Method of Evaluating Metal Cleaners - American Chemical Society

THE importance of having an adequate method for evaluating alkaline cleaners for the removal of contaminants from metal surfaces is generally agreed u...
0 downloads 0 Views 5MB Size
Table

V.

Tocopherol Content of Vegetable Tissues

Sample

Furter-3Ieyer

Mu. 0.23 0.44 0.39 0.36 0.72 1.24 0.42 0.34

Spinach leaf meal Turnip leaf meal Xlash, 5 % alfalfa l I a s h , 2 . 5 7 , brorroli Scratch corn FrPsh carrot root Fresh carrot t o p Frozen spinach a

20I

ANALYTICAL EDITION

March, 1946

...

0 . 14“ 0 . 79“ 0.47“

Emmerie-Engel 7 1 ~ rgram

0 26 0.71 0.42 0.39 0 70 1 20 0.40 0 30 0 02 0 02 0 00 0.13“ 0 56“ 0 43n

Calculated on moisture-fire hasis,

than the Emmene-Engel value. Hickman et al. (6) report an analgkis of a mixture of foods i n n-hich the Furter-Meyer value was 1400% higher than the Emmerie-Engel result. These extreme discrepancies are undoubtedly due to insufficient purifieation of the extracts prior to color measurement. The Emmerie-Engel procedure is much more sensitive than the Furter-Meyer, and hence is more convenient for low potency samples. SUMMARY

The Furter-lfeyer and Emmerie-Engel methods for determination of tocopherol have been applied to plant extracts. I n both cases, dry plant materials are extracted with Skellysolve B, and fresh materials with an ethanol-petroleum ether solution, the alcohol then being removed. I t is necessary to purify the sample before the final estimation can be made. Chlorophyll and xanthophyll are separated from tocofiherols by adsorption on a Superciil-activated magnesia column. Carotene and tocopherolquinonw are then destroyed by treatment with 85% sulfuric acid. In the Furter-Meyer procedure the tocopherols are finally transferred to ethanol solution, oxidized with nitric acid, and, after removal of ethanol-insoluble lipoids, determined with a photoelectric colorinieter at 480 nip.

I n the Emmerie-Engel procedure the tocopherols are finally dissolved in chloroform, and reacted with the ferric chloride-a,a’dipyridyl reagent. The pink ferrous dipyridyl complex is immediately transferred to aqueous solution to prevent further reduction of the iron by a slowly reducing fat-soluble compound. The colorimetric determination is carried out at 520 mp in a and 480 and 520 mp values obphotoelectric colorimeter. tained by the two methods with synthetic and natural a-tocopherols agreed within 5%. The values for natural y-tocopherol were 10% lower by the Furter-Meyer method and 35% lower by the Emmerie-Engel method. All these compounds can be quantitatively eluted from a magnesia-Supercel adsorbent and show a maximum loss of 57, when shaken with 85% sulfuric acid. Tests in which pure synthetic a-tocopherol was added to the extracts and carried through the entire procedure showed an average recovery of 95% with both methods. I n most cases the rmults obtained by the two procedures agreed within * 5 to loc;. ACKNOWLEDGMENT

The authors n-ish t o acknowledge the assistance of Margaret E:. Heller in this invehtigation. LITERATURE CITED (1)

Baxter, J. G., Robeson, C. D., Taylor, J. J., and Lehman, It. It-.,

J . Am. Chem. Soc., 65,918 (1943). (2) Devlin, H. E., and Mattill, H. A., J . Biol. Chem., 146, 123 (1942). (3) Emmerie, A., and Engel, Chr., Rec. trap. chim. Pays-Bas, 57, 1351 (1938). (4) Fisher, G. S., IND. ENG.CHEM.,ANAL.ED.,17, 224 (1946). (5) Furter, M., and.Meyer, R.E., Helv. C h i m . Acta, 22,240 (1939). (6) Hickman, K. C. D., Kaley, M. TV., and Harris, P. L., J . B i d . Chem., 152, 321 (1944). (7) Hove, E. L., and Hove, Z., Ibid., 156, 601 (1944). (8) Karrer, P., and Keller, H., Helw. Chim. Acta, 21, 1161 (1938). (9) Moore, L. A., and Ely, R., IND.ENG.CHEY.,ANAL.ED.,13,600 (1941). (10) Morton, R. A., “Vitamins”, London, Adam Hilger, 1942. (11) Parker, W.E., and McFarlane, W. D., Can. J. Research, B 18, 405 (1940). (12) Quackenbush, F. W., Gottlieb, H. L., and Steenbock, H., ISD. ENG.CHEM.,33, 1276 (1941). (13) Scudi, J. V., and Buhs, R. P., J . Bid. Chem., 141,451 (1941). (14) Smith, L. J., Chem. Rev., 27, 287 (1940). (15) Wall, M. E., and Kelley, E. G.. ISD. EXG.CHEM.,ANAL.F h . , 15, 18 (1943).

M e t h o d of Evaluating M e t a l Cleaners SAMUEL SPR,NG, HOWARD 1. FORMAN, AND LOUISE F. PEALE, Frankford Arsenal, Philadelphia, Pa. A quantitative method for performance evaluation of alkaline metal cleanevs is described and discussed. Reproducibility is rather good. The method involves coating of metal panels with various oils b y a specific dipping and drainage technique, followed b y a carefully controlled cleaning and rinsing procedure. The panels are covered with a fine spray of water, which condenses as droplets on the oil-covered areas, providing a pattern that remains constant for a sufficient time for a sketch to be drawn on paper divided into 100 squares. The average value for cleaned area of 5 panels is the cleaning efficiency index. Conditions influencing results and variations in the procedure are discussed.

T

H E importance of having an d e q u a t e method for evaluating alkaline cleaners for thc removal of contaminants from metal surfaces is generally agreed upon. Morgan and Lankler made an important contribution in this direction (3) in 1942 in devising a semiquantitative method that involved photographing fluorescent oil residues under ultraviolet light after a standard cleaning procedure. It was applied specifically to the removal of mineral oil by alkaline salts containing a sur face-active agent of the sodium

keryl benzene sulfonate type ( 2 ) . This procedure is rather unwieldy, particularly for the control and procurement of alkaline cleaners. The “water-break” method has been used for a long time as a criterion for evaluating metal surface cleanliness. This test is based on the ability of metal surfaces to sustain an unbroken film of water when “chemically” clean. It has not been found adequate, since the water-break pattern was observed to be dependent on the thickness of the water film. Smaller and smaller areas sustaining a complete water film were obtained as the water drained from the panel. As these areas began to reach a more or less steady state, evaporation of the water became a factor in obscuring the results. As a result of these factors, evaluations performed with the water-break test, as normally used, did not provide an adequate estimate of the efficiency of metal cleaners. For these reasons, a method has been devised which is fairly simple and has been found capable of yielding results of good reproducibility. PROCEDURE

Some of the modifications desirable for application to specific problems are obvious. In general, the method used a t the arsenal

t

202

Vol. 18, No. 3

INDUSTRIAL A N D ENGINEERING CHEMISTRY ..,"

"

.._,"~

with a viscosity a t 100" F. of.470 seconds Saybolt Universal viscosity. In the other case, the coating is obtained from a solution of a sulfurized lard oil, containing approximately 12% sulfur, dissolved in toluene to the extent of 1part of sulfurized oil to 9 parts of toluene. In the work performed with this procedure, the average weight of oil on each panel after coating with mineral oil was 0.283 gram and after coating with sulfurisod fatty oil was 0.038 pram. Maximum variations were % 4 ' for the mineral oil and

active agents rare dissolved in water t o make 2 liters of solution of

the temperature dE the solution is maintdned a t 60" + l o C. CLEANING AND RINSING OPERATIONS. The panel holder is inserted through a cover which tends t o reduce evaporation, and connected t o a small motor (Figure 2). The panel is positioned in the cleaner so that a t least 2.5 em. (1 inch) of the solution are above the top of the panel, t o reduce the effect of temperature

50" C. It is withdrawn from the water once each minute durina a

at (he panel from a distance of approximatdy 60 cm. (2 feet) by means of an atomizer connected to a compressed air line. (The head of a De Vilbis bulb atomizer was attached to a compressed air line equipped with a redlioing valve. A small spray gun such as is used for spraying paint, was also employed. The air'is filtered through glass wool.) The oil-covered regions are delineated

Standard doviation = Figure 1 .

4""' n-1

d = deviation of individual values from the mean n = number of 0hservatio.m

Coating Technique

involves coating of panels by a controlled dipping and drainage technique, followed by carefully controlled cleaning and rinsing

_.._

anmatinns T h m nsn& _r___"_l__l_

31s

-

than s n h i w t d to nmv ~ of ".._.. " _ a fine ...._9~ -. 111"11"11

.

-

~

water, which has the effect of sharDlv . . delincatine- the oil-covered regions. Quantitative estimation is made by sketching this pattern 0In paper divided into squares and counting the number of squares that arc free of oil.

P ~ E FRATION AI OF PANELS.The panels are of light-gage metal, 10 em. (iL inches) square, of cold-rolled SAE 1010 steel. Coldrolled brabss ?ralumi?pmmay also be used, Two hqles ar: $+ed in each panel GO enanie m e insertion or a a-prongra panel noiner (Figure 1) which maintains the panel in position without the use of nuts and bolts or similar arrangements. This was found neeessary because other holders trapped some of the oil, which afterwards spread and gave erroneous values. The steel panels are degreased before use with an alkaline silicate plus a synthetic detergent and rinsed with water until a completo water film is sustained. They are then pickled a t room temperature, for 1minute, in 6 N hydrochloric acid containing a small amount of wetting agent rinsed in running cold water for a short time, given two sucoehve rinses in hot alcohol containing 1% ammonia (70" to 80" C.), air-dried, and atoredin an evacuated desiceataruntilused. COATING OF PANELS.The panel is immersed in the oil to about hslf its height. Thc container is tilted so that the entire panel is covered with oil and then returned to an upright positron such that the e ~ c e s sof oil drains from around the hules in the panel. The panel bolder prmgs are inserted and the panel is hung in a rack a t 4 j a angle (see Figure 1). The panels arc allowed to drain at 25' * 2" C. for 1 hour. The globule of oil that remains on the bottom corner is removed. The panels should be shielded from drafts during the drainage period. It is desirable to determine the weight of oil on individual panels at frequent intervals to ensure control over these conditions. Two oils arc used for coating panels. I n one case, the coating is obtained from a hydraulic mineral oil of high viscosity index,

Figure 2.

Assembly before Test

203

ANALYTICAL EDITION

March, 1946

4

3

Ted Panels

..

Oil Light mineral oil Hesvy,minejal oil in toluene (1 to 3) Lard ml (Pnme No.,l) Sulfurired mineral oil Sulfurized iatty-mineral oil Sulfurizedfatty baae in mineral oil

(mirror finish) 99

97

[roughened) 89

11

17

94 95

23

84

11

88

2

variations in the temperature of the room during the drainage of oil from the panels. The frequency of this occurrence, however, is notsufficient'..:..+,."c""" ...:&L ".I+ ^F +L" .-"+L",!

.

.. .

.

.

__.. - . .

.

.. .

..

tion factor for various room temperatures during rhe hour-long drainage period. CLEANING AND RINSING OPERATIONS. A temperature of 60' C. (140' F.) was utilized in the procedure. This temperature is somewhat lower than that recorded for most industrial cleaning operations, although it is not uncommon for the temperature of production cleaning tanks to drop to this level. Cleaning efficiency improved rapidly as the temperature of cleaning was increased above 60' C. This tended to obscure differences among cleaners and the various factors affecting cleaning. up to 36 revolutions per minute, w&s ~h~ effect f,, found to be fairly small. Generally, the effectof concentration of dderable up to a minimal d u e , beyond which in Concentration caused relatively smdl improve.. .. ments in cleaning efficiency. The concentration ot 3% aIkaIIne salts in the above procedure NBS found to be above this minimum concentration for most condition?. Lower concentrations of alkaline salts tended to accentuate differences among surface-active agents. ~

S E L E C ~ ~OFOPANELS. N The surface condition of metal panels has a considerable bearing on cleaning efficiency. In utilizing this test to solve specificproblems, therefore, panels should be selected with surfaces similar t o those involved in the problem. So far us the test itself is concerned, the conditions listed above are out of the critical range for surface variations. COATING OF PANELS.It wus observed that the ease ofremoval of oil from the panels was dependent t o some extent upon the oil film that was used. Thus, specific differences were observed between sulfurized fatty base oils and mineral or lard oils under certain eonditiom. Another very important factor was the viscosity of the oil used. It may he seen from Table I1 that the ease of oil removal decreased as the viscosity of the oil and the weight of oil on the panel increased. The distribution of oil on each of the panels is not uniform. since during the drainage period a meater amount of oil is concentrated on the lower nortion of the panel. I n a great many tests, however, i t was observed that results were not consistently poorer for the lower half of the panel than for the top half, even though approximately two thirds of the total oil was on the bottom half of the panel. One of the factors that controlled the reproducibility of results to a great exteut,.in utiliaing the above coating procedure, was the temperature at which the excess oil was drained from the panel. In order t o obviate difficulties due t o this, i t was found necessary t o maiutain the drainage temperature within +2' C. at 25" C. during the I-hour period. Examination of the data in Table 111indicates that in this range the temperature effect is less than that due to random variables. Lower temperatures resulted in erratic results; higher temperatures in bctter results which were rPproducible. After considerable experience wirh a particular set of conditions, ir should be possible t o apply a corree-

Table II. Effect of Viscosity of Mineral Oil on Cleaning Efficiency Visoositv at Cleaning Weight of Oil No. of 100' F. Index on Esoh Panel Determinations

I

Figure

4. Cleaning Results

kff. Cleaning Index 73 (fair).

Right. Cleaning index 94 (good)

I N D U S T R I A L A N D E'N G I N E E R I N G C H E M I S T R Y

PO4 Table Temp.,

c.

25

25

111.

Typical Reproducibility Data

Medium Mineral Oila Cleaning efficiency Standard index deviation 56 53

Sulfurized F a t t y Oil Baseb Cleaning efficiency Standard index deviation 80 86 85 81 83

23 55 24 53 24 55 56 85 26 a Mineral oil of high viscosity index. Viscosity 470 sec. per 100' F. S.U.V. Av. weight 0.283 gram. b Sulfurized f a t t y oil in toluene (1 t o 9). Av. weight 0.038 gram. Temperature between 23' a n d 27' C. Cleaner, 3% sodium orthosilicate plus 0.15% sodium keryl benzene sulfonate (40%). Conditions, 5 minutes, 10 r.p.m., 60' C.

Variations in the rinse-water temperature were found to be not very important. I n the procedure a fairly long rinsing time is provided a t a fairly high temperature. Under these conditions it was found unnecessary to dip the panels in dilute acid, prior to the evaluation, t o avoid "false water film continuity" (1) due to the presence of some of the surface-active agent. The main reason for eliminating the acid rinse was that rusting of the panel developed after this rinse and tended to obscure the patterns of oilcovered areas that were being sketched. EVALUATION ASD REPRODUCIBILITY. While the spraying of the panel with water is not extremely critical, care should be exercised to prevent its becoming drenched with a large excess of water. Two or three fairly slow passes with a fine spray of water appear to be optimum. It is also desirable to avoid having the

Vol. 18, No. 3

water under very high pressure or the spray source close to the panel. One possible source of a subjective error lies in the sketching of the pattern of oil-covered areas disclosed by the condensation of the spray of water. Comparison of the values obtained with three different operators, on a number of occasions, indicated that this error was rather small, being less than 2%. For operators who might have difficulty in this evaluation, it is possible to provide assistance in the form of a viewing screen divided inro 100 squares or some similar arrangement. Some reproducibility data, obtained on different days over a period of several months, are listed in Table 111. The values for cleaning index fall well within the range to be anticipated from the magnitude of the standard deviation. Appreciation is expressed to Lt. Col. C. H. Greenall, officer-incharge, Maj. W. W. Culbertson, research officer, C. C. Fawcett, associate director, and E. R. Rechel, chief of the Chemical Research Section of Frankford Arsenal Laboratory, as well as the Ordnance Department, for permission to publish this paper. Special thanks are due J. W. Mitchell for his helpful review of the paper. LITERATURE CITED

(1) Harris, J. C., A.S.T.M. B'ull., 136, 39 (1945). (2) Morgan, O., and Lankler, J. G., IND. ENG.CHEM.,34, 1158 (1942). (3) Morgan, O., and Lankler, J. G., IND.ENG.C H ~ MANAL. ., ED.,14, 725 (1942). PRESENTED a t the spring meeting of the Philadelphia Section and t h e Meeting-in-Print of the AMERICAXCHEMICAL SOCIETY, Division of Analytical a n d Micro Chemistry, 1945.

Fluorometric Attachment for the Beckman Spectrophotometer MARY H. FLETCHER, CHARLES E. WHITE', AND MILTON S. SHEFTEL' Eastern Experiment Station, Bureau of Mines, College Park, Md.

A

fluorometric attachment for the Beckman spectrophotometer for use in the measurement of fluorescence in solutions consists of a light-tight cell compartment equipped with suitable Filters and a lamp housing. It uses the receiving and amplifying systems of the Beckman instrument. General Electric B-H-4 mercury lamp is light source. Lamp emission is controlled b y Sola constant-voltage transformer No. 30,852 and b y proper ventilation of lamp housing.

M

EASUREMENT of the fluorescence of solutions as an analytical procedure has been confined primarily to the field of vitamin chemistry, and for this reason, most commercial fluorometers have been designed for use with the brilliantly fluorescing solutions encountered in this field. For weakly fluorescing solutions, such as those encountered in the determination of beryllium @), the commercial instruments a t hand proved to be insufficiently sensitive for an accurate measurement. Since a Beckman spectrophotometer was available, it was decided to adapt it for use with these solutions. The receiving and amplifying system of the Beckman instrument ( I ) possessed the desired characteristics with respect to sensitivity and range, and consequently it could be used with a minimum of change. All that was needed to make it operative was the addition of a source of ultraviolet light and a compartment for holding the optical cell and filters. An attachment which consisted of a light-tight cell compartment and a ventilated lamp housing was built from sheet brass. The interior of the cell compartment was painted with a nonfluorescent flat black paint. The principal parts are shown in Figures 1 and 2. The original cell compartment was removed from the Beckman instrument and replaced by the attachment, the cell compartment of which was bolted to the original photo1 3

Address, University of Maryland, College Park, Md. Present address, Goring Products Co., Kenilworth, N. J.

tube housing. The arrangement is shown in Figure 1, in whlch 1 is the lamp housing, 2 the cell compartment, 3 the phototube housing, 4 the Beckman spectrophotometer, 5 the ventilating fan, and 6 the ventilating louvers. The ventilating louvers occupy two of the outer adjacent sides and the top of the lamp housing section. An opening is provided on the third side to permit a connection with the fan. Figure 2 is a plan view (covers removed) of the fluorometric attachment and shows the relative positions of the various parts of the instrument. The light source, a General Electric B-H-4 mercury lamp, was chosen because previous experience had shown i t to be generally satisfactory. Another possible light source, the hydrogen discharge tube furnished with the Beckman instrument, was tried on the beryllium-quinizarin solutions and proved unsatisfactory because of the very low intensity of the exciting wave lengths. On t h other hand, the B-H-4 lamp because of its strong emission a t 3660 A. is an excellent light source and provides an intense radiation of the proper wave lengths for an efficient excitation of fluorescence in solutions of the type under consideration. The emission of the mercury lamp is influenced by the voltage and ambient temperature of the lamp; therefore, t o ensure constant emission, both voltage and temperature regulation on the lamp were found necessary. For the former, a Sola constantvoltage transformer No. 30,852 specifically designed for use with the H-4 lamp was employed. For the latter, an Eastman darkroom ventilating fan, Model A, was used.

Table

Input Voltage 110 120 130 Spread, 20 volts

I.

Stability

Reading for 22.4 Micrograms of B e 0 in 25 M1. of Solution 98.0 101.4 103.8 5 . 8 scale divisions

Reading for 11.2 lllcrograms of B e 0 in 25 M1. of Solution 46.8 48.0

50.0 3 . 2 scale divisions