emission spectra of the discharge. The background spectrum is thus low, especially from within the coolant tube. In addition, the discharge offers a wide range of excitation energy for special applications. Greenfield, Jones, and Berry (5) have confirmed our expectations that matrix effects are negligible. Although only a convenience advantage, the discharge emits considerably less audible noise than a Beckman burner or a constricted d.c. arc plasma jet. Our observations indicate that this combination of plasma and aerosol generator is a practical and versatile source for analytical spectrometry. We are now exploring further analytical applications and will discuss these in a more complete communication.
LITERATURE CITED
(1) Beguin, C. P., Kana’an, A. S., Margrave, J. L., Endeavor 23 (89), 55
(1964).
( 2 ) Cannon, H. R., “A Study of an In-
duction-Coupled Plasma Operating at 400 Kilocycles,” M.S. thesis, Air Force Institute of Technology, Wright-Patterson AFB, Ohio, GA/Phys/62-2, AD-286404 (1962). (3) Greenfield, S., Jones, I. LI., Berry, C. T., Analyst 89, 713 (1964). (4) Kana’an, A. S., “Studies at High Temperature,” Ph.D. thesis, University of Wisconsin, Madison, Wis., 1963. ( 5 ) Lang, R. J., J . Acous. SOC.Am. 34, 6 (1962). (6) hlavrodineanu, R., Hughes, R. C., Spectrochem. Acta 19, 1309 (1963). (71, Mavrodineanu, R., Boiteux, H., Flame Spectroscopy,” pp. 51-3, Wiley, New York, 1965. ( 8 ) hlironer, A,, Hushfar, H., “Radio
Frequency Heating of a Dense, AIoving Plasma,” presented at AIAA Electric Propulsion Conference, Colorado Springs, Cola., (March 1963): preprint 63045-63, American Institute of Aeronautics and Astronautics, S e n York, -N . v _.
(9) Reed, T. B., Intern. Sci. Technol. (6), 42 (1962). (10) Reed, T. B., J . Appl. Phys. 3 2 , 821 (1961). (11) West, D. C., Hume, D. Tu’., A Y ~ L . CHEM.36,412 (1964). RICH.ZHD H. WEXDT YELMEHA. FXSEL Institute for Atomic Research and Department of Chemistry Iowa State University Ames, Iowa RECEIVED for review February 24, 1965. Accepted April 1, 1965. Work was performed in the Ames Laboratory of the U. S. Atomic Energy Commission.
Application of Electrolytic Moisture Meter to Measurement of Water Vapor Transmission through Plastic Films SIR: The water vapor transmission (WVT) of plastic films is commonly determined by a method, such as ASTM E96-53T1which measures weight of water passing through a film sample under specified conditions. This procedure cannot be applied to thick films or to materials with a very low water vapor transmission rate, because the weight of water passed in a reasonable time is too small to measure accurately. It appeared that very sensitive measurement of WVT would be possible with the Pz05electrolysis cells developed by Keidel ( g ) , and an instrument was designed and built for this purpose. This report describes the instrument and its application to the measurement of water vapor transmission rates.
MEASRING
CELL
Figure 1. saturator
Block diagram of WVT instrument and detail of
EXPERIMENTAL
PaOr Electrolytic Cells. The electrolytic cells used in this instrument were purchased from the Consolidated Electrodynamics Co., Pasadena, Calif. h thorough description of their construction and operation is contained in papers by Keidel and others (1, 2 ) . These cells quantitatively electrolyze the water vapor contained in a gas stream passing through them, and the measured electrolysis current is converted to weight of water per unit time by application of Faraday’s Law. Sensitivity of measurement is such that water diffusion rates of the order of micrograms per hour can be measured when the cell is connected to the water vapor transmission cell described below. 922
ANALYTICAL CHEMISTRY
Water Vapor Transmission Cell. The sample cell consists of two compartments separated by the film whose WVT is being measured. The lower compartment is maintained a t a known, constant water vapor content (usually 100% relative humidity). The upper compartment is continuously swept with a stream of dry nitrogen which maintains it a t essentially zero relative humidity and carries the water diffusing through the film into the P20acell for measurement. (Although the instrument reading is reasonably independent of nitrogen flow, the rate is maintained a t 90-100 cc./minute to ensure rapid removal of
water vapor from the upper compartment.) A series of concentric lands and grooves machined into the faces of the two halves of the cell to give a crosssection
forms a leak-free seal when a plastic film is clamped between the faces of the cell.
The cell constructed for measuring bhe WVT of Kel-F films has a sample area of ahout I 0 0 square cm. A smaller cell would be easier to seal, hut the amount of moisture diffusing through the Kel-F would he too small to measure accurately. (100 square em. of ,>mil Kel-F transmits onlv about grain of H,O per hour.) If the pressure difference between the two compartments of thc cell is too great, t h i film being tented may he distorted and pressed against the interior of the cell, thus rrdueing the area available for water vapor transmission and giving erroneous results. This occurred when the nerdle valves and flowmeters were installed at the nitrogen inlet, but has not hcrn a problem with the valves downstream and both compartments of the WVT cell as close as possihle to the common nitrogen inlet. (In recent work with rubher-like films a differential manometer and another needle valve havc been added to allow the pressure difference between the compartments to be measured and minimized.) Saturator. The WVT cell was originally operated with a few drops of liquid water in the lower compartment,. This arrangement did not permit the measurement of blanks (with a dry lower compartment) and also gave erroneous results if liquid water touched the sample during the measurement. Consequently, an external gas saturator with a by-pass was constructed so t h a t the flowing gas stream in the lower compartment could either be dry or be saturated with water vapor. Operation of Instrument. The various components of the instrument are connected as shown in the block diagram, Figure 1. Two P,O, cells are used, the first serving to d r y the nitrogen stream hefore it enters the WVT cell and the second to measure t h e moist,ure transmitted through the sample film. The gas saturator 1s constructed of glass. The assembled instrument, minus the gas saturator, is shown in the photograph, Figure 2. A measurement of WVT is made by clamping the sample in the WVT cell and passing wet Nf through the lower compartment. A very high meter reading indicates a sample too permeable to measure, or a pinhole, or leak in the sample. If the sample is measurable, the wet gm is continued until the meter or recorder shows a steady state condition in the upper compartment. This then corresponds to the moisture transmitted through the film plus the instrument background (micro-leaks, etc.). The saturator is then by-passed and dry gas passed through the lower compartment until a steady state-the instrument background-is reached. If this is a significant fraction of the brevious reading (which it is with Kel-F), the Iirevious reading is correctid accordingly. The time required to approach a steady state reading varies with the permeability of the samlilc. A 1-mil polyester film will take 10 to 20 minutes, while &mil Kel-F will require ~
~~~
~
Figure 2.
WVT apparatus (without saturator)
several hours. This is also true when going from the wet to the dry gas state to measure the instrument hackground.
Table
I. Results of Water Vapor Transmission Measurements
WVT (grams meter2 24 hours-l)_
RESULTS
Calculation of Results. The steady state current reading is directly convertible to weight of water per unit time by use of Faraday’s Law. Therefore, the current reading and the area of the sample film are all t h a t is required to calculate WVT in the units of ASTM E96-53T: WVT= (grams H20)/(meter’) X
ThiekMaterial
ness, Electrm mils lytic
ASTM ESfi-
53T
Unrdel -~ -. .
polystyrene 1 . 4 >26 Lexan polycarbonate 3 . 2 14 Celanese polyester 1 Kodar polyester 1 Mylar polyester 2 3M polyester 2
8.2
7.5 4.8 4.2
44.2
17.6 7.9 7.4 4.0 3.4
(24 hours) =80.6 (ma.)/cm.2
The WVT of a given material is usually inversely proportional to the thickness of the sample film, and by mult,iplying the WVT by the thickness, a constant which is independent of film thickness is usually obtained. Results of Measurements. Results of a number of \ V V T measurements made with the electrolytic moisture meter are summarized in Tahle I and compared with TTVT values measured by routine determinations using ASTM E96-53T. The results agree satisfactorily with those obtained by ASTM E96-53T a t the higher TVVT levels, hut are considerably lower than the ASTM results for the low WVT samples. Other Applications. The instrument as it now exists could easily be adapted to study the variation of W V T with temperature or ait,h water vapor pressure. I3y placing a hydrocarbon or other material in the saturator and using a suitahle detector, such as a flame ionization
Reported poor reliability Actual rurrent measured waa 0.035 ma., corresponding to about 2 p.p.rn. water st 100 er./rninute gaa flow. Instrument background WRS 0.1lni ma.
detector, the transmission of materials other than water could be measured. LITERATURE CITED
( I ) Cole, L. G., Czuhn, Michael, Masley, R. W., Sawyer, I). T.,A N ~ L CHEM. . 31,2048 ( I W I ) . ( 2 ) Keidel, F. A,, I b X , p. 2043. PAULE. TOREN
Central Research Lahorntories Minnesota hlining iu .\lnnufnrturing Co. Saint I ’ d , .\linn. 55119 VOL. 37, NO. 7, JUNE 1965
923
