Process Biochemistry 35 (1999) 391 395 www.elsevier.

com/locate/procbio

A preliminary evaluation of turbidity removal by natural coagulants indigenous to Venezuela

Received 20 April 1999; received in revised form 19 May 1999; accepted 12 June 1999

Abstract The ability of two plant materials, Cactus latifaria and the seeds of Prosopis juliora, to act as natural coagulants was tested using a synthetic water formulated to resemble the drinking water supplied to the city of Maracaibo. Turbidity was added as kaolin. The coagulation ability of the two materials was assessed by the use of standard jar test measurements. Both materials produced comparable turbidity removals and were able to produce a nal water whose turbidity was close to the required standard of 5NTU with both high (100200 NTU) and low (30 40 NTU) initial turbidities. This was comparable with the performance achieved by previous workers using Moringa oleifera extracts. The optimum coagulant dose was found to be lower than that for aluminium sulphate. Treatment of the raw vegetable solids with solvents produced coagulants which, in some cases, were different from the raw materials and suggested that non-polar material might be involved with the coagulation process. 1999 Elsevier Science Ltd. All rights reserved.
Keywords: Natural coagulants; Cactus latifaria ; Prosopis juliora ; Turbidity removal

1. Introduction The production of potable water from most raw water sources usually entails the use of a occulation/ coagulation stage to remove turbidity in the form of suspended and colloidal material. Aluminium and iron salts are the chemicals most commonly used together with synthetic organic polymers. With aluminium salts, there is always the concern about residuals in the treated water [1] and Alzeimers disease and, whilst iron salts are a cheaper option, the cost of any imported chemicals can be a serious problem for developing countries. Thus, in recent years, there has been considerable interest in the development of natural coagulants. Aqueous extracts of the dry seeds of Moringa oleifera are one such product and much attention has been paid to ways in which these extracts can be used [2,3]. These
* Corresponding author. Tel.: + 44-121-4145049; fax: + 44-1214143675. E -mail address: c.f.forster@bham.ac.uk (C.F. Forster)

studies have shown that a crude water extract compares quite favourably with alum and, as such, its use has been recommended for use as a water treatment agent in developing countries [4,5]. Similar studies have been carried out with other vegetable materials. These include extracts of Okra and Nirmali seeds [6], and Tamarind [7]. Two plants, which are indigenous to Venezuela, which are common in Venezuela and which have properties similar to those described previously for natural coagulants, were chosen for testing. These were the mesquite bean (Prosopis juliora ) and Cactus latifaria. The properties of the mesquite bean and some of its components have been examined for a number of reasons other than coagulation. A protein concentrate made from the bean has been considered as a food ingredient [8] and mesquite gum has been evaluated as an alternative to Acacia and Tragacanth gums as a binder in tablet formation [9]. Because of its ready availability in Venezuela, it seemed reasonable to assess its potential as a coagulant. This paper reports the results of a preliminary study using crude preparations

0032-9592/99/$ - see front matter 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 3 2 - 9 5 9 2 ( 9 9 ) 0 0 0 8 5 - 0

392 Table 1 Extraction procedures Extraction Methanol Ethyl acetate Petroleum ether

A. Diaz et al. / Process Biochemistry 35 (1999) 391395

Volumes used (ml) 410 310 210

of Prosopis juliora and Cactus latifaria as coagulating agents.

2. Materials and methods Prosopis juliora pods were collected in the city of Maracaibo and the Cactus latifaria was collected from the Can ada Urdaneta District. Both materials were processed at the University of Zulia. They were washed and dried at 80C. The resultant solids were milled (ThomasWiley, Model 4) and sieved to give a fraction of 6001300 mm. These were then used as a raw coagulant by suspending them in water (5% w/v). The milled solids were also treated by solvent extraction (Table 1). Methanol was used to remove polar substances, petroleum ether (45 50C) was used to remove non-polar lipids and ethyl acetate to remove substances with an intermediate polarity. In all cases, the solids and solvents were contacted for 5 min with gentle agitation and the solids were then removed by centrifugation (5 min; 270 g ). The solids were dried at room temperature for 24 h after each extraction. The treated solids were then used to prepare 5% (w/v) suspensions. A synthetic water was prepared so that its composition was similar to that of the water supplied to the city of Maracaibo (Table 2). A stock suspension of kaolin with a concentration of 2 g/l was used to provide turbidity in the range of 20 to 200 NTU. This range was chosen so that the results could be compared with data from previous studies using Moringa oleifera [4,10]. A standard jar test apparatus (Phipps and Bird, Model 300) was used in the coagulation tests. Samples of the synthetic water (1 l) were stirred at 80 rpm for 2 min and during this time the coagulants were added
Table 2 Composition of the synthetic water Component KCl NaHCO3 MgCl6H2O CaCO3 pH Concentration (mg/l) 32.85 125.70 40.00 75.70 7.0

Fig. 1. Effect of altering the dosage of the raw Cactus latifaria coagulant with initial turbidities of 50 NTU ("), 80 ( ) and 100 NTU ( ).

from a pipette to give the required dose. The stirring speed was then lowered to 30 rpm for 30 min after which the samples were allowed to stand for 20 min. The turbidity of the supernatant liquors was then measured using a turbidometer (Orbeco Analytical Systems, Model 965-10). Replicate measurements showed that the standard deviation of the results was 9 15%. Aluminium sulphate was also tested. This gave control data against which the performances of the vegetable materials could be assessed.

3. Results and discussion The effect of altering the dosage of the raw Cactus latifaria coagulant is shown in Fig. 1 for waters with three different initial turbidities, 50, 80 and 100 NTU. In each case, a dose of 1020 mg/l was optimum with the nal turbidity being less than or equal to 10 NTU. Fig. 2 presents the data for the effect of the raw

Fig. 2. Effect of altering the dosage of the raw Prosopis juliora coagulant (initial turbidity = 30 NTU).

A. Diaz et al. / Process Biochemistry 35 (1999) 391395

393

Fig. 3. A comparison of the turbidity removal by organic coagulants (data for Moringa oleifera from 4).

Prosopis juliora coagulant on a water with an initial turbidity of 30 NTU which was typical for the performance of this coagulant. Again the nal turbidity was low, less than 5 NTU which is the standard for drinking water, and the optimum dose was 20 40 mg/l. Both these coagulants gave a high removal of the initial turbidity and, as such, can be considered as potentially useful candidates for further examination. However, they do need to be measured against the yardstick of other natural coagulants which have been evaluated previously. The material which has received the greatest degree of attention is the aqueous extract of Moringa oleifera seeds. Muyibi and Okuofu [10] used it to treat waters with turbidities of 23 90 NTU obtained from three surface water sources in Nigeria. They found that, on average, a 50% removal of turbidity could be obtained when the Moringa oleifera extract was used as the primary coagulant. The optimum dose for coagulation with an extract using only the seeds of Moringa oleifera has been shown to be comparable to that of alum, 50 mg/l [5]. Ndabigengesere and Narasiah also worked with an aqueous extract of Moringa oleifera seeds but they used a synthetic water with the turbidity (105 NTU) added as kaolin [4]. The results of their study are shown in Fig. 3, together with the data from this current study. This comparison demonstrates that the three materials are very similar in their performance and that the Prosopis juliora and Cactus latifaria solids are worth further evaluation. This is conrmed by the results in Fig. 4 which shows a typical set of data for aluminium sulphate and one of the vegetable coagulants, Cactus latifaria. The nal, plateau turbidity values were very similar although the dose of aluminium sulphate which was required to achieve this was greater than that of the Cactus latifaria coagulant. This is comparable with previous work with extracts of Okra which found that the natural coagulant gave a better performance than alum [6]. The work which was reported by Ndabigengesere and Narasiah [4] was based on a synthetic water with

Fig. 4. A comparison of the performances of aluminium sulphate (") and Cactus latifaria ( ).

an alkalinity of 53 mg/l CaCO3 and showed that the addition of a range of different cations had little effect on the nal turbidity of the treated water. However, they did show that the addition of different anions had a slight effect on the nal turbidity. The alkalinity of raw waters will vary and, therefore, the effect of the Prosopis juliora coagulant was tested on waters with an initial turbidity of 200 NTU but with a range of alkalinity values (Fig. 5). This showed that the nal turbidity did increase slightly as the alkalinity of the water increased, although it was not possible to determine from the published data whether the two increases were comparable. Earlier work with Moringa oleifera has suggested that the presence of ions is necessary for coagulation to occur [4]. This work also indicated that there were some ion concentrations which resulted in

Fig. 5. The effect of alkalinity on the nal turbidity (initial turbidity (initial turbidity = 200 NTU) using doses of Prosopis juliora of the 20 mg/l (") and 40 mg/l ( ).

394

A. Diaz et al. / Process Biochemistry 35 (1999) 391395 Table 3 Composition of Prosopis juliora and Cactus latifaria Parameter Prosopis juliora (% DW) [11] Seed Grease Protein Fibre Ash 2.1 35.0 6.6 5.7 Pods 2.8 10.1 27.5 3.7 8.0 11.6 13.8 20.2 Cactus latifaria (%DW) [12]

Fig. 6. The effect of pH on the nal turbidity using doses of Prosopis juliora of 20 mg/l (") and 80 mg/l ( ) (initial turbidity = 200 NTU). Data for Moringa oleifera ( ) from [4].

there being a less efcient removal of turbidity. However, the coagulant dose was xed and there was no information about how variations in the coagulant dose:ionic concentration ratio might affect the nal turbidity. Obviously the role of ions will need further investigation. Conventional inorganic coagulants are used within specic pH ranges and their addition to aqueous suspensions can affect the pH. Earlier work has shown that the use of Moringa oleifera does not cause alteration in pH [5] and that there is no easily recognisable pH optimum [4]. Fig. 6 shows the effect of pH changes on the behaviour of Prosopis juliora and compares the data with the results previously reported for Moringa oleifera. Whilst it is clear that pH changes do alter the nal turbidity, in terms of the overall removal, the changes are not appreciable and suggest that there is no reason to consider using Prosopis juliora at a pH other than neutral. The general similarities in performance between Moringa oleifera and the two coagulants being evaluated in the current study are sufcient to suggest that their modes of action might be similar. It has been shown that the active agent in Moringa oleifera extracts are dimeric cationic proteins [5]. Both Prosopis juliora and Cactus latifaria contain protein in varying amounts (Table 3) and the protein from Prosopis juliora has been considered previously as a potential food ingredient [7]. The supposition that the modes of action might be similar does not, therefore, appear too improbable. Solvents were used to extract components from the Cactus latifaria solids and the effect of these extractions on the removal of turbidity (initial turbidity = 50 NTU)

is shown in Fig. 7. The effect of using the solids which had been treated with ethyl alcohol and with methanol was not really any different from the effect achieved with the raw, untreated solids. However, the data for the petroleum ether treated solids were signicantly different, with only a small percentage of turbidity removal being achieved. This implies that there is a non-polar component which has a signicant role in the coagulation process. A comparable set of data for the Prosopis juliora solids is given in Fig. 8 and shows a distinctly different pattern of behaviour. The solids which had been treated with methanol gave a turbidity removal which was not signicantly different from that achieved with the untreated solids. The solids which had been treated with petroleum ether gave a removal which was slightly worse than that achieved with the untreated solids, again implying that a non-polar component is involved in the coagulation process. In all these three cases, the turbidity/coagulant dose curves were typical of those described for coagulant sorption and charge neutralisation [13,14]. However, the data for

Fig. 7. A comparison of the effect of untreated Cactus latifaria coagulant () and solids subjected to the solvent extraction using ethyl acetate ( ), petroleum ether ( ) and methanol (") (initial turbidity = 50 NTU).

A. Diaz et al. / Process Biochemistry 35 (1999) 391395

395

Their optimum dose was 10 mg/l which is comparable with that achieved by extracts produced from Moringa oleifera. This performance meant that they could be used at lower doses than aluminium sulphate. References

Fig. 8. A comparison of the effect of untreated Prosopis juliora coagulant () and solids subjected to solvent extraction using ethyl acetate ( ), petroleum ether ( ) and methanol (") and methanol (") (initial turbidity = 50 NTU).

the ethyl acetate treated solids were typical of the condition where there was a degree of charge reversal. Since the concentrations of kaolin and coagulant were the same in all the cases described in Fig. 8, this suggests that the ethyl acetate treated solids had a slightly higher positive charge than the raw solids. Taken overall, the results show that the solvent-treated materials, which would incur a greater preparation cost, do not offer any advantage over the raw coagulants. 4. Conclusions The results show that: Cactus latifaria and the seeds of Prosopis juliora have the potential for use as water treatment coagulants.

[1] Miller RG, Kopfter FC, Kelty KC, Strober JA, Ulmer NS. The occurrence of aluminium in drinking water. J Am Water Works Assoc 1984;76:84 91. [2] Jahn SAA. Using Moringa seeds as coagulants in developing countries. J Am Water Works Assoc 1988;80:43 50. [3] McConnachie GL. Water treatment for developing countries using bafed-channel hydraulic occulation. Proc Inst Civil Eng Water, Marit Energy 1993;101:55 61. [4] Ndabigengesere A, Narasiah KS. Inuence of operating parameters on turbidity removal by coagulation with Moringa oleifera seeds. Environ Technol 1996;17:1103 12. [5] Ndabigengesere A, Narasiah KS, Talbot BG. Active agents and mechanism of coagulation of turbid waters using Moringa oleifera. Water Res 1995;29:703 10. [6] Al Samawi AA, Shokralla EM. Investigation into an indigenous natural coagulant. J Environl Sci Health A 1996;31:1881 97. [7] Bhole AG. Relative evaluation of a few natural coagulants. J Water Supply Res Technol 1995;44:284 90. [8] Holmquist Donquis I, deRey GR. Functional properties of mesquite bean protein. Arch Latinoam Nutr 1997;47:343 51. [9] Khanna M, Dwivedi AK, Singh S, Soni PL. Mesquite gum (Prosopis juliora ): potential binder in tablet dosage forms. J Sci Ind Res 1997;56:366 8. [10] Muyibi SA, Okuofu CA. Coagulation of low turbidity surface waters with Moringa oleifera seeds. Int J Environ Stud 1995;48:263 73. [11] Marangoni A, Alli I. Composition and properties of seeds and pods of the tree Prosopis juliora (DC). J Sci Food Agric 1998;44:99 110. [12] Velez Boza F, Chavez F. Los cactus en Venezuela. Ed Caracas 1978;11:46 68. [13] Tenney MW, Stumm W. Chemical occulation of microorganisms in biological waste treatment. J Water Pollut Cont Fed 1965;37:1370 8. [14] OMelia CR. Coagulation and occulation. In: Weber WJ, editor. Physicochemical Processes for Water Quality. New York: Wiley Interscience, 1972.