226 Ra (Bq/kg) ::: 700 + 232Th (Bq/kg) + 40 K(Bq/kg) + Vrorroatls