Science Inventory



A team of students and faculty from The Cooper Union and KNUST is developing a simple filter for the removal of fluoride ions from groundwater used by remote rural communities in the Developing World. Ingestion of large quantities of fluoride leads to dental fluorosis, a considerable problem in many rural communities in the Bongo District, in Ghana’s Upper Eastern Region.


Given that an essential component of the design is that it can be adapted for use throughout the world, the potential media investigated are those available in the regions containing fluoride contaminated groundwater. From the literature, wood charcoal, bone char, laterite and Moringa oleifera seeds were identified as potential media. To test the efficacy of these materials, it is self-evident that the most important procedure in this research is the accurate measurement of the concentration of fluoride ions in aqueous solution. For use in Northern Ghana, the method needs to be portable, require a minimum calibration procedure and ideally provide data under conditions of constant flow. We list below our major findings and outcomes:

  1. We have developed a reproducible procedure for the measurement of the free fluoride concentration in aqueous solution using an ion sensitive electrode (ISE). The electrode has a built-in preamplifier and a platinum resistance thermometer enabling correction for temperature. We have found that the ISE offers a reliable, efficient and robust solution for this measurement.
  2. During the development of this procedure, we have come to understand the nuances of the ISE and therefore believe that we are in a position to cope with, account for, and correct any unexpected behavior during our field investigation.
  3. We have developed procedures to characterize potential filter media using column and batch processing. We have found that particulate forms of laterite (from Pennsylvania and Bongo) and wood charcoal, and crushed Moringa oleifera seeds are all able to reduce the concentration of free fluoride ions in aqueous solution: the most efficient being laterite and the least efficient being charcoal.
  4. We have found that the magnitude of the reduction in fluoride concentration, after exposure to these materials, increases with decreasing particle size. We believe this indicates that the fluoride is removed from solution through a physio-chemical adsorption process.
  5. We have found that laterite and Moringa satisfy the requirements that potential adsorption media be indigenous and economic. We have also found that these materials do not require special preparation prior to their deployment as adsorption media. Bone char has been rejected for this part of the investigation since the communities, for whom the filter is being designed, consume little meat. Wood charcoal, never considered a serious candidate, for reasons of deforestation, was included to provide a standard since the absorptive properties of this material are well known and it can be obtained in pure form. We are in the process of replacing the wood charcoal with alumina for a standard.
  6. A design for a different culture and people is of little use without input and feedback from the stakeholders. Working with faculty and students from KNUST and a community in North Eastern Ghana, we have been able to learn something of the ways in which people in very poor, remote rural communities actually use water. They are teaching us about the technology gap that exists between them and us; we are learning about their preferred requirements for a water filter, their preferred procedure for recharging the filter media and have been able to observe at first hand, their usage of water.
  7. Working with the agencies and communities in Bongo, we believe that we have the necessary ingredients to develop an indigenous, sustainable filter for the removal of excess fluoride ions from groundwater. By the end of the project period, we intend to have prototype filters in use by the Bongo community – the quantity and type will depend upon the outcome of impending characterization of potential media and the logistics of monitoring their behavior. Based upon our preliminary findings we anticipate achieving our filtration goal of reducing 5 ppm to below 1.5 ppm.


Final Progress Report

Record Details:

Start Date: 08/31/2007
Completion Date: 07/31/2008
Record ID: 186616