Since it is easier to get rid of ammonium than nitrate in tertiary treatment (the ahoe plant makes the solution basic with lime and blows out ammonium gas), I would be desirable to have on-line monitors of nitrate to be able to stop the eration process before excessive conversion to nitrite. Sewage is of such variable oncentration between night and day, week days and weekends, that the optimum eration time will greatly vary.

Current analytical methods for nitrate in the parts per million concentration nge are not ideal. Automation of the colorimetric techniques has helped but as not essenetially reduced the need stressed by the Frank Long Committee of e A.C.S. and by Sawyer and McCarty for "a more refined and exact method of alysis."

The Turkevich moon box used two nuclear processes for analysis of the light ements. Firstly, they measured the back-scattered alpha spectrum from Curium2 alphas, making use of the billiard-ball collision laws that make the maximum ergy of back-scattering an increasing function of target mass-going from zero r atomic mass four up to full energy for masses large compared to four. Secondly they extracted secondary information by measuring proton spectra ising from (a,p) reactions on the material.

59-068 0-71-pt. 3-23

Tony Turkevich remarked to John when our ideas were first discussed with him, "The method is fairly slow." Speed of analysis was not a prime consideration in the lunar application, where the telemetered data stream was shared with various other measurements.

For the sake of faster analysis, simplicity of design, and freedom from interferences we favor now for a nitrogen or phosphorus analyzer a very different arrangement from the lunar box. We propose to take proton spectra only and operate with such close spacings that pumping a vacuum becomes unnecessary. We have placed the sample to be analyzed on filter paper with a Po10 alpha source immediately above it and an ORTEC silicon-gold surface barrier detector directly below the filter paper. The beauty of the a,p reaction with alpha emitter energies up to as high as the 6 MeV from curium sources is that in the most common elements, carbon and oxygen, the a,p reaction is energetically forbidden. At 6 Mer alpha energies the Coulombic repulsion for nuclei of atomic number above potassium is so great that (a,p) reactions will not occur. At 4.5 MeV alpha energy, there is a resonance in the nitrogen cross section where one might operate with the cheaper polonium source; at this lower energy the Coulomb repulsion will cut down drastically on (a,p) reactions from elements of higher atomic numbers than nitrogen itself.

The scientific research aspects have been thoroughly done by Patterson, Turke vich, and Franzgrote, as the next two slides from their papers illustrate. These slides show proton energy spectra at back angles from thick samples, using a Curium-242 source. Events from carbon at the top are essentially zero, and can only be due to slight impurities. Nitrogen gives a low energy spectrum, consistent with the energetic requirement than 1.19 MeV less energy is released in the proton emission than comes in from the alpha particle absorption. (We refer, of course. to center-of-mass energies.) Fluorine, sodium, and aluminum give characteristic proton energy spectra going to much higher energy endpoints. The spectrum from magnesium is similar to nitrogen, but at the lower Po0 alpha energies we propose for a nitrogen sensor the reactions from magnesium will be much reducer.

Fig. 8. Protons from (a, p) reactions from various elements. These data were obtained with the prototype instrument using a source strength of approximately 1.4 X 10" d/m of Cm". The detectors were covered with gold foils (approximately 15 μ thick) to remove the scattered a particles. The background in the system has already been subtracted. Data are presented for (4) carbon in the form of graphite, (B) nitrogen in the form of NINO,, (C) fluorine in the form of teflon, (N) sodium in the form of Na:CO,, (E) magnesium metal, and (F) aluminum metal.

FIGURE 3

The next slide shows additional proton spectra from their work, adding phosphorus and silicon to the elements of the previous slide.

We had hoped to be able to show a comparable library of proton spectra at 4.5 MeV incident alpha energy, but by mistake the polonium source we bought evidently had a much thicker mica window than we had specified. The energy was insufficient for (a,p) reactions, though we were able to check out our detector and pulse height analysis system on the alpha backscattering plateau from lead. lead.

FIG. 8. Response of the alpha scattering instrument in proton mode to three pure elements. The data are for aluminum, silicon, and phosphorus, The ordinates are events per channel per 1000 min as a fumetion of channel_number. The source strength was 6.72X10)1o dis/min. Background has been subtracted.

FIGURE 4

It seemed clear to use then that the application of the a,p reaction method to analysis for some of the light elements was not a matter calling for further research, but rather a developmental and hardware project and it seemed appropriate for an aerospace company, such as the Space Vector Corporation, to carry out preliminary experiments and propose a production of proto-type instruments with assistance from some federal agency. The practical problem then is one of optimum sampling methods. We have considered many schemes beginning from simple evaporation of water samples and anlysis of the residue. Generally, the main dissolved constituents of water are not of a great deal of interest in the pollution problem, and indeed sodium in large amounts will interfere. Thus, the problem is to prepare a sample containing the nitrogen compound of interest on a metal foil or a piece of filter paper with the nitrogen being at least several per cent of the weight of the material in order to get reasonable sensitivity. It may be necessary to remove species, such as, magnesium or silica, which might contrib ute low energy proton spectra difficult to distinguish from that of nitrogen. In an earlier slide we showed the concentration of a few constituents in the influent and effluent solutions from the secondary processing stage. Ammonium was in the range of 10 mg/1 and nitrate in the vicinity of 1 mg/1. Major dissolved constituents are sodium, chloride, sulfate, and bicarbonate ions all in the 200 mg/1 range. Magnesium and silica are at about 20 mg/1 level. In the sample preparation stages, we have wished to avoid elaborate wet chemistry, for then the hopedfor advantages of the method for automatic applications may be lost. At any rate. let me give an example of a specific sample-preparation we would suggest here for application to ammonium and nitrate analysis in sewage plant solutions. (Sludges and solids need only be spread on filter paper to be analyzed.)

The next slide is from the FWQA Primer on water treatment and is intended to illustrate electrodialysis as a potential means of desalting tertiary effluent so the water can be re-used. You have all read, and many of you know much more than I, about the electrodialysis methods that have been studied as a means of obtaining pure water from brackish waters. Membranes are commercially available which quite selectively will pass either cations or anions. We propose for the basic analysis equipment to employ an electrolytic cell with a mercury anode and a mercury cathode. The sewage process solution be analyzed would be in a central compartment separated from the smaller anode compartment by an anion selective membrane and from the cathode compartment by cation selective membrane. As the voltage is applied to the cell, the central compartment will gradually be depleted of ions. The cations concentrate in the cathode compartment. The solution in this compartment may, after a specified length of time, be made strongly basic and ammonia evolved by heating. The ammonia could be converted to an ammonium salt by drawing the vapor through a filter paper moistened

either with a mineral acid, such as hydrochloric acid or an organic acid such as oxalic. Another alternative would be the adsorption of the ammonia on a thin charcoal disc. The filter or charcoal plate would then be placed between the alpha source and proton detector and the proton pusle height analysis spectra determined. An added incidental advantage of this arrangement would be that a number of the metallic elements concentrated in the cathode compartment would form amalgams in the mercury, and the mercury cathode could periodically be distilled and analyzed for the trace metals that have so accumulated. Here the X-ray fluorescence method or spark spectrograph could be used. (Fig. 5 follows:)

FIGURE 5