111-DUSTRIAL A N D ENGINEERING CHEMISTRY
June, 1928
of Standards, is published through the courtesy of Director George K. Burgess (Table V). Table V--Analysis of C o t t o n s e e d Hulls P e r cenl
Moisture Ash Crude fat Crude protein Crude fiber Xitrogen-free extract
12.27 2.09 0 SO 2.88 32.49 49.47
Per cenl 0.46 22.23 43.14 37.96 20.91
Nitrogen Furfural Pentose Pentosan Lignin
The production of cottonseed hulls is over 1 million tons annually, all of which are readily accessible should the chemist find valuable uses for them. During the past two years the practice of harvesting cotton with “sleds” has brought a large amount of waste to the gins. This material is principally cotton burrs. It is estimated that from 250,000 to 400,000 tons of this stuff were piled up around the gins in the Southwest in the fall of 1926, with much less in 1927. An analysis of this material by the Bureau of Standards showed the following amount of cellulose: Total cellulose: Air-dry Oven-dry Alpha-cellulose : Air-dry Oven-dry
Per cenl 42 8 47.2 29.2 33.0
Cottonstalks are not likely to become a source of cellulose for manufacture unless some method of harvesting cotton should be found which xill bring the stalks together as a straw-pile is now accumulated. At any rate, cottonstalks do not often weigh much more than 1 ton per acre.12 Boll12
Fraps, Texas AgY E x p t Sla , Bull. 247 (1919).
591
weevil control calls for low-growing cottonstalks. Cellulose production would call for tall, rank-growing varieties, and the man who would be sure of his cotton crop cannot have both; so he will probably go in for the smaller varieties. Sugar-cane bagasse was used only as fuel until the Celotex Company made it valuable as raw material for insulating board. The sugar-cane mosaic threatened the extermination of the sugar business in Louisiana until the introduction of the Java canes squared it away with a new lease on life. From average yields of 10 tons of sugar cane per acre with the old varieties, yields were recorded last year up to 43 tons from the new “P. 0.J.” canes. For each ton of sugar cane taken to the mill about 500 pounds of wet bagasse are produced, carrying 45 to 54 per cent of water. There are no analyses showing the cellulose in the new canes grown in Louisiana, but an analysis of Louisiana Purple cane13 gives the cellulose of that variety as approximately 50 per cent of the dry-matter content by Cross and Bevan’s method. The P. 0. J. canes are said to contain about one-third more fiber than the old varieties, of which Louisiana Purple is an example. Louisiana planters estimate that 3 million tons of sugar cane will be grown in that state in 1928, which will produce 750,000 tons of wet bagasse, or 375,000 tons of bone-dry material. The Celotex Company is said to have contracts with about 75 per cent of the mills in Louisiana, but is not expected to use more than half of the bagasse produced in 1928-29. The company will not have to import bagasse next year. 13
La. A g r . E r p l . Sta., Bull. 9 1 (1907).
Accelerated Tests of Organic Protective Coatings’,’ Percy H. Walker and E. F. Hickson BURLACOF
STANDARDS, W A S H I N G T O Y ,
D. C .
Apparatus for exposing organic protective coatings to artificial light, water, and gases is described. A variety of materials, including varnishes, oil paints, enamel paints, lacquers, bituminous saturated felts, bituminous roofing materials have been tested in a proposed accelerated cycle. As judged by visual inspection, the nature of the breakdown is remarkably similar to the breakdown of the various materials on weathering. There is chalking with paints that chalk in service, cracking with those that crack, and similar changes in color. The characteristic differences in behavior of various asphaltic mixtures observed on outdoor weathering are duplicated in the accelerated cycle. The same may be said of varnishes and lacquers. In fact, the duplication of weather effects has been remarkable with all types of materials tested.
While failure in the accelerated cycle is similar to and more rapid than that on weather exposure, no definite ratio has as yet been fixed between the accelerated test and weather test. The difficulty of determining the relative condition of protective coatings is discussed and the unreliability of opinions based on visual inspection is pointed out. Several methods of quantitatively measuring the extent of failure of such coatings are described. These include measuring the amount of water vapor under definite conditions or amount of air under definite pressure passing through coating on wire gauze, and several plans for locating and measuring breaks in a coating on metal by electrical means.
HE rapid testing of coatings to determine their probable
in use for subjecting test coatings to a variety of agencies which cause deterioration will first be described. Methods of determining the degree of deterioration will then be discussed.
T
relative durability under service conditions involves two distinct problems. It is necessary first to subject the coating to conditions which will cause rapid deterioration of the character encountered in actual service. It is then necessary to determine the degree to which the coatings haye failed under the test conditions. The equipment now Received March 12, 1928. Presented before the Division of Paint and Varnish Chemistry a t the 75th Meeting of the American Chemical Society. St. Louis, Mo., April 16 to 19, 1928. 2 Publication approved b y the Director of the National Bureau of Standards. This paper is condensed from a somewhat longer Technologic Paper of the Bureau of Standards now in press. 1
Methods Used to Cause Accelerated Disintegration Light, moisture, temperature changes, and varying small accidental additions t o normal air are the important causes of decay of paint, varnish, and other organic protective coatings. I n general, the most important of these is light. APPARATUS-The chamber for exposure to light and moisture consists of a rotating cylinder made of No. 16 gage
galsanized iron, 76 cni. (30 iiicbes) in diameter, 38 cm. (15 inches) high, open at hoth ends, with tire light suspended in the center. This size was selected in order to bring the light as near the panels as possible, at the same time avoiding too high temperatures. The cyliiider has a capacity Sor sixtl, i.5 by I5 em. (a by 6 inch) panels. The test. panels are plnced i n two tiers iinrnediately opp0sit.e the light inside the cylinder, the panel? thus being 38 em. (15 inches) frrom the center of tlie light source. This gives a temperature of about 5"" t,o Xio C . at the panels vith the type of lamp used. Thirty slillteil holders, 7.5 ern. (.3 inches) u-idc and about 33 cui. (13 inches) long, for panels 7.5 em.
(3 inches) wide are atlaclied to tlie inner surface of t.he open cylinder. 'The exposure cylinder is provided with wat,er sprays: so ilint it k possible to EXpiiSe t.he panels in succession to intense light. and to a variety OS moisture conditions. A pan placed about 5 cm. below the hott.om of t,he light eylindor contains water and serves to keep the teinperat,ure next to the panels doivn to about 52" C.:as well as to humidify the air. Separate cabinets for exposure to gases and refrigeration are provided. LIGBTSOURCE -It is ad ible to choose a powerful light and one whose average iiitensity and spectral distribution will remain fairly eoristant. Whether or not t,he relative resistance to exposure of various coatings will be the same under sunlight as under a. liglit of very different spectral dktributiun has never been Jeteriiiiued, and urrtil it i s dsheriiiincd it seemed desiralble to use a light souriie as sirnilnr to sunlight as possible. Ouiiig t o absorpbion by the earth's atmosphcre, nu radiation of uiave length shorter than 290 nip roadies tlie eart1i.s The quartz-tube, mercury arc light, which liaij been advocated by Nelson and otiiers,' has not been used as a source of liglit in this work. Most of tlie work recorded iii this paper lias been (lone wit11 an enclosed type esrb(m arc light operated at 220 volts d. c. at 13 amperes. Since the light. is oi the enclosed type tile carbons can be operated %-it11t,his current about :14 hours wit,Ilout renewal. The glass globe sliould be clemcd on cacli renewd of the electrodes. Figure 1 sliows, in comparison with sunlight, Ihe spectrum of tlie arc used with tlie glass globe in position, taken on a panchromatic film in a quartz spectrogrtqh. I t is to be noted that the band spectra of both the arc and sun become more apparent as t.he time of exposure is increased. Note also
' Bur. Slondordr, Sei. Paper 859, 692. A ~ r o r A. ~ SOC. . T C S L ~ ..ticrieriuis, W aa, 920 w w : a@,PL.11, sa3 (I@%).
~ t ir, . 4s5 (1922);
a4. Pt. 11,
that, while this arc spectrum sliows uiuch radiation in the moderate ultra-violet (350 to 400 mp), it shows none of the very short waves not found in sunlight but given by the quartz mercury arc. The electrodes used are of solid carbon 12.7 mm. in diameter. The glass globe is opaque to ultra-violet wave lengths less than 320 mp and infra-red radiation longer than about 4500 mfi; hut the large arnount of infraaed radiation emanates from tlie glass globe that becomes heated by the arc. The total radiation at a distance of 38 em. (I5 inches) Srom the center and at right angles to the arc on 12.6 amperes i s 0.075 watt per square ceritiineter. Figure 2 shows two light units. In the left-hand unit the light i s shoxn hi position in the center of the cylindrical sample holder, II. In the right-hand unit the light, I,, is raised out of t.he cylindrical holder. Ram-In addition to light exposure the panels should he subjected from time to time to several hours of vigorous spraying with warm (100' F.) water. The panels can be removed to a special diamber for this purpose, but it i s inore converiient,t.o produce this artificial rain in the same clrainber used for the light exposure. For this purpose the liglit is r a i d out of the cylinder, ti coininon rotat,ing lawn spriukler put in tlie hott,orn of tlio cylinder, the bracket supporting tlie liglit protect& from spray by a brass cylinder, and tlie top of the cylinder holding the saniple covered by galvanized iron
re 3 sliows tlie arrangement for spraying the pmels with wit.er. The riglit-liand unit with the cover removed shows the rotating spray, the slots containing test panels, and tlic protecting tube over tlie lower part of the lanip suppwt. The left-hami unit shows the lid in place. In onler to simulirte a hot. lruriiid climate a fixed water spray Iias been niouiited in the b n k iri sue11 a manner that it may fuiiction wfiile the lights are operating. Connections for t h i s spray are made at the tee above and betveeen the t,wo tanks (Figure 3 ) . The light cylinders are also now nixie to rotate sloirly s t tlie rat,e oi three revolutions per liour. :rhus with the lights in operation arid in position (Figure a), a gentle spray ea11 wet, t,lre test surfaces periodically while exposed to the light.
F i ~ u r e2--Liehf Eapoauce Equipment
,.1
GW'BIU.TUAE CH~NGES (12~1.2iro~aa~io~)-.~,, mmnoiiia coil about 27 by 47 by 40 em. in an insulated chamber that. can be cooled to -25' C. (-13" F.) is used t,o chill the t,est panels quickly. On removal from the liglit chamber tlie test panels are plnced on tlie rack shown, and ihe rack with samples is
placed in the refrigerator arid left there for 1 hour and then removed. s--It is most convenient to transfer the panels to a separate cabinet for exposure t,o various gas mixtures. Such a cabinet for exposure to ozonized air is shon-nin Figure 4. Air is forced, by meaneof the small motor-driven fan shown in the lower left corner, through a silica-geldeliydratorr,then through a calcium chloride bottle, which swves to indicate that. the silica gel is working efficiently, to the ozonizing apparatus, and thence to the glass chamber on the right. for holding the panels. Some n-ater is placed in the bottom of this glass chaniber t o moisten the ozonized air. fly suitable manipulation of the stopcocks in the pipe line the rate of flow can be determined hy the flownreter, shown to the right. of the silica-gel holder, and samples for determiniirg the ozone concentration can he take11 through the pipe projecting toward the front. This equipment delivers about 660 liters per hour of air containing ahout 0.08 per cent of ozone by volume. METWJDor PI~EPAIIING TCST PANEIk-It is a waste of time to make tests on single coats of protective coatings. Tbis can be readily demonstrated by brushing and flowing paints on glitss and, after drying, examining by transmit,ted light. It is advisable always to use at least two and generally three coats on tlie exposed test surface. When using wood or nletal t,he back and edges of the panels are given three coats of aluminum paint (25 grams of polislicd ahimimim powder to 100 cc. of long-oil water-resisting spar varnish). This aluniinum paint is waterproof and gives exaelleilt protection to both wood and steel. Since during exposure it is not, exposed to light, it may he relied upon t o last Ioirgcr than any material being tested oil ilie exposed surface. In addition to tests ori wood or metal it is advisable to apply tlie roaterial to 10Umesh wire gauze or to cotton cloth. This will lie discussed later in connection with methods of rletemiining the extent to wliich disint,egration has taken place. I)):SCRIPTIOL. OF EXIWSURN CYcu:Since light is the most important destructive agency in exposure tests, it seemcd neccssmy to zn&e use of as much light as possible. Therefori!, the night, hours (17 hours) wcre given over to light. The idea in subjecting tlic panels t.o air containing a sinal1 amount (0.08 per cent) oE ozoiie was to accelerat,e surface oxidation. I n using refrigeration, the desired factor seemed til be a rapid change in temperature, rather than any dcfinite time of exposure to low temperature. Tho refrigerating unit was not assembled until some time after the lightexposure cylinder and gas chamher were made. In some of the earlier t,ests the panels were therefore not exposed to sudden temperature changes. With the weathering cycle I M ~ in T use t,he exposure j r i a week is as follows: Kcfrieeiation Ozoiiizcd air
wstw Light
riou*r
Pe. Len,
4 17 32 112
10.3 19.4 67.9
2.4
It will be noted that in this sclredule only 165 out of 168 hours are accounted for. The remaining 3 hours are used for the inspect,ionof the pancls. The light periods range from 3 to 41 hoiirs, t,he water from 3 t.o 17 hours, ozonized air from 2 > /to ~ 3l/2 hours, and refrigeration 1 hour. Methods of Determining Extent of Disintegration
O ~ ~ S E ~ W ~ in W I ExrN:nTs~~--tomiietcirt IIX experienced O ~ I servers frtquent,ly draw widely varying conclusions from insppetion of the same panelr and the same ohservers draw vsrying rrinciusions at different times.6 * Pro
