current and anticipated research plans
I moved to the U of M after spending four years at ASU. Every time I have moved from one university to another in my career, I have added another dimension to my research. In going from Tennessee Tech to Miami University as a graduate student, I began to focus on drinking water disinfection instead of the chemistry of water treatment as a whole. In going from Miami University to St. Lawrence University, I began searching for alternative colorimetric reagents that could be used for drinking water disinfectants. In moving from St. Lawrence University to Arkansas State University, I added a research component of drinking water disinfection by-products to the lab. Upon arriving at the University of Memphis, I moved firmly, with the help of external funding from the American Water Works Research Foundation (Awwa-RF), into the area of on-line real time monitoring of drinking water distribution systems. This has largely been my focus for the past three years. More recently, I have become interested in microfabrication, microfluidics and practical approaches to miniaturizing analytical instrumentation. In this document, I will briefly discuss the projects that have been completed in my research laboratory, those that are currently underway and the direction I plan to take in the future.
Organic Dyes for Disinfectant Analysis
I had mentioned my work with organic dyes as sensitive and selective reagents for drinking water disinfectants. These projects made up a significant portion of my research effort when I was a doctoral student, as well as my post graduate research experience at SLU and at ASU. Since coming to U of M, we have not been working in this area as much as before, more recently, we have revived some of this work with an eye toward providing analytical alternatives for the developing world. For example, this summer my Project Seed student, Thirston Johnson started to work on a relatively familiar high school chemistry experiment. He extracted a dye from cabbage. Our twist on this project was to use the anthocyaninn dyes as a reagent for drinking water disinfectants. Thirston evaluated the bleaching effect (decolorization) that free available chlorine and chlorine dioxide (common drinking water disinfectants) had on the cabbage dye. He discovered that chlorine dioxide would effectively decolor the cabbage dye while free available chlorine did not. This means that people in the developing world could possibly use cabbage dye as a reagent to measure chlorine dioxide concentrations in treated drinking water. Thirston worked on developing relatively simple colorimetric procedures for chlorine dioxide measurements. He also worked with one of my graduate students (Lucy Thurston) in constructing an inexpensive light emitting diode (LED) spectrophotometer ttailored to allow for spectrophotometric analysis of cabbage dye concentrations. The decrease in the absorbance of the dye can be related to the concentration of chlorine dioxide in the treated drinking water. Ultimately, this work should lead to relatively simple and inexpensive alternative for drinking water analysis using a reagent extracted from cabbage and a LED spectrophotometer that can be constructed from commercially available electronic components for under $50.
Supported Capillary Membrane Sampling-Gas Chromatography
While at ASU, I developed a collaboration with Global FIA Inc using supported capillary membrane sampling-gas chromatography (SCMS-GC) to eliminate sample preparation in on-line monitoring studies of drinking water disinfectants. We have continued that work at the U of M. In fact, this research effort was a large factor in the awarding of the grant from AwwaRF. We have improved upon the prototype instruments built at ASU and have interfaced the probe with additional detectors since coming to U of M. The progress we have made is summarized below.
We had previously interfaced the SCMS-GC with the flame ionization detector (FID), the electron capture detector (ECD) and the pulsed discharge photoionization detector (PDPID) while I was at ASU. We have not pursued the FID work further at this point, though very recently, I have come to believe this approach might be very useful in the problem of endocrine disrupting compounds in the environment-an area of great interest to NASA, EPA and AwwaRF. I plan to pursue this possibility in the coming year.
As for the work with the ECD detector, since coming to U of M, we have constructed two additional SCMS-GC instruments that improve on this approach. We have published a peer-reviewed article in Talanta (Talanta, 63 (2004) 675-682) that demonstrates the utility of SCMS-GC for monitoring trihalomethanes in drinking water. Since then, we have constructed a prototype portable SCMS-GC-ECD in conjunction with our AwwaRF grant and carried out extensive studies with this instrument. These studies include very detailed and controlled experiments designed to evaluate possible interferences in the method as well as instrumentation considerations leading to full automation of the SCMS-GC-ECD. We have also used this portable SCMS-GC-ECD to carry out extended monitoring studies directly from the MLGW distribution system and compared the portable SCMS-GC-ECD to EPA Method 502.2 in side-by-side studies. This work is detailed in our recently submitted final report to AwwaRF.
We have continued our studies using the SCMS-GC with the PDPID. At U of M, we have recently successfully miniaturized the SCMS-GC-PDPID in a new approach that we have developed here at U of M-this approach is called "gas chromatography on a valve" (GCOV). We have used the SCMS sampling probe in conjunction with our GCOV approach to construct a miniaturized version of the SCMS-GC-PDPID originally build at ASU. We have combined work from ASU with our new studies and a manuscript detailing this work was recently accepted for publication in Analytica Chimica Acta.
We have also explored interfacing the SCMS-GC to a mass spectrometer (MS) so that we can confirm our other methods if the need arises. One of my graduate students who completed her MS degree at U of M (Ms Gang Cao) worked on this project. While this project is relegated to use as a confirmatory method in our current AwwaRF study (the MS simply cost too much to meet our criteria set by the AwwaRF project), we are continuing to explore this technique. Portions of this work are presented in our final report on our phase II activities under review at AwwaRF. A manuscript is in preparation detailing this work is currently in preparation while we are wrapping up some loose ends of this research project.
The Capillary Membrane Sampling Device
We have recently developed a device that is similar to the supported capillary sampling probe in the manner in which it can be used as a sampling device. We have published this research to date in two recent papers (Water Research, 39 (2005) 16, 3827-3836 and Analytica Chimica Acta, 2005, in press). This work involved three of my U of M doctoral students, Gija Geme, Michael Brown and Paul Simone. In this device, we do not support the capillary membrane on a metal probe-hence we have named this device the capillary membrane sampling (CMS) device. The greatest advantage of the CMS device is its versatility. It can be used as a sampling device for GC, flow injection analysis (FIA) and possibly other applications. The CMS device costs about 1/10 the cost of the SCMS probe and can be built with commercially available components. It also uses much less water during a monitoring study than the SCMS probe.
The prototype CMS-GC uses an ECD detector and has been evaluate as a possible method for on-line monitoring of THM4 directly from the Memphis drinking water distribution system. In comparative studies, the CMS-GC offers several advantages over SCMS-GC. We have carried out seven day monitoring studies and side-by-side comparisons to EPA Method 502.2 with very good results. We plan to use the CMS-GC approach as one of our methods for on-line monitoring this summer in Memphis and Houston. As with the SCMS-GC method, we have completed very detailed method detection limit, accuracy and precision studies of it use as an on-line monitoring method and are very close to completing a manuscript for submission to a peer reviewed journal. We expect to submit this paper in early summer.
We have used the CMS device to distinguish between volatile and non-volatile disinfection by-products. This approach has allowed us to develop a new on-line monitoring approach that used the CMS device in conjunction with FIA for measuring the total concentrations of the four THM species and the total concentration of the five HAA species. This is done by using two sample injection loops, one sampling from inside the capillary membrane for the THM4 species while the other samples from outside the membrane for a sample effectively enriched in the non-volatile HAA5 species on the outside of the membrane. In either case, the sample is injected into a flowing stream of mixed sodium hydroxide and nicotinamide, and then reacts to form a fluorescent product that is detected. This automated analyzer alternately measures total THM4 and total HAA5 in drinking water distribution systems. It has been compared in side-by-side studies to the standard EPA methods for THM4 and HAA5 with very favorable results. More detailed studies of the chemistry of the fluorescent reaction of nicotinamide with THMs and HAAs are underway.
Evaluating and Developing New Ion Chromatography Based Methods
We had originally hoped that traditional ion chromatography-meaning ion chromatography with membrane suppressed conductivity detection -- would be a useful tool for on-line monitoring of HAA5 species in drinking water (At the pH of drinking water. All five HAA species are dissociated). This seemed reasonable as we had reviewed two papers that indicated that IC-MSCD cold be used in drinking water samples. Over the past two years, Paul Simone and I have put a great deal of effort into developing IC-MSCD as an on-line monitoring method. Unfortunately, we came to the conclusion that this is not a possibility. We are preparing a manuscript detailing our findings in this area. In our research, we were able to get very good results with the method in laboratory water, but once we moved to real samples (Memphis drinking water) we experienced several issues. Ultimately, the problem comes down to one of relative analytical signal. We are trying to measure HAA5 concentrations at the single microgram/liter level in the presence of common ions such as chloride ion, nitrite ion, nitrate ion and sulfate ion among other that are present at the milligram/liter level. With the conductivity detector, one sees very large peaks for common ions and very small peaks for the HAA5 species. Resolution of the HAA5 is a demanding task, even in the relatively simple matrix of Memphis drinking water. We tried several approaches to get around these problems and none of them worked. For example, we tried to use commercially available pretreatment cartridges to remove some common ions. This worked well for halide ions and sulfate ion, but there are no commercially available cartridges to remove nitrite ion. Nitrite ion was giving us resolution problems with one of the HAA5 species. We tried a series of oxidation procedures to convert nitrite ion to nitrate ion, but none of them seemed to work. While one might have the opportunity to simply wait for nitrite ion to oxidize to nitrate ion in batch methods, this is not a possibility when one is trying to do on-line monitoring. Ultimately, we have shown that these issues are problematic in Memphis water. If the method will not work in the relatively clean matrix of Memphis water, there is no doubt it would be useless in more complex systems.
It turned out that all of our work with IC-MSCD was not in vain. Once we knew that the HAA5 species would react with nicotinamide (from our FIA studies discussed above), we believed it would be possible to use the post column approach to develop a new IC based method for HAA5 species. A significant research effort has been devoted to this goal. By the time this document is being reviewed, we will have submitted a manuscript detailing this new method for on-line monitoring of haloacetic acids. This is quite possibly one of the largest impact findings from our AwwaRF grant. The advantages of this method are that only the HAA5 species appear in the post column chromatogram. Sampling rates for this new post column-IC (PC-IC) method are ~one sample per hour-a significant improvement over the cumbersome EPA method.
Microfabrication and Microfluidics-A New Research Direction and Simplified and Practical Procedures for Analytical Chemists
Two of the hottest new topics in analytical chemistry are the related areas of microfabrication and microfluidic devices. I have become interested in this area as the natural progression of my earlier studies in FIA methods. Moreover, these areas are in line with my interest in miniaturizing analytical instruments for true portability. Unfortunately, many of the techniques require specialized and sometimes very expensive facilities in order to fabricate these micron-sized structures. Since I do not have easy access to such faculties, I have decided to develop techniques of microfabrication that do not require any specialized facilities. Last year we began to make progress in this area. My studies in microfabrcation and microfluidic devices will be summarized in the following section.
Our first approach was to develop procedures for making microchannels from the commercially available polymer polydimethylsiloxane (PDMS). PDMS can be bought in a commercially available kit containing two materials-the monomer and the polymerizing agent. Once the two components are mixed, the resulting mixture is a viscous liquid that can be cured to form a elastomeric polymer in about one hour in an oven or overnight if cured at room temperature. We cut pieces of aged fused silica GC columns into sections and use super glue to place them in a pattern on a perti plate. Once we pour the polymer in the Petri plate and allow it to cure, we can remove the elastometic polmer that contains the negative relief of the fused silica. In short, we have made the three walls of our channels. We can place a piece of glass, with drilled holes (using a Dremel tool and diamond bit) that is fitted with commercially available tubing ports on top of the PDMS structure to seal our channels. Using this approach, we have been able to make microchannels with widths of 200 to 500 microns. This is roughly ½ to ¼ the size of FIA tubing. With this approach, we are moving toward making practical "lab in a shoebox" sized instruments that are truly self-contained as opposed to many "lab on a chip" applications presented in the literature with micron sized channels on a microscope slide and a $100,000 laser sitting beside it to enhance detection. For detectors in our "lab in a shoebox" designs, we are using light emitting diode (LED)-based spectrophotometers and spectrofluorometers. These devices were built in house by Lucy Thurston under the guidance of Mr. James Aschberger and my post doctoral associate, Dr. Kyoo Jo. This collaborative effort is developing relatively fast. I expect a manuscript from this work on our simplified fabrication procedures to be competed by the end of summer 2005. Using the simple template approach along with the LED-based detectors, we hope to miniaturize the spectrophotometric and spectrofluorometric method we have developed for disinfectants and disinfection by-products. This new area should be particularly fruitful for publications and funding.
In a related area, many researchers that construct miniaturized devices for monitoring that are used in the body or in the environment are hampered by the formation of biofilms. In the body, the immune response will recognize a foreign body and form a cyst around it, sealing it off and in many cases ending the usefulness of the biosensor. As for environmental sensors, microorganisms will often use a sensor as a good place to set up housekeeping and colonize the sensor. The result is the same as in the body-a sensor that can no longer function. After attending several seminars in Biomedical Engineering this semester, I was exposed to the idea of making materials that will release over time inhibitors to the formation of these biofilms. We have started to develop methods for doping PDMS with disinfectant species so that the PDMS will release the disinfectant over time. Disinfectant species are well known as inhibitors of biofilm formation. To date, we have been able to dope PDMS with sodium chlorite solutions and after aging, chlorine dioxide is formed in the PDMS. We have carried out studies detecting the chlorine dioxide using UV spectrophotometry and have demonstrated its release into solution. Additional doping studies have been done using hypochlorous acid as the dopant. Preliminary studies with it indicate that it too can be released from PDMS over time.
Additional studies have focused on coating PDMS with chemicals to change the properties of the polymer. For example, we have doped PDMS with graphite to block stray light and used this technique in preparing a LED based fluorescence detector so that we could miniaturize the CMS-FIA methods discussed earlier (the Water Research paper). We have also been able to use the silver mirror test (Tollen's test) to coat the inside of PDMS microchannels. More recently, we have been working to make a polymethylmethacrylate coated channels in situ inside a PDMS constructed flow manifold.
I am working to get funding to enhance this research. I envision the establishment of a "practical" polymeric microfabrication laboratory that will allow us to make smaller channels-on the scale of 10 to 50 microns. With these channels, we would be able to carry our capillary electrophoresis on a chip. I would hope we might be able to adapt out post column IC techniques to post channel CE methods. We might be able to use our LED-based detectors in this research as well. If I can get the funding, I hope to be able to integrate a workshop once a year for area undergraduate and MS program faculty. The idea would be to have a two day workshop, where faculty at our regional universities (the universities we recruit our doctoral students from) would attend. Over the two day period, two experiments would be carried out. One would require no specialized facilities and could be done at their home institution. The other would require the use of our regional laboratory and the equipment we have-with the knowledge that they are welcome back to use the facilities when they need them. I hope to get NSF support for this laboratory and will be pursuing this goal in the future.