Testing drinking water with a very long path cell
The quality of drinking water is determined by three main factors: (a) salt content (dissolved solids), (b) organic and chemical content, and (c) the content of microorganisms such as bacteria and viruses. The Environmental Protection Agency has established a series of testing methods to determine these contents in the specialized environmental laboratory. However, for on-line continuous monitoring, and for ordinary business and industrial plants, only item "a" can be readily determined by using a conductivity meter. Usually water is rated by "how many megohms" without knowing its organic and microorganism content. While most organic molecules exhibit strong absorbance in the UV range and most microorganisms scatter light strongly in the UV range, conventional UV/Visible spectrophotometers are not sensitive enough to detect significant differences between tap water, filtered water or distilled water.
WPI's new Liquid Waveguide Capillary Cell (LWCC) will significantly increase the sensitivity of UV/Visible spectrophotometers' sensitivity, making continuous monitoring of quality factors "b" and "c" possible. The difference between using a standard 1 cm path cell and a 50 cm long path cell to measure water absorbance can be likened to the difference between seeing water in a cup or the ocean.
To really "see" the difference in the water we use every day, we first took a spectrum of tap water coming from our county water treatment plant. As the reference, we used HPLC grade water (Sigma Chemical Co., #27,073-3). The LWCC is so sensitive that it was necessary to dilute the tap water ten times with HPLC grade water to allow the UV absorbance measurement to reduce to the spectrometer's absorbance range
.
|
|
Figure 1 displays the spectrum of 10-times diluted tap water. Much of the light scattering is due to colloids and microorganisms. There is also a small "bump" at 260 nm range probably due to the organic molecules in the water. Next, we took the spectra of some purified drinking water samples as shown in Figure 2. From the scale of Figure 2 we can see that these water samples are much cleaner than the county tap water shown in Figure1. The absorbance of the distilled water at 240 nm was only 0.3% of that of tap water. On the other hand, the difference between each treated water sample is very clear. Distilled water showed more absorbance in UV than HPLC grade water; probably, some low boiling point organic compound is distilled together with the water molecule. The reverse osmosis (RO) water from a vending machine did not show more scattering of light than the HPLC grade water. The baseline is very flat. However, there is an absorbance peak below 235 nm, indicating it contains some chemical molecules. To our surprise, RO water from a home drinking water system showed lower absorbance than the HPLC grade water in UV. The shape of the spectrum is a mirror of typical colloidal particle light scattering spectrum, indicating it contained even less colloid than the HPLC grade water.
|
|
Just as the conductivity reading doesn't reveal the content of specific heavy metal ions (but nonetheless provides useful information about the performance of the desalinating system), so likewise the UV spectrum measurement does not indicate which organic compound or bacteria is present in the water; however, very low absorbance of UV indicates that the water purification system has removed most of the organic compounds and microorganisms-very valuable information for on-line monitoring.
Clearly, WPI's LWCC is sensitive enough to detect the difference between different types of purified water. With the current technology the length of the waveguide cell can be built even longer for greater sensitivity. We predict that this technology may soon become a practical way for continuous on-line monitoring of water quality in food processing and other industrial plants.
-S.Y.L.
Fiber Optic Cables & Accessories