STUDIES ON THE PRODUCTION OF PLASTER OF PARIS FROM DISCARDED MOULDS

Abrtroct .t Investigations have been made t o develop a simple method to convert discarded moulds in ceramic industry t o Plaster of Paris. With this in view, physical and chemical properties of used moulds, Plaster of Paris, gypsum and the products obtained by firing used moulds under different conditions have beea determined. Monoclinic form' of CaS04.2H20 has been identified in the used moulds while Plaster of Park contains 0 form of CaSO .HH20. DTA and XRD analysis show that a temperature lower than 3 5 0 ~ ~ but hi& than 1 7 0 ~ ~ is required for the conversion. Rapid heating rates are ~equired for the formation of 0 -CaSO ?4H 0 which gives 0 Q soluble anhydrite ai 170'~. At temperatures higher than 350 C, solu&e nnhydrite is transformed into an inactive form. Good quality Plaster of Paris could be produced by heating powdered moblds at 1 8 0 ' ~ for 2 houls in an open atmosphere. Breaking. strength, chemical composition and other properties of the product thus obtained conform to the standard requiremenix of comrpercial Plaster of Paris.


I. Introduction
Plaster of Paria is produced by cdcining gypsum.The main deposits of gypsum are found in United States, France, West Geniany, Poland and Austria.A large quantity of the world production of Plaster of Paris is used in the ceramic industry for the manufacture of moulds and as a cementitious material for construction purposes.2It is also used for the rrianufacture of wall boards and for dental and orthopaedic work. 2 In Sri Lanka, plaster of Paris is mainly used for making tableware and sanitaryware moulds in the ceramic industry and as a binding material for plastering purposes.In 1986, the total .production of modds at the Ceylon Ceramics Corporation, Piliyandala, was around 32-35 metric tons per month.After using the moulds for a certain period, these are mainly discarded as a waste material.Some of these have been used for the manufacture of writing chalk and for land filling.Plaster of Paris is imported to Sri Lanka at a very high cost and therefore, it is important to study the possibility of recycling the Plaster of Paris.In order to develop a method of recycling, it is essential to study the physical and chemical properties of used moulds and Plaster of Paris.
In the present ,study, investigations have been made in the laboratory scale, to determine the most suitable conditions of regenerating Plaster of Paris from used moulds.

Experimental
Used moulds were ground to pass the sieve (300 Pm, mesh No.50) and burnt under two different conditions, (i) in a muffle furnace and (ii) in a special vessel (Figure 1) using burners.The constituents in used moulds, Plaster of , Paris and the fired products were determined by powder X-ray diffraction with Cu K, radiation.Thermal analyses (DTA, TG ind DTG) of used moulds, gypsum, Plaster of Paris and the fired products were done using a Thermal Analyzer mbdel NETZSCH STA 409.Transverse breaking strengths have been determined by a standard procedure.5Chemical analyses of used moulds and the fired product for CaO, A1 0 , 'so3 were done using gravimetric methods.'FeqO was determined a by a tltrimetric method.'Na20 and KqO were estimated y using a Flame Photometer and MgO was determiqed by using an atomic absorption spectrometer.Si02 was deiermined by the HF treatment.'

Physical ~r o ~k r t i e s
of used moulds and Plaster of Paris

X-ray if fraction Analys<s . .
Powder X-ray diffraction patterns of crockery and sanitaryware moulds showed the presence of CaSO .2H20 in the monoclinic form.Other compounds~ were not found in Ltectable quantities.Imported Plaster of Paris showed the presence of hexagonal type of CaS04.0.5H20.XRD patterns of used mod& and Plaster of Paris are given in Figure 2.

Thermal Analysis
Thermal analyses of used moulds, gypsum and Plaster of Paris were carried out under different conditions and the DTA curves are given in Figure 3. DTA curve of gypsum showed two significant endothermic peaks.The large --It-* * .

HEAT
J ., endotherm at lower temperature ( 1 3 0 ~~) is due to the loss of 1.5 mdecules of X20 from CaS04.2H20.The reaction for the conversion is as follows.

(anhy drite)
A very small endothermic peak was observed at 7 3 0 ~~.This is probably due t o the decomposition of magnesite which is present as an impurity.
The DTA .curves of used mould (1,2,3,4,5 and 6) were similar to gypsum, but slight variations of the peak positions were observed.These results show that the CaSO 2H 0 form of the used plaster undergoes dehydration through the formazin 0 % CaSO 0.5H20.In addition to these two major endotherms, samples 4,5 and 6 wgch were analyzed at higher heating rates (lo0, 15O and 20° Clmin.)'showcd a small exothermic peak at 3 5 0 -3 8 0 ~~' and this is due to the phase transformation of soluble anhydrite (CaS04) to an insoluble form.It has been found that this kind of transformation is only possible for 0 form of C ~S O ~. O .~H ~O .~ There are two types of hemihydrates, a! and fl.The & form is more stable2 and.does not undergo is type of transformation These results indicate that the anhydrite is formed through the formation of P -CaSO4.0.5H20 at higher heating rates whereas at low heating rates (2 and 5OC/minute) the anhydrite is produced through the formation of a-CaS04.0.5H20.The exotherm at 350-380°C was found to be more sharp and the intensity increaseswith increase of heating rate.The insoluble form of CaS04 is an inactive form and. does not undergo rehydration as the normal active CaS04.Therefore temperature shoidd not be allowed to exceed 350°c, since the active form of anhydride is transformed into an inactive form at this temperature.

DTA curve of commercially available Plaster of Paris showed an endo-
t h e m at 170°C and a small exatherm at 350°c.These are due to the loss of 0,5 molecules of H 2 0 from CaSO 0.5H20 and the phase transformation of soluble anhydrite to an insolub f e form.This result indicates that the commercial type of Plaster of Paris constitutes fl -CaS04.0.5H20.TG curve of the used moulds showed one large and one small weight loss and these are due to the loss of 1.5.and.0.5 molecules of H 2 0 respectively.DTG curve also showed two sharp peaks due to these two weight losses.Figure 4 shows TG, DTG and DTA curves of the used moulds.
Table 1 shows the positions of the two endotherms and percentage weight losses of the used moulds.I t was found that the position of the endotherms depend on the heating rate of the sample.The peak temperatures of each reaction increases with the increase of heating rate and the separation between the two endothermic peaks was clearly observed when the heating rate was 2OC/minute.The percentage weight losses are almost the same in all the samples and the ratio of these two losses in each sample is roughly three.This is because the weight losses are associated with the loss of 1.5 and 0.5 molecules of N20 respectively.One sample analyzed at 2OC/minute was kept in an oven at 45OC for 4 hrs.before the analysis.The results showed .thatthere is no appreciable effect of preheating of the sample at this temperature on percentage weight loss. . . . .

Chemical Analysis of used Plaster of Paris Moulds and the Product
Table 2 shows the chemical composition of used moulds and the product obtained by burning at 180°C for 2 hrs.Major impurities such as Si02, Fe 03, A1203 and MgO were found in appreciable quantities.Na 0 and ~~6 were also found in detectable quantities.It has been observed ;hat the contents of bound H 0, CaO, MgO and SO3 show significant differences in the two samples.The 7 ow amount of bound H 2 0 in the product is due to the dehydration of i ~? molecules of H 2 0 The contents of CaO and SOg in the' product are high when compared to that of the used mould.This is due t o the loss of 1% molecules of H 2 0 hring'calcination process.Slightly higher quantity of the MgO in the product indicates the dehydration of hydrated MgO during 'thermal treatment.The amount of CaS04.21120 in the moulds and the amount of CaS04.%H 0in ihe product are 95.8% and 95.1% respectively.The contelits 0% CaS04.2H20 in moulds and CaS04.HH 0 in the product conform to the requirements of gypsum and Plaster 0 % Paris specified in ASTM: The samples obtained by burning at 1 4 0 ~~ and 180 C contain'the hexagonol type of CaS04.0.5H20.Then the dehydration occurs with increase of temperature and the samples obtained at 300°C and 550°C show the presence of CaS04.The XRD patterns were taken at room temperature although anhydrite is formed at 180°C, on cooling in air it transforms into hemihydrate.As such the XRD powder patterns of samples heated at 140°C and 180°C are similar showing the presence of hemidydrate in both cases.The XRD patterns of the fired products are shown in Figure 5 and their d values and the relative intensities are given in Table 3. 'Table 3 .X-ray Diffraction Data of Fired Samples 1 4 0 ~~1 2 hrs. 1 8 0 ~~1 2 hrs. 3 0 0 ~~1 2 hrs. 5 5 0 ~~1 1 .hrs.

~z r r n i n ~. i n a special vessel using burners
obtained at 1 8 0 ~~ is very close t o that of Plaster of Paris.Therefore this result shows that the most suitable temperature of burning used plaster t o obtain Plaster of Paris is 180°C.The strength increases upto 180°c and decreases with increase of temperature.It is significant that there is a sudden drop of strcngth when temperature rises from 3 0 0 ' ~ to 5 5 0 ~~.This is bccause thc activc form of CaSO converts t o an inactive fonn during this temperature range as shown in D A. Used mould samples were also burnt at 180 k 3 ' ~ for 1,2,3 and 4 hrs and breaking strength of the products were determined It was found that the breaking strength of the product increases with increase of duration of burning upto 2 hrs, and remains consta-nt.It has been observed that there is no appreciable effect of further burning beyond 2 hrs , on the quality of the product.These results indicate that the most suitable duration of burning of used moulds is 2 hrs.Breaking strength of the products are given in Table 5.  3 while the XRD pattern of the product is given in Figure 5.
It is apparent that the material produced by calcination of the used moulds at 180°C consists mostly of the hemihydrate.The calcination temperature 180°C is however higher than the dehydration temperature of hemihydrate.Therefore the product of calcination at 1 8 0 ~~ is probably the soluble anhydrite, which on cooling and exposure to air forms the hemihydrate.
Breaking strengths of the products obtained by burning the used moulds at different temperatures are compared with the breaking strengths of Plaster of Paris in Table 4.The transverse breaking strength of the product obtained at 180'~ (375g) is very similar to that of Plaster of Paris (392 g).The products obtained at other temperatures have much lower breaking strengths.Thus, it is clear that the optimum temperature of conversion is about 180°C.
Plaster of Paris can be produced by the heat treatment of discarded moulds under different conditions and the quality of the product thds obtained depends on the temperature and duration of burning.No app~eciable.differences have been observqd in the products obtained by the two methods of burning used in the present study.Burning in a kilfl requires high expenditure and as such open burning method seems to be more economical for a commercial process.DTA studies have shown that a temperature higher than 140°c and a rapid heating rate are required for thk formatjons of 0-CaS04.0.5H20 which under oes further dehydration at 170°C to give an an activq form of anhydrite.Since &e active form of CaSOp is transformid into inactive fonn a t about 3,50°C, the burning temperature should-be controlied below 350°C.
A temperature of 180°C w d a heating duration of 2 hours are found to be most suitable for the conversion of used maulds to Plaster of Paris.
Physical and chemical properties of the product thus obtained are very similar to th6se of commercial Plaster of Paris.Therefore, the product obtained under these conditions can be used for the manufacture of new moulds and as a cernentitious material for construction purposes.

Figure 1 .
Figure 1.Vessel used for burning used nioulds

Figure 2 .Figure 3 .
Figure 2. XRD patterns of used mould and Plaster of Paris 0

,Figure 5 .
Figure 5. XRD patterns of products obtained by burning used moulds N.B. * gives the most suitable burning condition

Table 1 . ~k a k Position and Percentage Weight Loss of the used Mould Heating Rate O~frninute Large endotherm (OC)
-Note: *This sample has been preheated at 4 s 0 c for a period of 4 hrs.

Table 2 .
Cllemical Composition of Used Moulds and the Product

Table 4 .
Breaking Strength of thc,Products and Plaster of l'aris

Table 5 . Breaking Strength of the Producs obtained by Burning Used Mould at 1 8 0 ~~ for Different Durations
fi CaSO Y2H 0 to form anhydrous CaSOe and the transformation of soluble anhy&te ?o the insoluble form respectwely.Relevant DTA curves are shown in Figure