INSTRUMENTATION BY
R A L P H H.
MÜLLER
Advances in Photometric Design HE
INCREASED
ATTENTION
being
T given to filter photometers is timely
and encouraging. F o r a long time the analyst contemplating a long series of simple colorimetric analyses has been pitied or scorned if he failed to use a n elaborate a n d expensive spectrophotometer for measuring absorbance. T h e respective advantages and limitations of spectrophotometers and filter photometers have been recognized for m a n y years. One would assume t h a t the important factors of cost, speed, and convenience would always be k e p t in mind even in our affluent society. With all their flexibility and information gathering capabilities, only a few spectrophotometers are capable of high photometric precision. On the doubtful assumption t h a t a filter photometer is necessarily crude, and in m a n y cases an unavoidable compromise, too little attention has been given to the very high precision which it can afford. T o be sure, purely chemical studies have provided us with colorimetric reagents of relatively high selectivity and which produce light absorbing entities with high extinction coefficients. This has been good news for the biochemist, clinical chemist, and the trace analyst, and p u t s the subject in the microchemical category. There are many, however, who don't have to worry a b o u t sample size b u t must have a decent regard for sampling errors and high p r e cision. Several factors have contributed to the relatively small effort to improve filter photometers. One of these has been the practice of presenting t h e entire transmittance (or absorbance) on a three- or four-inch meter scale. P r e sentation on a 10-inch range recorder chart improves readability b y a factor of only 2.5. More t h a n t h i r t y years ago a t New York University, we were using vacuum tube voltmeters to read out phototube currents, the measurements being made b y precise voltage compensation in the input circuit. I n this case, the o u t p u t current was read merely as a balance indication and with a potentiometer of ICh4 volt sensitivity used for compensation, precision of greater t h a n 0 . 1 % was easily attained. Circle No. 43 on Readers' Service Card
As early as 1935, photometers with direct logarithmic o u t p u t were described [Miiller and Kinney, J. Opt. Soc. Am. 25, 342 (1935)], one of which used a variable μ tube with appropriate cathode loading such t h a t the o u t p u t signal is directly proportional to the logarithm of the input voltage. When one considers the enormous progress in electronics since those days it m a y be said t h a t perhaps no more t h a n ten percent of the present re sources have been applied t o this p r o b lem. One m a y ask—what about volt age-to-frequency conversion and subse quent counting? Or integration of radiant flux for a precisely measured time interval? Perhaps we have not recovered from the period when simple barrier-layer cell photometers were developed and manufactured in profusion. These were barely as good as a visual pho tometer; they merely eliminated fa tigue and the necessity for some degree of capability of visually matching in tensities. Entirely too much emphasis was placed on the fact t h a t barrierlayer or self-generating cells require no external source of electrical energy. This dubious advantage must have im plied t h a t the source of illumination in the photometer was either a candle, sunlight, or a kerosene l a m p . W i t h all their idiosyncrasies as compared with a high vacuum phototube, one cannot ig nore the great advances which have been made in selenium and silicon solar cells. They are now the principal sources of power in space vehicles and it would be surprising if some of their initial defects had not been eliminated. Optical factors involved in photo electric photometry have not been ex ploited to the degree warranted by pres ent day technology. Too little a t t e n tion has been given to the errors arising from differences in refractive index when comparing a colored solution with the blank. Such errors can be com puted from Fresnel's Law. I n this and other respects, it was shown m a n y years ago b y Kortiim t h a t much greater precision can be attained b y comparing a n unknown solution with a standard solution of nearly the same concentra-
tion. An alternative approach which retains all these advantages was p r o posed as a theorem [Miiller, R. N . I N D . E N G . C H E M . A N A L . E D . 11, 1, (1939)]
in which a solution is measured in a rectangular cell, first through one thick ness and then b y 90° rotation, or al ternative optical methods, through the second thickness. The advantages accruing from the use of monochromatic lamp sources have been known for a long time. T h e work of Ashley was particularly con vincing in this respect. The great amount of work which has been done in developing hot cathode lamps for atomic absorption spectrophotometry would seem to be incentive enough to develop discharge tubes, perhaps with a mixture of rare gases, with a reason able number of strong lines throughout the spectrum. These could be isolated with interference filters or a relatively crude prism or grating monochromator. Historically there is ample evidence in all branches of instrumentation t h a t a technique can often be greatly simpli fied and applied to other techniques which afford increased resolution b y virtue of the selective n a t u r e of the second technique. F o r example, chro m a t o g r a p h y a n d ion exchange are in themselves highly selective. If efflu ents are measured with ultraviolet light which is absorbed by all fractions to some appreciable degree, it m a t t e r s lit tle t h a t the wavelength which is used is not necessarily the best choice for each eluted fraction. A recent development which we find most interesting and provocative is the D u P o n t 410 precision photometer; E . I . du Pont de Nemours & Co., Inc., I n struments Products Division, Wilming ton, Delaware, 19898. Through the courtesy of D r s . R . A. Baxter and D . R. Johnson we have obtained some p r e liminary details of this development. The 410 precision photometer is a n outgrowth of the D u P o n t 400 process stream analyzer. T h e development is described by Puai B . Hamilton [Rev. Sci Instr. 38, 1301, (1967)] as a "Sensi tive Linear Flowing Stream Photometer for Amino Acid Analyzers." I t p r o vides a system with linear output suitVOL. 40, NO. 6, MAY
1968
•
109
A
INSTRUMENTATION
BEA» SPLITTER PHOTOTUBE
SAMPLE CELL
LENS
LENS
LENS
• EMURIN6 λ FILTER
CONTROL RECORDER STATION
PHOTOTUBE MIRROR REFERENCE λ FILTER SOURCE SECTION
SAMPLE SECTION
DETECTOR
SECTION
OPTICAL UNIT
SOURCE
SEAM SPLITTER
' |
I I I Γ»
,
LENS
,
c u e
SAMPLE CONTROL |
RECORDER
PHOTOTUBES StATIO»
I LENS • IRfiOR
SOURCE SECTION
τ
, _ _
SAMPLE SECTION
DETECTOR SECTION
OPTICAL UNIT
Figure 1 . Schematics of DuPont 4 0 0 precision photometer. beam arrangement. Lower: Dual beam arrangement
Upper:
Split
able for monitoring the effluent from ion-exchange columns after reaction with ninhydrin in the chromatographic analyses of amino acids. Phototube outputs are converted to voltages in log arithmic amplifiers, further amplifica tion is provided, and the final voltages are subtracted yielding log I0—log I values [or log ( I 0 / I ) ] which are propor tional to concentration. In one ar rangement, the absorbance can be mea sured at 4400 and 5700 Â. A split beam arrangement allows two levels of sensitivity to be recorded independently and simultaneously, each to any degree of attenuation below maximum so that a wide range of concentration can be accommodated. The range is 0-1.5 absorbance; at 2.5 absorbance the departure from linearity is 2.2%. Drift and noise for the absorbance range 0 to 0.01 is approximately 2% of full scale, equivalent to 0.0002 absorbance. This paper should be consulted for interesting details such as design of small flow cells, calibration, and other applications such as bacterial counts and rate of bacterial growth and reaction kinetics. Another paper contributing to further developments of the 410 photometer is that of J. J. Kirkland, ANAL. C H E M , 40, 391, (1968) which is en-
Control yοur temper with a Lauda Circulator Need t o temper glassware or other laboratory equipment? Let a lauda Circulator d o it. A variety of different models (K-2 shown here) circulate water, oil or other liquids at constant, electronically controlled temperatures up to 300° C , or as low as —120° C. Each features stainless-steel construction of all com ponents in contact with bath liquid, manual drainage and flow valves, and automatic excess load protection for pump. Our new 36-page catalog gives you all the details. Just write: Lauda Circulators, Division of Brinkmann Instruments, Cantiague Road, Westbury, N.Y.I 1590
A DIVISION OF
B Lauda Circulators
BRINKMANN
Circle No. 105 on Readers' Service Card
110
A
•
ANALYTICAL CHEMISTRY
titled "A High Performance Ultraviolet Photometric Detector for Use with Efficient Liquid Chromatographic Columns." This paper in addition to instrumental details gives examples of determinations in which nanogram amounts of substances are easily detectable. A close-up of the control station module and optical module of this photometer is shown on the top of page 124 A of this issue. More informative are the schematics shown in Figure 1. In the upper schematic a beam splitter allows the measurement of the difference in absorbance at two wavelengths, thus providing compensation for interfering materials or particulates. The dual beam arrangement (lower schematic) measures absorbance at a single wavelength through two beams (reference and measuring) to give precise differential absorbance. With appropriate filters, any wavelength between 380 and 650 m^ can be selected, using a quartz-iodine light source. The 253.7 ταμ, mercury line can be obtained from a low pressure Hg lamp. Other regions are promised from 1000-210 πιμ at a later date. The unique logarithmic amplifiers provide outputs which are linear in concen tration over a 10,000:1 range. Out put is standard at 0-10 mV. Two outputs are provided in 10:1 sensitivity ratio.
