**2.2 Water permeability and juice recovery (%)**

Before each MF process, the membrane water permeability was measured at transmembrane pressures of 0.5, 1.0 and 2.0 Bar, respectively, to verify its recovery, according to the following formula:

Organic Acid Lime (*Citrus Latifolia* Tanaka) Clarified by Microfiltration 305

299

percentage of peels in the BLJ samples. However, a yield of 47% was previously reported by Pedrão et al. (1999) for conventional lime juice, cv. *Tahiti*. Other factors may affect juice yield such as crops with high concentrations of nitrogen and potassium, leading to increased peel thickness, as well as relative humidity, and ambient temperature (Chitarra et al., 1990). The maturity stage of the fruit also influences the yields of acid lime juice that have been

Mean water permeability flux was 1,241.95 L.m2.h (± 0.48) in the three processes using the

In order to determine the best operating transmembrane pressure for all MF processes, 8.0 L of CLJ were initially used in the feed, resulting in 6.0 L (75%) of clarified juice. The mean fluxes obtained at transmembrane pressures of 0.5, 1.0 and 2.0 bar were 49; 47 and 35 L.m2.h, respectively (Fig.2). A slight difference was observed between the fluxes at the applied transmembrane pressures during the first twenty minutes of the process, but after this time there was a significant flux decline for processes operated at 1.0 and 2.0 Bar, demonstrating that concentration polarization was the result of polarization on the membrane surface and/or fouling, providing a decrease in the clarified juice flux over time (60 minutes) in all cases, which was more pronounced at 2.0 Bar. For both UF and MF processes, flux decline is high even when the water permeability of the membrane is completely recovered by

observed, ranging from 55.60% in green fruit and 59.40% in ripe fruit (Ziena, 2000).

Fig. 2. Effect of process mechanism on the permeate flux at 0.5, 1.0 and 2.0 Bar

declined gradually declined until the end of the process, probably due to fouling.

Thus a 0.5 Bar transmembrane pressure for 60 minutes was adopted to process BLJ by MF, resulting in a mean volume of 6.0 L of clarified juice from a 8.15 L feed and yields of 73, 76 and 54% of clarified juice in the first, second and third processes, respectively (Fig. 3). In the first and third process, after the first 20 minutes there was a decrease in the flux that

transmembrane pressures with conventional lime juice.

**3.2 Water permeability and permeate fluxes** 

0.3 μm membrane.

washing (Oliveira et al., 2006).

$$\text{Flux} \left( \text{L.m}^2 \,\text{.h} \right) = \frac{\left( \text{L} \,\text{/h} \right) \times \text{P} \left( \text{Bar} \right)}{0.05 \,\text{m}^2}$$

Where: 0.05 m2 = total membrane area; P = applied transmembrane pressure

The water permeability (L.m2.h) was measured with a chronometer, at 10 minutes intervals. Cleanup procedures (alkaline and chlorine-alkaline) were performed before and after each process, aiming to recover water permeability of the membrane.

Membrane alkaline cleaning was performed by adding 80 mL of a NaOH solution (Tedia, Rio de Janeiro, Brazil) (1N/pH 11) to 10 L of distilled water at 40 ºC, and re-circulating the solution in the system for 30 minutes. The solution was then discarded, using 80 L of distilled water. Chlorine-alkaline-cleaning was performed by using 40 mL of NaOH (1N/pH 11) and 130 mL of NaClO (Tedia, Rio de Janeiro, Brazil) (6% sodium hypochloride) diluted in 10 L of distilled water at 40 ºC, with recirculation in the system for 30 minutes, followed by washing with 80 L distilled water. After each cleanup, pH was checked.
