ANALYTICAL EDlTION
NOVEMBER 15, 1940
681
and the glossmeter has been found a valuable aid in the study TABLE V. GLOSSMETER RATINGSOF BAKEDUREA-FORMALDE-of alkali-resistance of baked urea-formaldehyde finishes, HYDE FINISHES I S .4LKALI TEST Typical data are listed in Table V. Vehicle, 40% urea formaldehyde, 60% As Table V indicates, the glossmeter recorded in numerical alkyd Alkali, 3% NaOH Pigmentation, 1:l on resin solids, 5% form the progressive loss of gloss produced by the attack of zinc oxide, 9570 Ti03 Panels, on glass 3 X 5 inches the alkali. Also, the improved resistance to alkali attack Film, 0.003 inch (Filmograph), 55% solids under longer baking conditions )vas readily measured. Since Gloss, Gloss, Resin Gloss, Gloss, Gloss, So. 1 Hour 2 Hours 4 Hours 5 Hours 6 Hours the alkali test undoubtedly is based upon the loss of gloss due Baked 30 minutes a t 200' F. to pitting of the surface, this instrument, designed to measure 1 Over SOa Over 90 32.5 l5b .. surface irregularities, is excellently adapted for application 2 Over 90 80 30 25b 3 Over 90 Over 90 40 30 3Ob in such a test. 4 5
1 2
3 4 5 0
b
Over YO Over 90
Over 90 50 Over 90 Over 90 Over 90
Over 90 60 SO Over 90 Baked 45 minutes a t 200' F.
Baked 90 minutes a t 200' F Over 90 80 l5b 30 40 Over 90 Over 90 72.5 Over 90 Over 90
40 47.5
50 b
40 72.5 90
32.5 32 5
45
.. 55 90
Over 90 indicates specular reflection a t normal incidence Film failure.
with carbon black as pigment, but zinc is equally effective with titanium dioxide. Lead and zinc are about equally effective in a toluidine red baking enamel. A third use of the instrument has been in following the loss of gloss of a baked urea-formaldehyde finish as observed during the alkali test, This application of the instrument, which approaches most closely the fundamental theory of the observation and measurement, was developed by K. E. Martin and W. H. Graeff in this laboratory. The test is in daily use
Summary
A glossmeter has been designed and constructed to measure the distinctness-of-image gloss of painted surfaces. While the theory of the instrument is based upon the existence of minute irregularities on, in, or beneath the surface of the paint film, the data collected thus far do not prove that distinctness-ofimage gloss is solely affected by the presence of such irregularities. The instrument is flexible in its application and has been found valuable in various problems dealing n-ith a gloss denoted as distinctness-of-image gloss. Its chief points of advantage are : simplicity of construction, simplicity of operation, and capability of yielding a reproducible gloss value in numerical form. Literature Cited (1) Hunter, R. S., K'atl. P a i n t , Varnish Lacquer bssoc., Sci. Sect., CZ'TC. 493, 268-80 (1935). (2) Zbid., 503, 141-53 (1936). (3) Hunter, R. S., a n d J u d d , D. B., Am. SOC. Testing Materials, BUZZ.97, 11-18 (1939). PRESENTED before t h e Division of Paint and Varnish Chemistry at the 9 9 t h Meeting of t h e American Chemical Society, Cincinnati, Ohio.
Precision Feed Device for Catalytic Experiments ROBERT L. BLRWELL, JR.', Trinity College, Hartford, Conn.
W
ITH increasing frequency, the necessity of employing
rare or expensive chemicals in catalytic experiments is arising. This article describes an apparatus designed for the precise feed of liquids on a semimicro scale, though i t may find application for ordinary feed rates where the precision of flow achieved would u-arrant the increased complication. The following characteristics were required for t h e author's purposes: ready setting of feed rate to any predetermined value, ready change of this rate, ready use of the apparatus at any pressure from atmospheric t o about 50 mm., provision for evacuating the catalyst under high vacuum, and absence of stopcocks in the feed line. Devices in which the feed liquid is displaced b y mercury flowing in through a capillary have been widely used in experiments with ordinary feed rates. These may be modified for use a t lower rates, as has been done, for example, by Goldwasser and Taylor (1). Such apparatus fails to meet all the above requirements. Furthermore, the achievement of uniform flow of mercury through fine capillaries is difficult if the mercury becomes in the slightest fouled. T h e device described in this article satisfies the conditions named above. T h e feed is positive in action and yields very 1
Present address, Northwestern University Evanston, I l l .
even and very reproducible result's. It could be simplified for purposes in which all the itemized characteristics were not required.
Description of Apparatus The apparatus is illustrated herewith. hIercury is displaced by the gas evolved by the electrolysis of a 30 per cent potassium hydroxide solution. The displaced mercury forces the feed liquid into the preheater. An electric current from a storage cell is passed through a 4-decade resistance box, &, and through the cell, K. The voltmeter, R, allows a check upon the potential fall through the cell. The cellis provided xith two nickel electrodes brazed or spot-welded to the tungsten rods sealed into the apparatus. It would probably be satisfactory, however, to replace the tungsten seals with leads passing through a rubber stopper. The gas evolved in K is fed to the container, H . Mercury is forced from H up into the calibrated 10-cc. buret, F , which is surrounded by a water jacket, E. The feed liquid passes into the capillary, C, and thence through the finer capillary, B, into the preheater, A , the jacket of which should be heated at least 40' above the boiling point of the liquid. The liquid volatilizes within the fine capillary at the edge of the preheater. In the author's experiments, the meniscus of the feed liquid would not vary 1 mm. from this point during an experiment. The proper diameter for the fine capillary depends upon the desired range of feed rates and upon the viscosity of the feed liquid. With a
682
INDUSTRIAL AND ENGINEERING CHEMISTRY
A
I
FIGURE 1. APPARATTX
VOL. 12, NO. 11
diameter of about 0.2 mm., feed velocities of 0.12 t o 4.2 cc. per hour were obtained using sec-propyl and sec-butyl alcohols. I t is, of course, necessary to thermostat vessels H and K . This may be done by immersing them in Dewar flasks with ice. The buret, F , may be estimated to 0.01 cc. Since the coefficient of cubic expansion of most organic liquids is about 0.001, the buret temperature must be maintained to within 0.1" to allow volumes of liquid fed into the preheater to be known to within 0.01 cc. The author was able to achieve this degree of constancy for periods of an hour by passing a steady stream of tap water through the buret jacket. The buret is filled by forcing the mercury to the top by the application of compressed air through the connection at J . The tip of D is broken off and inserted in a tube containing the feed liquid. Upon releasing the compressed air the mercury sinks, at a rate controllable by tap G, drawing liquid into the buret. The liquid in the filling tube may be kept boiling during this process to eliminate dissolved air. This is necessary if the feed is to be used a t reduced pressures, in which case gas bubbles would be evolved in the liquid. The tip a t D is then resealed. The small quantity of gas remaining a t D may be dissolved in the liquid higher in the tube by heating with a small flame a little above D. Small bubbles of gas will disengage and will rise and dissolve. In starting a run, the pressure in H must equal the pressure in the apparatus plus the hydrostatic pressure of the mercury column. This may be adjusted by reference t o a manometer connected through J . The catalyst may be evacuated, provided the buret is emptied through D. The last traces of liquid may be removed by passing a reverse stream of a gas which does not affect the catalyst out through D with the mercury almost a t the level of the side tube to D. Warming this section of the tubing hastens this process. The constancy of the feed rate is good. For example, delivering 0.3 cc. in 10 minutes, successive deliveries in each period were constant to +0.01 cc. until a total of 2 cc. had been delivered. However: since the gas volume in H increases as mercury is driven out, the rate of feed for a given current through cell K decreases with the height of mercury. This effect is the smaller, the larger the gas volumes in H and K . It may be compensated by an occasional decrease in the resistance.
Literature Cited (1) Goldwasser, S., and Taylor, H. (1939).
e., J . A m . Chem. Soc., 61, 1260
Aspirating Unit for Collecting Air Samples LESLIE SILVER>I.iN, Harvard School of Public Health, Boston, 3Tass., AND WESLEY B. W.IRDLOW, Texas S t a t e Health Department, Austin, Texas
I
N@IAKI?jG air analyses for industrial hygiene and other purposes, there is frequent need for a portable sampling unit which mill operate without an external power source or where the direct use of electric pumps or blowers constitutes a fire or expbsive hazard. Zhitkova (1) describes such a unit', but it is not available in this country and is not of adequate capacity for sintered absorbers. The unit described here is easily handled and can be used for laboratory, field, or routine plant determinations.
Construction The unit is shown in Figure 1. It consists of two standard 19-liter (5-gallon) gasoline containers connected by two 10-cm. (4-inch) sections of 0.625-cm. (0.25-inch) brass pipe and bushings and a 0.625-cm. (0.25-inch) gate valve. The bushings are soldered into holes drilled into the filling caps. Two 3.75-cm. (1.5-inch) sections of 3.125-em. (1.25-inch) thin-walled brass tubing are soldered into the can bottom for the air inlets and waterlevel indicator. Copper tubing, 0.938-em. (0.375-inch), lvith brass draincocks (automobile crankcase draincocks) is used for the air inlet and wat,er-level indicator. These connections are placed in No. 6 rubber stoppers which are easily removed for filling. Small draincocks are soldered to each of the pouring caps for air removal.
The tanks are mounted in a frame constructed of two 73-em. (&foot) pieces of 1.9-em. (0.75-inch) angle iron 0.47-em. (0.1875inch) thick. The carrying handle lugs are bolted to the angle iron after removing the handles, A light metal 0.3 X 2.5 em. (0.125 x 1 inch) strap is fastened around each container and a 10-cm. (4-inch) wooden baseboard is drilled for the filling caps. These baseboards are connected t o the angle irons by means of corner braces. Thirty-centimeter (12-inch) lengths of angle iron are fastened to the ends of the angle iron to act as a base when the unit is removed from the wheeled stand. Thus, the unit can be used with or without this mobile section. The iron pipes are slotted, as indicated, to allow ready removal of the aspiration unit. The turnover catch shown is made from 0.3 X 1.25 cm. (0.125 x 0.5 inch) steel strap bent into a U-shape and pivoted, as shown. An elastic band serves as a spring allowing foot operation for turnover releasing and catching.
Operation The rate of flow varies slightly x i t h falling head, as shown in Figure 2. These curves were obtained using a large wet meter of very low resistance to air flow and with the gate valve completely open. The rate a t any time is found by dividing ordinate by abscissa. For practical purposes when sampling a t low rates (0.5 to 1 liter per minute), it may be
