INSTRUMENTATION

---

INSTRUMENTATION by Ralph H. Müller

Epitasimeter, which measures surface tension continuously, is being used as process control instrument . . . . N e w ideas in photoelectric circuitry badly needed r r i o CONFIRM our long-standing con-

A viction that progress in analytical research is not confined to the "polish­ ing" and improvement of ancient tech­ niques we found a fascinating example in the research of a South African in­ vestigator. W. H . Aarts of the Re­ search Department, African Explosives and Chemical Industries, Ltd., Northrand, Transvaal, describes the Epitasimeter: an instrument for process con­ trol by means of the continuous meas­ urement of surface tension [J. Sri. Instr. 36,252 (1959)]. As the author points out, "Physical methods of determining chemical con­ centrations are very useful in the con­ trol of process plants. While many physical properties have been so used, little use seems to have been made of surface tension. Surface tension measurements are frequently used in the laboratory, and generally an abso­ lute value of the surface tension is re­ quired. For process control, however, absolute measurements are often not essential. The method described here uses a jet of liquid issuing from an elliptical orifice to form a standing wave. Viscosity effects are either eliminated or allowed for in the tem­ perature compensating circuits." As far back as 1909-1910 Niels Bohr found that an oscillating liquid jet forms a system of lenses and mirrors. He derived an explicit expression re­ lating surface tension to the wave length of the oscillation. When all the quantities in this fonnula are known or measured, it is possible to obtain an absolute value for the surface tension. Even the simplified expression is for­ midable, b u t as Aarts points out, ab­ solute determinations are not neces­ sary and the system can be empirically calibrated in terms of the parameter which is to be measured. Th Epitasimeter makes use of the fact tnat an oscillating jet consists of a series of lenses. T h e refracting sur­

faces of the lenses have two curvatures and the jet therefore forms a contin­ uous chain of symmetrical anastigmatic lenses. One of the double convex lenses is chosen in the Epitasimeter for locating the position of one wave. For the case of a system in which the liquid is maintained at a constant height above the orifice and for which the density of the medium (air) sur­ rounding the issuing liquid is very small compared with the liquid and for which the effective radius of the el­ liptical orifice is very small compared with the wave length, the γ/ρ value or ratio of surface tension to density can be empirically related to the only re­ maining variable—the wave length. Aarts uses an ingenious servo-system to locate the standing wave. A de­ tector head is mounted on a servo-con­ trolled carriage. Two right-angle prisms are placed a small distance apart, and, if light focused by the liquid lens element strikes the upper prism, a photomultiplier produces a signal moving the carriage upwards. If the light strikes the lower prism, another photomultiplier causes the reverse mo­ tion of the carriage. I n the very nar­ row intermediate position a third photomultiplier closes a relay which operates a lamp to indicate that the carriage is in the equilibrium position and is correctly measuring the position of the standing wave. The position of the image-locating carriage is measured in terms of the po­ sition of a drum potentiometer driven by the same shaft as that which raises "and lowers the carriage. This poten­ tiometer forms part of a Wheatstone bridge, one arm of which also includes a platinum resistance thermometer. Inasmuch as the ratio γ/ρ is a function of temperature, the distance, x, of the image locator below the orifice will be a function of the temperature for any given concentration. A temperature coefficient, a, of χ can be defined as:

XT = Χ * [1 - α (Γ-15)] where X 1 5 and Χτ are the distances at 15° and T°C, respectively. A suitable potentiometer can be constructed for setting in the appropriate correction for temperature. The bridge indicator is a 2000-ohm 25-0-25 microammeter and a series resistor in the bridge bat­ tery circuit enables one to set the meter at full scale for the highest concentra­ tion of solution to be measured. In the case of a liquid such as strong nitric acid giving off noxious fumes, self-cleaning windows are required. The author uses a simple but effective system in which a capillary admits some of the liquid to a chamber pro­ vided with overflow. Liquid from this chamber flows in a thin flat film over the protective window and effectively eliminates drop formation on the win­ dow. The instrument was originally de­ veloped to determine the concentration of very concentrated nitric acid con­ taining 0.1 to 2.0% nitrous acid and 1 to 7% water. Ordinarily such acids are analyzed titrimetrically with stand­ ard alkali to determine total acid con­ centration. The nitrous acid is deter­ mined by titration with standard po­ tassium permanganate and the nitric acid concentration calculated by differ­ ence. To avoid the tedium, several physical methods have been tried, but such schemes as density, electrical con­ ductivity, etc., give ambiguous results. With the Epitasimeter, the location of a particular standing wave is practi­ cally a linear function of nitric acid concentration. The mean error for chemical analyses is given as 0.05% and for the Epitasimeter 0.03%. The author is careful to note that the method measures the ratio of sur­ face tension to density and both are a function of concentration. In the case of brine solutions the two functions are so similar that γ/ρ remains practically VOL 31. NO. 10. OCTOBER 1959 ·

91

A

INSTRUMENTATION independent of concentration and the method is quite unsuitable. The water-ethanol system on the other hand is particularly favorable. At first thought it would seem that surface active agents woud be extremely effective but Aarts indicates that in this dynamic method the molecules do not have a sufficiently long period of time to migrate to the surface and hence affect the surface tension. In our opinion this investigation is a fine example of original thinking and the true research spirit. It emphasizes the fact that general instrumental techniques are more than adequate for the continuous and automatic measurement of difficult and obscure phenomena. Certainly, before the advent of this study, few investigators would have thought of using surface tension as a means of process control and in accomplishing this, the author invoked some elegant principles of hydrodynamics, optics, servomechanisms, electronics, and automatic temperature compensation. Photoelectric Circuitry Ideas Needed

With the continuing epidemic of papers on the optimum conditions for performing speetrophotometric measurements it seems to us that some new ideas in photoelectric circuitry are badly needed. It would seem that the analyst has been presented with a very limited set of techniques and it is assumed all too frequently that phototubes have severe limitations in reproducibility. We do not believe this and some day we hope to prove the point. It is well known that among all electrical measurements frequency can be measured with fantastic precision. There are a number of ways in which a phototube can control the frequency of an oscillator. In this connection some recently developed surface-barrier photodiodes have been developed which act as fast response photo-capacitors. If these are used in the tank circuit of a suitable oscillator, the frequency will be a function of illumination [Ahlstrom, E., Matthei, W. G., Gartner, W. W., Rev. Sa. Instr. 30, 592 (1959)]. The storage of photocurrents on precisionJ capacitors also presents great possibilities for increased precision. Schemes of this sort are used in some direct-reading spectrographs, but in such cases high photometric precision is of no great consequence because excitation conditions in arc or spark sources would hardly warrant high photometric precision. In connection with capacitor storage systems, improved Schmitt trigger circuits provide sharp criteria for the attainment of a precise charge level.

the EDITOR'S column L.T.Hallett, E d i t o r

Notwithstanding rapid progress in recent years in developing continuous onstream instrumentation, Robert T. Sheen, president of Milton Roy Co., Philadelphia 18, Pa., wrote to us recently, "the control laboratory still has had to analyze many variables. "Development of industrial colorimetric analyzers represents a substantial forward step in reducing laboratory analyses," he said. Today, it is often necessary to make accurate determinations of materials in solution in concentrations as low as a few parts per billion. At these low concentrations almost all materials follow Beer's law and colorimetric analysis can usually differentiate between single ions, often in the presence of other materials at several thousand times the concentration of the ion being analyzed. An analyzer-recorder combination has been developed for on-stream colorimetric measurement in which the color to be measured is the result of chemical reaction. Criteria include rugged construction, elimination of internal laboratory-type glassware; positive, accurate, volumetric measurement of both sample and reagents; and use of single wave-length light. The light, passed through the solution, is measured and related to the concentration of a specific material in the solution. The analyzer houses the sample and reagent metering pumps, the sample cells, light source, light filters, photocells, and the program timing system. A pneumatic programmer admits a 20 p.s.i. instrument air supply to a manifold. Small orifices open to the individual control lines. Each line terminates on the program drum at one end and at a pressure actuated component on the other. The slotted drum programs a complete cycle of analysis. As the cycle is started the bridge circuit is automatically zeroed after both sample cell and the comparison cell are filled. This compensates for irregularity in turbidity, coating of cell walls, dust on lenses, etc. The reagent metering pump consists of two identical segmented spherical cavities clamped together and containing a slack diaphragm. Reagent under pressure flows to the metering pump moving the slack diaphragm in the pump completely displacing exactly two

Schematic diagram of a two reagent system with flow-throagh type cells •sed f o r hardness analyses

ml. of the other reagent. When the inlet and outlet valves change position, this latter reagent is similarly metered. A sufficient time interval is allowed in the programming for proper color development. The recorder then indicates and records the results of the analysis. A complete cycle for hardness in water is 6 minutes. With this timing, 480 ml. of each reagent is required per 24 hours. The five-gallon supply in polyethylene containers is sufficient for over a month's operation. Provided the original instrument is supplied with sufficient number of metering reagent pumps, the same instrument can be used for a large number of different analyses by a field change of the interchangeable program timers and the insertion of the proper monochromatic light filters. Procedures are available for free chlorine, chlorides, chromium, copper, cyanide, fluoride, formaldehyde, total hardness (Ca + Mg), hydrazine, iron, dissolved oxygen, phosphates (ortho, P0 4 ), sulfites, and dissolved silica.