Document 2qqXwwbxOQq5g3bB4krvq34rR

E. I. DU PONT DE NEMOURS & COMPANY KREBS PIGMENTS DEPARTMENT 256 VANDERPOOL STREET NEWARK. NEW JERSEY Serial No. KN-51-41 Copy No. 1. Numerical NEWARK PLANT PIGMENT COLOR RESEARCH REPORT "AURIC BROWN", F-4-P - PROCESS STUDY Period Covered FEBRUARY - AUGUST, 1951 FILE: DATE, 210.2 10/15/51 4 Copy to: il - Numerical file 2 - Research Office (210.2) % * Library File (210.2) 4 A. Siegel/J. N. fully, Newark 5 * H. Arnottl, Newark 6 - R. A. Hageman, " 7 - B. H. Perfcins/R. R. Denslow, Newark 8 - Extra 9- n Serial Ho. KN-51-41 Copy Ho.1. Numerical NEWARK PUNT PIGMENT COLOR RESEARCH REPORT Progress Report "AURIC BROWN", F-4-P - PROCESS STUB! FEBRUABT - AUGUST, 1951 CHARGE: 1100-31-8 BATE SUBMITTED: 10/2/51 BATE ISSUED: 10/15/51 DUP050068642 mmmm* This report summarises the results of a process study carried out on "Auric Brown", F-4-P during the period February * August If51* . The work was initiated as a result of complaints from F. & F. on the quality of F-4-P shipments, especially during 1950* Zt has been shown that the poor quality of our shipments was due to the character of copperas used in manufacture* Specifically, material currently received from Sdge Moor is considerably higher in manganese content than that used several years ago in our original work* This is the result of a change in titanium ore (Qullon to Trail Ridge)* Products which are satisfactory to Parlin have been made from Baltimore Copperas (high Kg, low Mn} and from Stanley Copperas (low Mg, low Mn}* In the latter case it is necessary to add magnesium sulfate* Products of fair (border line) quality have been made from Sdge Moor copperas (low Mg, high Mn) by adding magnesium sulfate and precipitating out the manganese with potassium permanganate* An apparently satisfactory testing method has been developed in ooopor&tion with P&rlin* Sew limit dry plate (masstone) standards have been set up. Indications are that products meeting these standards can be manufactured without difficulty* A new plant formula ("D" process) has been written incorporating the information presented below. A summary of this process is attached as an appendix* m m*m The manufacture of F-4-P has been under the supervision of the Production Division since April 1949* At that time, ten batches had been made on a single formula and evaluated critically by Parlin, Seven were rated as satisfactory} two of the others were of borderline quality; one was definitely yellow and out of line in transparency* Since then, a number of adverse comments have been received on shipments*. During 1950, the complaints became more serious and the Chemical iivielon was requested to take up the problem again* Active work was started in February 1951* Manufacture of the pigment had been carried out on the formula used in the above standardisation run and there was no evidence of deviations from the specified operating procedures. Parlin'e complaints have been based on their difficulties in matching their "Metalllchrome" color families* Specifically, it was stated that our shipments of pulp when dispersed in single pigment DUP050068643 --2-- lacquers were usually too yellow and dull ("milky") in masstone. There were also indications of poor reproducibility, The evidence was some what confused but it was apparent that their difficulties were real and were largely the result of abnormal pigment quality* Our work had two objectives: (1) To determine definitely the type of product desired by F. & 7* and to set up a testing method for evaluation purposes* (2) To establish a manufacturing procedure for consistently producing material of the desired type. Close cooperation with Parlin has been maintained throughout the work. The basic Parlin patent on flushed lacquers is U.S. 2140745 (B. T. Bucks). The use of hydrous ferric oxide of the F-4-P type is covered by U.S. 2335760 (also Bucks), The claims of this latter are for coating compositions; the general method for making the pigment Is disclosed. The Barmon (Tesce) "Bough Gold" patent is U. S. 2,384,579. The product claims cover coating compositions rather than the pigment. In our opinion, nothing of a patentable nature has been developed as a result of the present study. mmmm. j?g.JTA^I^BD8 The original "Hetallichrome" color families were set up several years ago using a blend of semi-works F-4-P batches. To the best of our knowledge, these are still the goal standards. However, it appears that in some eases, compromise standards and some new shades have been set up by Parlin since then, using random batches of pulp; no details (lot numbers, etc.) are available at Hswark. All work on the problem of standardisation and reproducibility has been complicated by two factors; (a) color drift and (b) inadequacy of testing methods. Aged samples of pulp give enamels which are lighter and more opique in masstone and redder in tint than enamels made from the fresh pulp* The redness is easily seen in a metallic and the opacity shows up as dullness of "milkiness" and lack of two-tone flash, the serious ness or even the existence of this drift was not known to either Parlin or ourselves until well along in the research program with the result that many of our early conclusions are open to question. Some drift occurs during processing of the pigment and immediately after manufacture; DUP050068644 3 however, it la ordinarily not too serious for a period of 3-4 weeks. Since little or no drift occurs in the lacquer vehicle, it is Pariin's practice to disperse the pigment into semi finished or finished lacquer as soon as possible after it is received. Aside from its effect on any stocking programs, the chief effect of color drift in our work is the impossibility of setting up a pulp standard for matching purposes and the consequent necessity for Batching against panel standards of some type. Belatlvely little work has been done either in the past or during the recent study to reduce the amount of drift and it is uncer tain whether or not it represents a workable problem. Its physical explanation is apparently a particle else growth. The testing of P-4-P (experimental samples and plant production) has always been a serious problem and the methods used have been open to question either in respect to practical significance or reproducibility. The final test has been Parlin*s ability to use the pulp to match the particular shades in which they are interested at the time. They have made use of a laboratory flushing technique which is quite time con suming; moreover, correlation of the results with those obtained in their plant has not been entirely satisfactory; this and the matter of reproducibility have been the subject of considerable discussion. Additional complications have been introduced by various changes in Parlin*s plant dispersion technique. As a guide in our work, considerable use has been made of a laboratory flush-out technique (in varnish) in which the tint is compared with dry plate standards. This was written up as TP-7007-1. This test is of some value in predicting, in a general way, what type of "metallic" will be obtained from a given pulp. It also has some value in intercomparing members of single laboratory series. However, it has occa sionally failed in spotting batches alleged by Parlin to be unsatisfactory. Because of this and the obvious difficulties in correlating results with actual lacquer tests, the method has been abandoned except for certain research purposes* As indicated above, Parlin has made various changes in their plant dispersion techniques and in their laboratory dispersion methods. Of particular significance is ths fact that their current criterion of quality is masstonscolor (lacquer). Metallic* are made up only occasionally ana then only to guide their plant color shaders. At the beginning of the present study ws were supplied with a new laboratory dispersion procedure for testing work* It involves flubbing the pulp on a two roll mill In nitrocellulose/plasticiser, sheeting out and drying. The dried roll stock is stirred under specified conditions with the additional lacquer components (solvent, resin, etc.) and the resulting masstons lacquer sprayed to complete hiding on card board panels. Sufficient work has been dons to indicate reasonably good dupllcabllity both between check beets at one laboratory and between teste carried out here and at Parlin. As a result of the work, Parlin DUP050068645 -4 has sat up new light and dark limit standard panels. These were made from single pigment lacquers prepared from the following recent batches of F-4-Ps Light limit t Lot 20450 Dark limit : Lot 20579 Products siren darker than the "Dark limit" would probably be acceptable, if otherwise satisfactory in color. Parlint chief objection is apparently to yellowness and dullness or "milkiness". Batches lighter than the flight limit" would normally give difficulty because of their opacity which is reflected in lack of "flash" when the lacquer is metallised. Parlin has carried out some work fairly recently designed to cut the time required for testing by using forced drying to remove water from the roll stock. This tentative short method carries their code TM-69-B. Further work on this method is in progress. Attempts were made (at Newark) to develop a laboratory dispersion procedure based on the use of a small amount (ca.5C^ of pigment. Flushout e were made in 1/2 pint ball mills and the jars were subsequently rolled with the other lacquer ingredients. Products of good gloss were obtained but we were unable to match the depth of 2 roll disperiions (Bef. 1373-77). saamg o f u bo bat o r y morass The Work on lacquer testing, discussed above, was carried out concurrently with studies of miscellaneous process variables. In most of the latter, it was necessary to depend on spatula flnshouts as described in TP-7007-1. Generally speaking, there was considerable irregularity in the results, if and when farther work on this problem is required, a number of the experiments should be repeated, evaluating the results by lacquer tests. The major portion of the work, however, was concerned with the effect of impurities in copperas, in this case, the spatula flushout tests have been supplemented by lacquer tests and the conclusions are believed to be valid. .. Copperas quality was investigated briefly some years ago in connection with the original development work on F-4-P. At that time, the chief impurities appeared to be titanium and magnesium. The latter was observed to have a powerful effect; however, since it was found to be present in batches of satisfactory tinctorial quality, uo consideration was given to methods of removal or to detailed studies on the relation between amount of metal and color properties. The effect of titanium (normally present in duPont copperas to the extent of 0.2-0.3%) le similar to that of magnesium but much less marked. It appears to run fairly constant in amount and is ordinarily no cause for concern. The recent work has confirmed the above but has Ohown that DUP050068646 5 certain types of copperas contain significant amounts of manganese as well* - F-.4-P made in the laboratory from CP Copperas is very light and red in masstens (2 roll lacquer dispersion) and correspondingly opaque* This effect was cheeked in the plant using a rather pure commercial grade* This light* red material is unacceptable to P. & f* As indicated above, magnesium, has a powerful effect giving a dark masstens and correspondingly weak, yellow tint. As little as ca 0.0556 of metal added aa magnesium sulfate to OP copperas has a markid effect. To obtain standard color <soo previous section), ca 0.4-0.5% (as Mg) is required. The mechanism of its effect is not known but is believed to be physical rather than chemical, resulting in a pigment of smaller particle sise. The affect of manganese is also marked. Like magnesium, it Sives depth; however, the most noticeable result of its pressncs is sllneas of the type regarding which F. & F. has complained. While as little as ca 0.02$ of metal (as Ms) has a noticeable effect on color, satisfactory products can be made with as much as 0.05-0.06$. The upper limit of tolerance is not known; 0.15$ is definitely too much. It has bean concluded that to make acceptable F-4-P, the copperas must contain a fairly high percentage (0.4-0.5$) of magnesium and as little manganese as possible, say, 0.06$ maximum. Titanium in normal amounts (0.2-0.3$) is unobjectionable. Presumably the pigment used in setting up Parlin*s original standards was made from copperas of this type. PmimWJismsuL Because of a change in titanium ore (Quilon to Trail Bldge), the composition of Edge Moor copperas (N-391) differs from that made trior to circa 1950"current material contains about 0.1$ Mg and 0.16,11$ Mn. Prom the above,It is obvious that this will give light, dull pigment. Current Baltimore copperas (8-724, formerly B-391-S) is high in magnesium (0.2 - 6.5$ Mil and low in mangansse, 0.05$ Mn. All batches mads from Baltimore copperas have been satisfactory. Since few analytical data are available, it is being recommended that for the immediate future . shipmenta be analysed and the magnesium content adjusted by adding magnesium sulfate in the strike. Several plant batches were made from Stanley Copperas (8-391-SS), a pure grade low in both magnesium T< 0.05$ g) and manganese (ca 0.05$). By adding a sufficient quantity of magnesium sulfate this has given satisfactory pigment. A sample of copperas was received from National Lead. This was high in manganese and was therefore given only very limitedattention. DUP050068647 -6 The present filtration procedure for N-24 liquor removes most of the titanium hut does not remove magnesium or manganese which are present as soluble sulfates. ho method is known for removing magnesium aside from reeryatallisation. There have been indications that it may be possible to counteract its effect by adding Sodium fluoride* Additional work would be required to establish amount and the optimum procedure. Since the amount of magnesium required Is relatively high* the problem of its removal is not believed likely to arise. Two methods for the removal of manganese have been investigated. The first involves oxidation of the manganese with lead peroxide after oxidation of the iron has been completed. A dark colored sludge (MnOg ?) le formed but the main reaction it the following* 2 MnS04 + 5 Pb02 + 3 H2S04 Z 5 PbS04 + 2 H20 + 2 H Mn04 The permanganate formed should be unobjectionable. Because of cost considerations, more work has been dons on a second method, namely adding potassium permanganate to the oxidised iron solution. 3 MnS04 + 2 K n04 + 2 H20 - 5 HnCfc + KgS04 + 2 H2S04 Separation of the solution from the major portion of the eludge presented no great difficulty In the laboratory and both methods appeared to work satisfactorily (in the laboratory) with Idge Moor copperas. The second method was tried with fair success in the plant. However, with existing equipment, considerable sludge ran into the strike vat. Possibly because of this, the resulting pigment was of borderline quality, being slightly dull. It is probable that a satisfactory purification procedure might be worked out but additional equipment might be required. Normally, no difficulty is encountered in oxidising the iron with the specified amount (lOffof theory) of chlorate* With th#! theo retical amount, a slight test for Fa ^ was obtained but the pigment was normal. A large reduction (to $5%) in chlorate gave a dark product, presumably ferroso ferric oxide. At an intermediate point (92m) S Very red product, resembling Mapico Red rather than F-A-P, was obtained. It is recommended that the present (10? %) figure be. retained. Sodium pyrophosphate (e.g. l#/moi. copperas) added to the iron or soda ash gave very dark yellow material. The effect of sodium fluoride in giving light, red pigment has been mentioned above. The experiments with this material gave rather DUP050068648 7 irregular result0. further work is indicated especially in respect to the effect of sodium fluoride on magnesium and manganese. Some additional work has been carried out on the effect of the iron/soda ash ratio as judged by the pH of the slurry. Increasing the JH from 7 to 7.5-8.0 gave a lighter masstone. Reducing it to 5*5 a considerably darker masstone . It is recommended that the prei practice of adjusting to pH 7.0 after the strike be continued. four plant runs of five batches each wars followed and tested in lacquer. In most cases teste were mads both at Hewark and at Farlin. Agreement between the results, while not perfect, was fairly good and was considered satisfactory. The procedures used are indicatsd in Table I which should ha self-explanatory, Three of the hatches were made by standard formula using Bdge Moor copperas (N-391). They ell showed the masstone dullness regarding which farlin has complained and which is characteristic of products mads from high manganese copperas. The hatches mads from Baltimore copperas (high Mg, low Mn) were all satisfactory although one (20450) was unexpectedly light; this material represents the new light limit standard. Stanley copperas (low Kg. low Kn) used *as is* gave e light red product similar to what la obtained with Of material* By adding magnesium sulfate to approximate the composition of Baltimore copperas, satisfactory products were obtained. The attempts to remove or inactivate the manganese in Bdgs Moor copperas by permanganate treatment after oxidation were only partly successful, hot 27090 was quite light presumably because of iasufflcient magnesium or possibly because of the bisulfite which whs added after the strike to remove HnOo . Bisulfite was not added in the later rune since laboratory work Indicated it to serve no useful purpose; the sludge was allowed to settle and an attempt made to keep as much as possible from running into the strike vat* In these latsr runs, the masstone depth was satisfactorily adjusted by adding magaeeium sulfate; however, they were all slightly dull compared with the standards or batches mads from Stanley copperas plus magnesium sulfate. Additional work would obviously bs required to develop an entirely satisfactory process based on the use of Bdge Moor copperas. As mentioned above, farlin stated that the color of lot 20459 was satisfactory as a light limit standard* At the same time* lot 20579 (also from Baltimore copperas) was selected as a dark standard. It was indicated that material even darker than lot 20579 would be aoeeptable provided that it was not lacking in intensity. Observations made during these plant studies indicated the making operation to be under good control. The batches made early in May showed evidence of slight contamination In the press room and one corrected SSJscondition? d# *** DUP050068649 a H PLANT BATCHES BADE UNDER CHEMICAL DIVISION SUPEEVIS DUP050068650 9 FUTURE WORK: On the bauia of the work covered in this report,a new F-4-P process ("D") haw been written and submitted to Quality Control, the Production Division will be asked to resume supervision of the process. So laboratory work* aside from testing plant batches* has beet carried out since June and no additional work is contemplated for the present. Should it be necessary to take up work at some future date* attention might well be given to the followings Further studies on copperas purification. (Removal of Manganese} Further studies on physical variables during manufacture* for example temperature* strike time* agitation* concentration* digestion* washing* water content of pulp etc. A number of experiments along these lines have been carried out out with inconclusive results, possibly due to inadequate testing methods. gffect of foreign ions (fluoride, chloride* miscellaneous cations! during the strike. Seeding - Possible effect of material from previous batch. Intentional addition of seed material. Surface treatments. RBFBR8HC8Sg Details of the experimental work are recorded in Notebooks 137$ and 1402 and in DSD 51-31. Progress reports will be found in the Inorganic Monthly Summaries beginning with February 1951* Reports on various discussions with Parlin will be found in the following Technical Service Reports: WPS ;#5) SCB #10 WPS ,#23 VPS #24 VPS #33 VPS #3?) RED |1> VPS 1#43) 2/12/51 2/27/51 5/16/51 5/22/51 6/25/51 7/20/51 6/15/51 9/16/51 Correspondence on Copperas! Moomaw - Perkins 4/24/51 Moomaw - Perkins 6/7/51 Hageman - Smith 6/11/51 Bowles - Cantwell 6/20/51 DUP050068651 10 * Establishment of Standards: H. L. Priddy/R. 0. Isele - Spengeman E/3/51 See also TSR IPS #43 above Testing method: letter: Spengeman - Andrews, 2/7/51 and attached Letter: Isele - Spengeman RRP:lg Att DUP050068652 APPENDIX rrJHP The following is & summary of the D process which should be consulted for details. It differs from the "C* process in that the use of Baltimore copperas is specified and provision is made for maintaining the magnesium and manganese content within definite limits. Batch Site: The estimated yield is 2133 lbs. This is based on the assumption that half the pstronate used is retained in the pulp. ImrsMmte* 6250# Baltimore Copperas. (Total iron exprei Manganese (as Mn) 0.06#/100# PeSOj Magnesium (as HgS04.?H20) ca 200-i 1191# Sulfuric Acid (100*), N 27 428# Sodium Chlorate* N 12 (ca.) 25# Magnesium Sulfate* MgS04.7H20, N 563 3800# Soda Ash* N 3 100# Pstronate* N 490 ^ >0 (max.) 7100# FeS04.7H20 Approximately the desired amount of copperas is dissolved in 283 vat and the volume of solution noted. A sample is analysed (#/gal.} and the total weight of 70804.7^0 in 283 vat adjusted to 6250# by adding additional N 724 or by removing a calculated amount of solution. Magnesium sulfate N 563 is added if necessary to a total of 325#/batch. After adjusting to a specified volume* oxidation is carried out by adding sulfuric acid and sodium chlorate and heating to 180*1'. Soda Ash is dissolved in 182 vat and heated to lS0*f. (Prior to use* the vat is thoroughly cleaned.) Precipitation is carried out by running the iron solution into the soda ash solution in 182 vat. 75$ is added in 30 minutes and the remaining 25$ in 90 minutes. After the strike, the pH of the slurry is adjusted If necessary to 7.0 (soda ash or sulfuric acid), the pstronate added and the batch stirred 45 minutes before pressing. The pulp is washed in the press to 1*000 ohms. DUP050068653