I
Ii 1
,'I%+
1 1
Il
1
,,Microburette
delivery tube
n Ti t r oil on
Apparatus side view along the pivot axis with the solder k n o b striking the vessel to cause it to swing away. B. Transverse section through the apparatus across the pivot axis with vessel in vertical position during the swing Figure 1 . A.
densation reactions, in which conventional methods of stirring, to prevent clumping of nondissolved particles, have proved unsatisfactory. The requirements of this titration were similar, and small titration vessels were made which employed the same principle. The vessel (Figure 1) was straight sided with the base as a rounded cone protruding into its lumen. Apparatus Design and Procedure. The vessels were made from 12-mm i.d., medium wall thickness, borosilicate glass tubing. This tubing was blown to give a test tube round end and this was flattened in the flame. When the flat base was yellow to white hot, the tube was placed firmly and vertically base down onto a piece of asbestos board. The air in the asbestos directly under the hot base expanded and the increased pressure forced up the molten base as a smooth cone into the vessel lumen. The tube was then cut to give a vessel of external height, 2.0 cm. Once a known volume of titrate is delivered into the titration vessel, evaporation of water is irrelevant as long as sufficient volume of liquid remains to allow adequate dispersion of the precipitate. Volumes delivered from a breaking pipet in this study were found by back titration to be 7.95 g1 (*170). For practical convenience, this volume was increased by the addition of 0.3 ml of A.R. ethyl alcohol. This addition may affect the titration when titrating for very low concentrations of chloride with silver ni(5) K Schwarz, and C. Schlosser, Mkrochernie
13 ( N F 7 ) , 18 (1933)
trate (5), but here the range of concentrations was too high for any such effect to be important. Independent control titrations showed that the addition of small volumes of ethanol did not affect the end point. With vessels this small, the lateral motion required to effect the rotary swirling could be produced by swinging the vessel through a limited arc a t a controlled frequency. A rubber collar was cut from bunsen tubing and two pins were pushed through the collar from the inside, along a diameter. These were the vessel supports and pivots when the collar was fitted to the vessel. This type of collar allowed precise adjustment of the pivot height above the vessel's center of gravity. The pins pivotted in grooves of a fixed metal support ring within which the titration vessel swung. Regular swinging of the titration vessel is effected by its being struck by a soldered knob on the turning shaft of an electric motor. The turning speed of the motor is controlled through a rheostat as the striking frequency is critical. A small wave propagated in the titrate passes around the conical base rather than across the vessel and is reinforced each time the vessel is struck. The resultant stirring efficiently mixes the liquid and prevents precipitateclumping about the microburet and electrode tips. A rubber tubing collar fitted over the microburet delivery stem within the vessel, acts as an internal buffer, limiting the swinging travel. The collar also supports the silver anode wire which is pushed down through it to a position closely adjacent to the microburet tip. The collar is turned to align the anode and the delivery stem perpendicular to the pivot axis. The microburet tip and the anode are then positioned close to the side of the vessel in the greatest depth of liquid and adjacent to either of the pivots. With this precise positioning and buffering, neither the microburet nor the anode are struck by the swinging vessel. The electric motor is clamped so that it can be swung to or from the vessel. At the end of a titration, by lowering the pivot support ring away from the microburet, the vessel is easily removed for cleaning and refilling and can be readily replaced. Once the apparatus has been precisely assembled, a whole series of titrations may be performed rapidly without further adjustment. The electrodes are connected directly to the input of a potentiometric recorder (set to 500 mV full scale deflection) and the whole process of titration is highly automated. Received for review January 29, 1973. Accepted August 2, 1973.
Fluorescent Integrating Radiation Dosimeter W. A. Salmon and E. A. Chandross
Bell Laboratories, Murray Hill, N.J. 07974
The field of high-energy radiation processing of materials on an industrial scale is slowly but continuously expanding. A partial list of processes in operation includes the manufacture of heat-shrinkable film and tubing, the cross-linking of wire insulation, and the sterilization of medical supplies. A need exists for a simple indicator which would show a t a glance whether or not a supposedly 2446
processed object has indeed been exposed to the intended radiationdose. Various dye dosimeters are known which undergo a visible color change upon irradiation, i. e., the blue cellophane of Henley and Richman ( I ) which bleaches on exposure to (1)
E. J.
ANALYTICAL C H E M I S T R Y , VOL. 45, NO. 1 4 , DECEMBER 1973
Henley and D. Richman, Anal. Chem., 28,1580 (1956).
radiation. In other dosimeters, a color is formed or intensified (2). Dyes which bleach are unsatisfactory for the present application because the fading of a color is an ambiguous indicator. Those which form a color are, in general, limited to use with light-colored backgrounds in order to be distinguishable. Dosimeters based on radiation-induced fluorescence of inorganic phosphate glasses have been extensively studied ( 3 ) . We have developed a new dosimeter based on organic compounds in polymer films which, upon exposure to electron beam radiation, become visibly fluorescent under ordinary "black light" excitation. The material can be applied in the form of a paint. As a result, exposure can be qualitatively detected with high sensitivity by visual examination, In addition, an optical absorption is developed which can be correlated quantitatively with the absorbed dose. We began our investigation with the assumption that any system which underwent the desired change on exposure to ultraviolet light would behave similarly in an electron beam. Very recently, Zweig has reported several organic materials which became fluorescent after exposure to short-wavelength UV light ( 4 ) . This is a very simple assumption, and it appears to hold in general. There are many photochromic substances known, but they are not suitable for our purposes for a variety of reasons. First, we require a chemical system which is insensitive to ordinary daylight. Further, the changes caused by exposure should not be erased on standing. Finally, we would like to be able to use induced fluorescence as a means of detecting the change because fluorescence can be measured with great sensitivity. Ordinary photochromic materials do not meet these criteria and thus are not suitable for radiation dosimetry. The chemical system we chose is the photochemical cleavage of the photodimer of a suitable aromatic hydrocarbon to two molecules of its precursor. The photodimers formed by the anthracene family are a straightforward example of this ( 5 ) . The photodimers are formed by irradiating solutions of the monomers (I) with relatively longwavelength ultraviolet light. Anthracene dimers (11), which contain only isolated benzene rings as chromophores, do not absorb at wavelengths above 310 nm; thus they are not affected by ordinary daylight. The thermal cleavage of these photodimers is usually very slow at temperatures below 100 "C and is insignificant a t room temperature for most of them. R
.
X >310nm
L
heat, X < 310nm
I
I1
Once the dimer, in a plastic matrix such as a PMMA film, is broken to monomer, the characteristic near-UV absorption and the visible fluorescence of the latter are (2) A. Charlesby, "Atomic Radiation and Polymers," Pergarnon Press, New York, N.Y., 1960, pp 101-7. (3) J. H. Schulman, "Survey of Luminescence Dosimetry," in 'Luminescence Dosimetry," F. H. Attix, Ed., U. S. Atomic Energy Commission, Division of Technical Information, Oak Ridge, Tenn., 1967, pp 3-33. ( 4 ) A. Zweig, Pure Appl. Chem., 33,389 (1973). (5) W. J. Tomlinson, E. A. Chandross, R. L. Fork, C. A. Pryde. End A. A. Lamola, Appl. Opt., 11, 533 (1972) and references therein.
0.5
DOSE: 10 Mrad
0.4
0.3 w
z
4
a: v)
U
0.2
0.
C WAVELENGTH
Figure 1. Absorption spectrum of
inmi
n-amyl anthroate monomer
readily apparent. If the matrix is sufficiently soft, redimerization on exposure to "black light" is very inefficient and is unimportant for most purposes ( 3 ) .A variety of dimers could in principle be used, but it seems that anthracene derivatives represent the most attractive choice in terms of availability and the criteria given above. The solubility of the photodimer in the solution used for casting the polymer film is important. Most dimers are not very soluble but can be made more soluble by the incorporation of suitable substituents which do not alter the basic photochemistry. For our studies, we used the photodimer of amyl anthroate (11, R = C0&5Hll), which is quite soluble in the chloroform-toluene solvent used for the preparation of the PMMA films.
EXPERIMENTAL Anthracene-9-carboxylic acid (Aldrich Chemical Co.) was converted t o the acid chloride by treatment with excess thionyl chloride in boiling benzene. After evaporation of the excess reagent and the solvent under reduced pressure, the crude acid chloride was used directly for the next step. The anthracene-9-carbonyl chloride was dissolved in excess namyl alcohol, and the solution was boiled for 5 minutes. After evaporation of the excess alcohol, the residue was crystallized from benzene-hexane t o give yellow prisms, melting point 39-41 "C. of amyl anthroate. The ester was dissolved in toluene (1 g/5 ml) and, after deaeration with nitrogen, the solution was irraidated with a mediumpressure mercury lamp (borosilicate glass flask) until precipitation of the product appeared to cease. Crystallization of the dimer from toluene gave colorless crystals. The photodimer-polymer mixture was spin-coated onto 2.2- X 2.2-cm microscope cover glasses by means of a photo-resist spinner (Headway Research, Inc., Garland, Texas) from a solution having the following composition: n-amyl anthroate, 1 gram; poly(methy1 methacrylate), 5 grams; toluene, 25 ml; and chloroform, 25 ml. The film thickness, determined by weighing the coated cover glass, was 4.6 pm. The thickness was very uniform from specimen to specimen. This was ascertained by spinning a more dilute coating onto a series of cover glasses, breaking the dimer with shortwave UV light. and measuring the absorbance of the monomer a t 386 nm (A386). There was no significant variation from one film to another.
ANALYTICAL C H E M I S T R Y , VOL. 45, NO. 14, DECEMBER 1973
2447
V."
DOSE RATE: 0 0.045
-
-
0.4
o.8
s a 4
0.3
-
0.2
-
0.1
-
t
b
Mradlsec
1.3 Mrad/sec
w
z
g in 4
I 5
I 10 DOSE (Mrod)
I 15
I
3
20 DOSE
(Mad)
Figure 2. Formation of n-amyl anthroate monomer in the dose range of 0-20 Mrad
Figure 3. Formation and destruction of n-amyl anthroate mono-
The films were irradiated with 1.0-MeV electrons from a Van de Graaff accelerator. Two dose rates were investigated, 0.045 and 1.3 Mrad per second. The formation of monomer was followed by measuring A 3 8 6 . The increase in the base line, due entirely to the development of color centers in the glass during irradiation, was measured after the film was removed with chloroform. The baseline absorbance was subtracted from the measured value of A386 in the results described below. A typical spectrum is shown in Figure 1.
nounced. The curve reached a maximum a t 40 Mrad. Further irradiation resulted in destruction of monomer until, a t 97 Mrad, it was all destroyed. These results are shown in Figure 3, which is a plot of the data over the entire dose range studied. Commercial electron radiation processes for polymeric materials utilize doses typically in the range 1-30 Mrad. The photodimer system described above, therefore, is a useful dosimeter in this dose range. Its great advantage, of course, lies in the fact that it forms a fluorescent product and makes possible the qualitative detection or confirmation of irradiation, regardless of background color, The G-value for dimer breaks was found to be 29, which is unusually high for a reaction that cannot involve radical chains. It should be possible to correlate the fluorescence intensity of the broken dimer quantitatively with the absorbed dose. The conditions required for this are more involved than those of the absorbance readout. In order for fluorescence intensity to be linearly proportional to concentration, one must work at low concentrations to avoid the problem of reabsorption. I t would be advisable to use a somewhat thinner film, and a lower concentration of photodimer if fluorescence intensity is to be used for quantitative dosimetry.
RESULTS Figure 2 shows the results in the dose range 0-20 Mrad. The initial portion of the curve up to 5 Mrad is linear and fits the equation:
where fil = 0.0324 Mrad-1 and D = dose (Mrads). Making the reasonable assumption that the absorbance follows the Beer-Lambert Law, Equation 2 below describes the general case for the region of concentration and dose rate in our experiments: where c = concentration of dimer in film (g/ml); t = film thickness (cm); and k 2 = 600 cm2 g-l Mrad-I. Increasing the dose rate from 0.045 to 1.3 Mrad sec-l had no effect on the rate of breaking the dimer. As the dosage was increased, the departure of the absorbance us. dose curve from linearity became more pro-
2440
mer in the dose range 0-100 Mrad
Received for review June 21, 1973. Accepted September 10, 1973.
ANALYTICAL CHEMISTRY, VOL. 45, N O . 14, DECEMBER 1973
