Seems TiO2+UV do not have effect toward HA removal

I almost finished my current short term goal. I did some tests, using membrane 50 kD, 15 kD. For both, I used same comparison, one using TiO2 coating plus UV, the other one without them, just naked membrane. Seems that percentage removal of TOC between TiO2+UV membrane and just naked membrane, better for naked membrane. Wow.. I keep wondering why..

I am currently still running on 15 kD, TiO2 coated+UV membrane to remove humical acid (HA), 9 ppm under HA pH, around 5-6. After I gain the data, I will compare the result, is it true that the use of TiO2+UV+membrane even produce worse result compare to naked membrane? I know that TiO2+UV can remove HA by breaking mechanism. Meaning that the HA will be removed by OH radicals produced by TiO2+UV results in smaller molecule of HA. If this is true, the combination with membrane by physical screening for HA, should have produced better TOC removal for HA. But why the lab result shows the opposite??? I keeps wondering.. However, I will post the result here, perhaps you have some ideas…

Several experiments:
5 kD, no TiO2, just naked membrane, pH 7, HA 9 ppm, TMP 10 psi (G)
50 kD, naked membrane, pH 4, HA 9 ppm, TMP 10 psi (A)
50 kD, Coating TiO2+UV, pH 4, HA 9 ppm, TMP 10 psi (B)
50 kD, naked membrane, pH neutral, HA 9 ppm, TMP 10 psi (C)
50 kD, coated+UV, pH neutral, HA 9 ppm, TMP 10 psi (D)
15 kD, naked membrane, pH neutral, HA 9 ppm, TMP 10 psi (E)
15 kD, coated+UV, pH neutral, HA 9 ppm, TMP 10 psi (F)
Total 7 experiments, actually there were 9, but both were failed.

Now, the results:

Permeability and Rejection

I put 5 kD since it does not have pair to compare.
A:B:C:D:E:F:G = 0.015 : 0.0145 : 0.0129 : 0.01 : 0.0131 : 0.0115 : 0.034 (gr/cm2.min.psi)

Let’s compare one by one. Between A and B, normal, the permeability is almost the same, so problem. Between C and D, D slightly higher permeability, means higher chance for water and HA penetration. That would means, D should produce higher DOC percent rejection. Note that, permeabiliy of C and D are lower than A and B, because C and D is membrane renegerated from A and B by chemical cleaning. Now, E and F, 0.0131 and 0.0115, respectively. The permeability is lower than A and B, simply because both have smaller pore size.

Now, let’s see the results. Between A and B, the permeability almost the same, but B is using TiO2 and UV, shoud have provided better removal for DOC and UV254.

and another one

The removal for UV254 is fairly equal, while the opposite for DOC removal. Majority of samples taken, were showing lower DOC rejection for membrane with TiO2 and UV. That surprised me a lot. Kind a unexpecting result. Ok, let’s move to another result, see if this result confirmed.

C and D, the permeability of D a bit lower than C, thus hypothesis would be DOC rejection should be higher due to low permeability. Note, this has not taken HA particle size distribution into account. Let’s take as well E and F. F is much lower than E, should have given better DOC and UV254 removal percentage. Notice that from C to F, the lower permeability is D, then F, C, the last E. It is seen than under normal condition, D should produce better rejection.

another graph:

F provide highest rejection rate, than C and D, fairly equal, than E!!! There must be something wrong with this! E is 15 kD, and with TiO2 and UV, at least its ability to remove DOC as high as naked 15 kD membrane. This just doesn’t make sense. In my opinion, 15 kD experiment has to be conducted once again. Why? Let’s take a look of A and B. The permeability is almost the same, with DOC and UV254 removal are better for naked membrane, my explanation would be the permeability of A is slightly lower. C and D, D has lower permeability, thus it gives slight better DOC and UV254 rejection rate. But 15 kD? the permeability of membrane with TiO2 and UV (F) was lower than naked membrane (E), thus it was supposed to be better rejection. My another opinion is, it is likely that TiO2 and UV do not enhance natural organic matter (NOM) removal represented by HA significantly. Because the residence time of water get contacted with TiO2, is really short around 34 seconds. Other reasons would be that TiO2 has higher affinity than that of the membrane. Thus it will provide higher permeability and lower DOC removal.

Flux

Now the flux. I will directly show you the graph.

In 50 kD coated + UV, initial flux was higher, followed by a decrease flux to a final flux almost similar from flux 50 kD naked membrane. So, in general, TiO2 coated membrane give better result in term of normalized flux for 50 kD, pH 4. The opposite for 50 kD and 15 kD. Seems that the result are not in favor of TiO2+UV. In 50 kD pH neutral, the flux of coating membrane slightly worse than naked membrane. Especially for 15 kD, it is clear that TiO2 and UV do not give any enhancement at all at this point.

However, there are several factors that contribute this kind of results. First, TiO2 is known to increase affinity within membrane surface, thus it fasten the rate of water penetrate toward membrane. Second, the contact time between feed water and TiO2 itself does not occur within reasonable time due to the water velocity, in this experiment, the contact time was only 34 seconds, which even be shorter after short period of time. It would be good idea if I try with longer contact time, thus reduce the water velocity.

Hah.. That is why I need to do the research again to confirm the result.