Document 10g6w8opQdwyEkBa2jzEk4EM5

RESTRICTED TO P & F, FILM, AND PHOTO PRODUCTS PERSONNEL WITH A NEED TO KNQW - NOT FOR GENERAL DISTRIBUTION. DO NOT DISCUSS CONTENTS WITH ANY OTHER DEPARTMENT. FOR DU PONT USE ONLY Distribution on last page Report Number:EX-71-33 Copy Number: JL3.-- Indexed File: 1865 E. I. du Pont de Nemours & Company F & F, Research & Development Division Experimental Station Laboratory Research Report FLUSHING OF PIGMENTS Date Issued: DEC 0 8 1971 Period Covered: February 1971 to September 1971 Project Number: 211104 Previous Reports: None Notebook Numbers: 538E, 567E PREPARED BY .9. c_ Cl , I. C. Chu APPROVED BY When this report is no longer needed, please return It to the File Room, F & F Department, Experimental Station. TABLE OF CONTENTS INTRODUCTION---------------------------------------------------------SUMMARY AND CONCLUSIONS-------------------------- ------------ ~ ACTION TAKEN OR PROPOSED---------PATENT STATUS---------------------------------------------------------PUBLICATION STATUS---------------------- -------------------------------ACKNOWLEDGEMENTS --------------------------------------------------------------DISCUSSION -------------------------------------------------------- Page 2 4 5 5 6 I. General Studies of Flushing Technology------- 6 A. Flushability of Presscakes----------------- 7 (1) Surface Energy------------------------------(2) Thermo Dynamic Consideration ----------------- 7 8 B. Selection of Flushing Aids-------------------------- 10 (1) Surface Charge of Presscake --------------------- 10 (2) HLB of Flushing Aids;------------------ ----- --------- 12 C. Agglomerates and Fundamental Particle Size^------------------------------ 15 (1) Agglomerates------------(2) Fundamental Particle Size---------------------- 15 16 II. Flushed Dispersions for Organosol Lacquers 16 A. Flushed Transparent Iron Oxides------------------ 16 (1) Transparent Iron Oxide Yellow (W682)~- 16 (a) Characterization of Presscake ------(b) Process Consideration----- ----------------(c) Economics ------------------------ -----------------(d) Evaluation------------ 16 17 18 22 (2) Synthesized Hydrous Iron Oxide Pigment (3) Transparent Auric Brown (W46l)-------------- 26 23 B. Flushed Organic Pigments - CPC Blue and Green-------------------- 29 (1) General Characterization (2) Process Consideration & Economics ------- 29 30 REFERENCES TABLES I - IV FIGURES 1-15 APPENDIX 1 & IX ABSTRACT DISTRIBUTION LIST I 5 DUP030013834 INTRODUCTION Most synthetic pigments at certain stage of manufacture exist as a water wet presscake. The transfer of presscakes directly from the aqueous into an organic phase is referred to as "flushing". Flushing is an old art. Some of the early literature dates as far back as the late l800's. The advantage of "flushing" have often been cited as (a) It avoids the aggregation of pigment particles taking place during the forced drying and the subsequent mechanical pul verization and (b) it gives dispersions of better quality. Flushed formulations revealed in early patents are mostly for specific vehicles such as cellulose esters, printing inks, etc. With the advent of new coating compositions, many are no longer applicable. In F & F, a flushed dispersion of hydrous iron oxide pigment was used in the 195P*s in Duco lacquer. Its use has since declined with the advent of acrylic lacquers. This report will describe our general studies of the flushing technology and specific applications in organo sol lacquers. Also included will be the comparative costs of various dispersing processes and preliminary considerations of the role of flushing in mill base manufacture in terms of presscake characteristics, pigment types, cost and flushed base quality. OBJECTIVES The overall objective is to assess the role of flushing in F & F's dispersion manufacture. The technical objectives for the reporting period have been to develop better understanding of general flushing technology and to demonstrate applicability of flushed dis persions of selected pigments in a specific product line, i.e., organosol lacquer. DUP030013885 SUMMARY AND CONCLUSIONS * A simple and economical flashing process has been developed. The process involves low, shear mixing of a mixture of presscake/solvent/flushing agent/ AB dispersant followed by a continuous centrifuga tion. The actual flushing is carried out at the mixing stage and the removal of residual water as well as coarse agglomerates accomplished simulta neously in the centrifugation step. The process should generally be applicable to presscakes that can be well deflocculated in organic medium and contain only moderate amount of agglomerates (0-40% depending primarily on the cost of the pigments). The processing c.pst was estimated to be $0.360.44/lb. of pigment vs. $0.65 for "47" process and $1.52 for two-roll milling. e The "mixing/centrifugation" process was success fully demonstrated with transparent iron oxide type of pigments. The flushed dispersions, intended for use in the organosol vehicles gave organosol lacquers of good shelf stability. However, the ultimate dispersion quality of the lacquer and that of the final film depends on the fundamental par ticle size of the presscake. This conclusion stems from investigation of the following practical systems: (a) Using a Hilton Davis transoxide yellow presscake (precursor to W682), which was found to have fundamental particles ranging from 0.05-0.20u, we obtained a flushed dispersion that gave metallic lacquer films of nearly equivalent or slightly lighter side-tone than a two-roll milled control dispersion. Improvement of the flushed quality possibly exists if fresh presscakes are used. In view of the above findings, the potential quality -2- DUP030013886 improvement with fresher presscakes and the superior shelf stability, we have recommended the implementation of flushed transoxide yellow (W682) in the organosol lacquers. 1 (b) Using a fractionated hydrops iron oxide pigment--an experimental sample synthesized in the laboratory which had a particle size range of 0.01-0.05M-, we obtained a flushed dispersion of excellent quality. The said dispersion was very stable in the organosol lacquer and gave drawdown films of exceptional trans parency and metallic panels of very good two-tone characteristics. w For presscakes that form floccules in the organic media or contain excessive agglomerates (>40$), a flushing process which includes a grinding step will be needed. Preliminary studies indicated that a "mixing flushing" or "W & P Mixing" followed by "sand grinding" would be a feasible process. The "mixing flushing/sand grinding" procedure was demonstrated on a CPC (copper phthalocyanine) presscake (precursor to W765). The flushed dis persion was superior than a "47" process control dispersion in terms of transparency and cleanli ness of the drawdown film. The comparative econo mics (for the W765 dispersions) are yet to be determined. The translation of improved quality into cost savings is highly specific to the pig ments involved and has to be dealt with on indi vidual basis. e studies of general flushing technology revealed: (a) There is a positive correlation between pigment flushability and surface non-polarity. Presscakes with surface non-polarity significantly more than 70# can be flushed easily into the organic phase. And those significantly less would require flush ing aids. -3- DU P030C1388 (b) The selection of flushing aids can be guided by the surface potential (positive or negative) of the presscak.es determined with a Zeta meter. Positive surface potential will require an anionic water insoluble flushing agent and vice versa. (c) The significance of pigment particle size on trans parency and two-tone characteristics of metallic finishes can be explained in terms of a theorectical scattering coefficient v.s. particle size curve. The curve shows that maximum scattering for iron oxide pigment occurs at 'vO.lSu and substantial trans parency for particles 0.o4p, or less. Experimental results of our studies are in good agreement with the above predictions of Mie theory. * Some considerations, based on available information, of the role of flushing as an alternative way of manufacture of mill bases are as follows: (a) The "mixing/centrifugation" process can offer sub stantial cost saving and two-roll equivalent qua lity for pigments to which it is applicable. The key factor is the availability of appropirate presscakes (in terms of particle size and compatibility of additives). (b) The "mixing (low shear or W & P) sand grinding" process can offer superior quality than "47" pro cess (at equal amount of grinding). The merit of flushing will then depend on the practical signi ficance of the improved quality and how much It can offset the additional mixing (i.e., flushing) cost. .ACTION TAKEN OR PROPOSED The status of flushed transoxide yellow (682) dis persion has been summarized in a separate writing (I. C. Chu to T. A. Ashe, 9/3/71) with copies sent to all concerned and a recommendation has been made for preparing the flushed dispersion in semiwork quantity for further evaluation and implementation in the organosol lacquers. DUP030013888  Pigments Department will be contacted regarding the know-how for manufacture of F-llJ-D (hydrous iron oxide) which, we found, consisted of very fine fundamental particle sizes in sub-light scattering range but unsuitable for use in acrylic finishes due to incompatible additives. An added goal of our program is the implementation of flushing technology in the refinlshes area. PATENT STATUS Based on limited prior art search, we believe that the "mixing/centrifugation" flushing process is unique and unusual in terms of simultaneous removal of agglomerates and residual water; no heat nor vacuum being used in the removal of residual water; pigment deflocculation, high fluidity and low density being required of the flushed dispersion. A patent proposal will be made on the 11 mixing/ centrifugation" flushing process. PUBLICATION STATUS Our studies on (a) pigment surface energies and their correlation with flushability and (b) HLB (hydrophile lipophile balance) requirements for flushing effectiveness probably contain publishable results. Additional data and examples are desired in both cases. Publication will be contemplated at an appropriate time. ACKNOWLEDGEMENTS Acknowledgements are due to K. J. Brzozowski, H. Jakubauskas, L. R. Harper, S. Wu, V. Kasmer, J. G. King (Flint) , W. Ross (Pigments) and ;J. Jackson (Pigments) for their helpful discussions and contributions. Much of the laboratory work was skillfully carried out by M. Clark. DUP030013889 DISCUSSION I. GENERAL STUDIES OF FLUSHING TECHNOLOGY Most synthetic pigments are prepared by a pre cipitation procedure. The precipitated pigments are washed free of acids and other water soluble impurities by pressure filtration. The filter cakes (frequently referred to as presscakes) of various pigments still contain 4o-8o# resi dual water which has to be removed by other means. The most commonly employed method has been oven drying followed by mechanical pulverizing. The products thus obtained are in the form of a dry powder. The dry pigments normally exist in a highly aggregated state, and considerable amount of mechanical work will be required to break-up these aggregates and redisperse them in appropriate liquids and resins. For many pigments--particularly those with a "hard texture", it would be difficult to reduce them to the original degree of fineness (as found in the precipitated state) by mechanical grinding. "Flushing" offers an alternative route to manu facture of pigment dispersions. In "flushing", the presscakes are transferred directly from aqueous into the organic liquids or resins. The incentives of "flushing" are (a) it circumvents the processing steps of "drying" and "pulverizing" and (b) since the pigment particles in the presscakes are in a less aggregated or loosely agglomerated state, it requires no grind ing or less grinding to achieve the same result of particle size reduction. A key factor in "flushing" is the relative wetability of the pigment particles by water and oil (i.e., organic liquids or resins). In order to affect "flushing", the pig ment particle must be preferentially wetted by the oil phase and water must be effectively displaced from the pigment sur face. This factor has been studied in terms of the polar and non-polar surface energies of the presscake. -6- DUP030013890 A. FLUSHIBILITY OF PRESSCAKES (1) The total surface energy of solids can be resolved into two components--the polar and non-polar contributions. A procedure for estimating the magnitude of the polar and non-polar components for polymeric materials has been worked out by S. Wu. (Ref. 1). K. J. Brzozowski, (Ref. 2) applied the technique to determine the surface energies of pigments and found there was a good correlation between the pigment surface polarity and deflocculation by AB dispersant (H5581). We employed the same technique to characterize the surface energies of presscakes. The procedure involves oven drying the wet presscake, pulverizing the dried presscakes with mortar and pestle, making compressed pellet and measur ing its contact angles with water and methylene iodide. From the contact angle data, the polar and non-polar contributions to the total surface energy can be computed with the aid of a computer program (Ref. 5). Some of the experimental difficulties encountered during the contact angle studies are worth mentioning: Presscakes and pigments with a hard texture were difficult to compress into a coherent pellet. The resultant pellet was very brittle with little co herent strength. Leaving the pellet in the press longer and applying higher pressure did not seem to help the situation. It was found that screening out the coarse fractions and using only particles that could pass through a fine sieve (<200 mesh) would increase the coherent strength sufficiently so that an intact surface large enough for contact angle measurement might be obtained. For a Monastral red (W8l6) pigment, the compressed pellets prepared a few weeks apart gave significantly different contact angles with H2O and methylene iodide. The cause of this discrepancy has not yet been identified. ~7- DUP03C  Duplicate (or triplicate) pellets made on the same date by standardized procedure usually gave con sistent and reproducible contact angle results. I Taking the precautions as stated above the sur face energies of various presscakes were determined. The results are summarized in Table I. The flushabillty of a Klven presscake into an organic phase (i.e., toluene) is related to the $ surface nonpolarity of the presscake (with the except of W558 of which the contact angles with HaO and methylene iodide were suspected to be misrepresented due to artifact introduced at the pellet preparation). Presscakes with surface nonpolarity greater than 70$ will most likely be flushed into the organic phase spontaneously, while those with surface nonpolarity less than 60$ will most likely re main in the water phase. (2) In another series of experiments, we considered the flushability of presscakes by the thermodynamic argu ment. The free energy change (<G G) accompanying the trans fer of a pigment particle from aqueous into an organic phase is given by 4G A }Tso sw (1) Where A * Surface area of the particle. y so Interfacial tension between pigment surface and the organic liquid. Jt" 8Vl 35 Inter facial tension between pigment surface and water. The interfacial tension; of a small particle with liquids is difficult to determine experimentally. We choose to use a harmonic-mean equation developed by S. Wu (Kef. 4) to calculate the interfacial tension from the surface tension properties of the individual phases (i.e., pigment surface and liquid). -8- DUP030013892 YX2t 3 /"i + /a . Tj_Zl /}*/ /f+/l l {S) Where/xs * Interfacial tension between phases 1 and 2. =* Surface tension of phase 1 ^ a * Surface tension of phase 2 Superscript d Dispersion (or nonpolar) component of the surface tension Superscript p * Polar component of the surface tension Equation (2) has been shown (Ref* 4) to give better fit to experimental values than Foukes' equation (Ref. 5) for liquid/liquid and liquid/polymer systems where polar materials are involved. In order to apply equation (2), we need to know the polar and nonpolar components of the surface tension of water, organic liquid, and the pfesscake. The surface tension data for water and most organic liquids are available in the open literature and those for the presscakes can be determined in a manner described in the previous section. The flushability data for a variety of pigments are given in Table II. These results show: The calculated free energy changes predict reasonably well the flushability of various pigments. Such correlation also indicates that equation (2) and the contact angle/pigment pellet technique have useful application to the practical flushing problems. For some pigments (i.e., W293, W675, W8^9, W537 and W552), although there was clear evidence that "flushing'' had taken place, noticeable amount of pigments still remained in water. Flushing efficiency was generally poor. -9- DUP03001 3893  Optimum flushing efficiency was observed when the free energy change had a larger negative value (-28 to -45 evgs/cm2). e A good correlation existed between surface non polarity and pigment flushability. Optimum flushing efficiency was noticed when surface nonpolarity exceeds 85$. B* s e l e c t io n o f f l u s h in g Aid s 1. SURFACE CHARGE OF PRESSCAKES For presscakes that do not flush into the organic phase spontaneously or have a poor flushing efficiency, it would be necessary to use a flushing aid (or transfer agent). Flushing aids are materials that adsorb on the pigment sur face and alter the surface characteristics so that flushing (or flushing efficiency) can be promoted. Adsorption can be divided into two types--chemi sorption or physical adsorption. Chemisorption is very specific to the species involved and will be more difficult to predict on a generalized basis. Examples are Oleaterutile (TiOs) and Oleate-haemotite (Fea03) system (Ref. 6). Physical adsorption is due to Van der Waals interaction, hydrogen bonding and electrostatic attraction. Many known flushing aids are surface active materials of anionic or cationic types. Their physical adsorption behavior can be predicted from the ionic type, and the hydrophilic and hydrophobic portions of the molecule. We have found that presscakes which required flush ing aids consisted of particles either positively or negatively charged and they generally responded to ionic transfer agents of the opposite type. The results are shown below: - 10 - DUP030013894 Presscake Tested Surface Charge Type of Flushing Aid Required Transoxide yellow (EXP920-65) Positive Anionic (dialkyl phosphate) Lab. synthesized hydrous iron oxide (567E-28A Positive Anionic Monastral Blue (BT-463P) Negative Cationic (Tri alkyl quaternary ammonium salt) Monastral Blue (BT413-P) Negative Cationic Bon Red Dark (RT-695-P) Negative Cationic The surface potential of transoxide yellow pigment (W682) varies with the pH value of the aqueous medium. A Zeta potential* v.s. pH plot is shown in Figure 1. The sur face charge was found to change from positive to negative at ^pH=10.3. In order to further test the selection rule illus trated in the preceeding section, we brought the pH value of the transoxide yellow (W682) presscake to 10.5 where, in accordance with Figure 1, the surface charge should be reversed. As would be predicted, we found that under such circumstances, it required a cationic transfer agent to effect the flushing. These results also indicate that in the flushing systems (i.e., water/oil mixtures), the ionic attraction is a major factor for promoting adsorption onto the pigment surface. The Zeta potential of a solid particle is the surface charge on the plane which divides the rigidly adsorbed charge layer from the diffuse layer. - 11 - DUP030013895 2. HLB (HYDROPHILE - LIPOPHILE BALANCE) OF FLUSHING AIDS While the Zeta potential (or the sign of the surface charge) of the presscakes at the conditions of flushing will guide us to select the proper ionic type for the transfer agent, we noticed an additional requirement. This has something to do with the hydrophile-lipophile balance (HLB) of the transfer agent. Water soluble surfactants were shown to be poor transfer agents (i.e., for flushing). The following results illustrate this point: EFFECTIVENESS OF FLUSHING AGENTS FOR TRANSOXIDE YELLOW PRESSCAKE (W682) AT PH 6.8 Ionic Type Water Soluble Effective as Flushing Aid Bis(2EH) hydrogen phosphate Anionic ! No Yes Stearic Acid Anionic No Yes Tri Octyl methyl ammonium chloride Cationic No No Alkanol BG Anionic Yes No Compound 88 Anionic Yes No Sodium Lauryl sulfate Anionic Yes No Transoxide yellow presscake (W682) at PH=6.8 is positively charged and has a Zeta potential of +44. It therefore requires an anionic flushing agent. As anticipated, the cationic material (tri octylmethyl ammonium chloride) failed to act as flushing agent. Among the various anionic materials tested, however, only those with water insolubility were effective flushing agents; water soluble agents had either poor efficiency or no flushing effect at all. - 12 - DUP030013896 HLB puts the water and oil loving properties of amphipathic molecules on a numerical scale. It may serve as a useful criterion for general selection of transfer agents and a basis for simplification of experimental observations. In the following experiment, the flushing efficiencies of a series of fatty acids for the transoxide yellow presscake were determined (The fatty acids are known (Ref. 7) to form chemisorption on iron oxide surface). A range of HLB numbers was included in this study by varying the size of the hydrqpholic portion of the fatty acid molecules. The effect of HLB on flushing efficiency is Illustrated below: FLUSHING OF TRANSOXIDE YELLOW fW682) PRESSCAKE INTO TOLUENE USING FATTY ACIDS AS FLUSHING AGENT Flushing Agent HLB*1 Flushing Efficiency*2 Propionic acid 7.7 1 n-Valeric acid n-Caprylic acid Decanoic acid Laurie acid Stearic acid 6.7 5.3 4.4 5.4 1.5 2 4 5 4-5 4-5 *1 - HLB values were computed by the Davies' equation (Ref. HLB (hydrophilic group number) - n(CH2 group number) -f 7 -COOH (Group number) 2.1 -CH2 (Group number) * 0.475 *2 - Flushing efficiency was rated by the amount of residual pigment in the water phase 1 ------------- > 5} No flushing ---------- > complete transfer. - 15 - DUP030013897 The flushing efficiency is seen to increase with the size of the hydrophobic portion of the fatty acids up to 10 and 12 carbon atoms and begin to level off thereon. HLB values for optimum flushing efficiency Appear to be in the range of 1.5 to 5.3. In comparison, the optimum range of HLB values is 3.5-6.0 for water/oil emulsification; 7.0-9*0 for wetting; 13.0-15.0 for detergent uses and 15.0-18.0 for colloid sta bilization. Other implications from the HLB/flushing efficiency studies are: Stearic and lauric acids are slightly less effective than decanoic acid. This indicates that the hydrophile and lipophile balance rather than the size of the hydrophobic portion is the dominating factor. Along this line of reasoning, we found that succinic acid end--capped AB dispersant (H558I) did not func tion as a transfer agent. Presumably the 5000 MW acrylic B-segment of H5581 outweighed the succinic acid A segment and the resultant HLB value for H5581 was lower than the optimum range (1.5-5*3). e HLB values for ionic surfactants are not well docu mented because of the dependence of degree of ioniza tion on environment. But the ionizable groups are generally assigned with very high hydrophilic values. A few examples from the published data (Ref. 7) are: Ionic Group -SO4 Na+ -COO Na+ Hydrophilic Number 38.7 19.1 It is seen that one sodium sulfate group can balance out the hydrophobic contribution from Cep hydrocarbon chain (MW 1100). It is therefore feasible that low molecular weight dispersant of the A-B type can be tailored to function also as a flushing aid. - 14 - DUP03001 3898 A single experiment was conducted to test the above idea. An A-B dispersant (5000 MW) having a quaternary ammonium A group (546E-136) was prepared through the courtesy of L. Harper. We found dispersant 546E-136 functioned effectively as a flushing agent. (Simi lar result would be anticipated if the succinic acid capped dispersant H558I is neutralized). Unfortunately, the dispersant efficiency of 546E-136 was inadequate. It did not serve the dual function. More study will be needed to determine whether this problem can be circumvented and a new series of molecules with a unique combination of properties can be created, C. AGGLOMERATES AND FUNDAMENTAL PARTICLE SIZE IN PRESSCAKES 1. AGGLOMERATES The degree of agglomeration and ease of breaking up varied from presscake to presscake. Monastral blue (BT463P) and green (GT-751-P) presscakes contained agglomerates that required moderate mechanical shear (such as Cowles Dissolver and W & P mixer with high viscosity conditions) to break them up. Optical micro graphics of presscake BT-636-P (W552) subject to various mixing procedures are shown in Figure 2. Agitation in a simple pro peller type mixing container obviously did not generate enough shear to effectively break up the agglomerates. This type of information must be on hand in order to select the appro priate process and equipment for flushing. On the other hand, an auric brown transparent iron oxide presscake (F-113-P) was flushed over into the organic phase (toluene) with minimum mechanical shear and gave a dis persion containing little agglomerates. More discussions about this presscake (F-113-P) will be found in a later section. - 15 - DUP030013899 2. FUNDAMENTAL PARTICLE SIZE The fundamental particle size was shown to have a prevailing effect on film transparency and two-tone chara cteristics apart from the overall pigment flocculation appearance in the final film. The details of this effect in relation to various pigments, lacquers and drawdowns will be discussed in their respective sections. II. FLUSHED DISPERSIONS FOR ORGANOSOL LACQUERS A. FLUSHED TRANSPARENT IRON OXIDES A survey conducted in the Initial phase of our work (Ref. 8) revealed that considerable interest existed in obtaining an alternative method for making transparent Iron oxide dispersions. Current transoxide dispersions are either made by the costly two-roll milling process or pur chased from outside dispersion houses at a premium price. Furthermore, these dispersions are often found unstable in the vehicles and lead to gellatipn and flocculation problems. In view of the stated needs, we devoted a signifi cant portion of our work to the flushing technology of the transparent iron oxide type of pigments. Our studies will be discussed in detail In the following sections. 1. TRANSPARENT IRON OXIDE YELLOW (682) (a) CHARACTERIZATION OF PRESSCAKE The presscake (Code EXP. 920-65), which was a precursor for W682, was obtained from Hilton-Davis. As received, the presscake contained 62-67^ water. The press cake was further characterized in terms of various proper ties shown in the following table. - 16 - DUP03001 3900 CHARACTERIZATION OF PRESSCAKE (TRANSOXIDE YELLOW - 682) Fundamental Particle Size O.05-0.2|i Agglomerates (or coarse fraction) 10-20$ (on wt. of dry pigment) Surface Charge (at PH=6.8) Positive Zeta Potential (at PR*6.8) +44 MV Flushability Would not flush into organic media sponta neously Flushing Agent Required HaO - insoluble amionic type (Bls(2-ethyl hexyl hydrogen phosphate) The fundamental particle size was determined by electron microscopy (Figure 3). Particles in the 0.05-0.2p. range caused significant light scattering. This was evi denced by the slight haze of the wet dispersion and its milky appearance when viewed from a low angle. The agglomerates (or coarse fractions) in the presscake could not be effectively broken up by a low shear propeller type of mixing. But they could be separated from the fine fractions by centrifugation (i.e., clinical centri fuge commonly used in the laboratory). These agglomerates, if present, would significantly degrade the dispersion qua lity. Therefore, they must either be de-agglomerated or be removed from the flushed dispersion. Surface charge, Zeta potential and the flushability of the presscake have been discussed in Part I. Bis(2-ethyl hexyl) hydrogen phosphate was selected as the working flushing agent because it had excellent flushing efficiency and insig nificant effect on the organosol film appearance. (b) PROCESS CONSIDERATION Conventional flushing is usually carried out in a heavy duty mixer (such as W & P mixer) where the presscake - 17 - : DUP030013901 and the polymeric resin form the hjighly viscous mass. The aqueous phase separated from the viscous mass is mechanically decanted followed by vacuum distillation or sheeting on a two-roll mill to remove the trapped residual water. The thick mass resulted from the above operation requires additional mixing or grinding to reduce its consistency to a manageable level. The process is costly because of the multi-step operation and not suitable for handling the fluid A-B dis persions used in the organosol lacquers. The ''FB" (flushed base) flushing process at Parlin Plant is simpler and more economical. The equipment consists of a large mixing tank (type used for "49" process) equipped with propeller type agitators, heat jacket, vacuum, condenser and water trap. Flushing is carried out in the mixing tank under fluid and low shear conditions. When flushing is completed, the mixture is let stand until the water separates into a distinct phase which is than removed by siphoning. The residual H2O trapped in the organic phase has to be distilled off. Since the process consists of no high shear operation, a major pre-requisite for applying this process is that the presscake must be reasonably free of agglomerates and unaccept able coarse size fractions. (The presscakes (F-4p) used in the Parlin flushed hydrous iron oxide dispersion for the Duco products was provided fresh by Pigments Department. F-4p has since been discontinued due to sub-standard particle size and shape required by the current glamorous finishes). The transoxide yellow (W682) presscake, as shown by the characteriza tion studies, contains significant amount of agglomerates and, therefore, can not be used directly in the "FB" process without introducing additional size reduction operations. The "FB" process, however, has the merit of simplicity and economy, it will receive due consideration whenever it is applicable. The flushing process developed for transoxide yellow (W682) presscake is illustrated in the following schematic diagram Flushed Dispersion H2O Stream Rotor Protection Cage Coarse Agglomerates deposited on wall of Rotor DUP030013902 In the laboratory, the flushing was carried out in a wide mouth jar with an air drive stirrer. Presscake, water, solvent (toluene) and transfer agent (Bis (2-EH) hydrogen phosphate) were charged into the container in proper amounts. The mixture was subject to mild agitation with an air stirrer for 20-30 minutes. By then,the aqueous phase should be well separated from the organlc/plgment phase and decanted manually. Following the decantation of the bulk water; an A-B dispersant (H5581) was added to deflocculate and stabilize the flushedover pigment particles. The flushed dispersion of transoxide yellow was then centrifuged in a standard clinical centrifuge (A.H. Thomas Co) designed for laboratory use. Residual H2O and coarse agglomerates were simultaneously removed and a fluid and stable dispersion of W682 was obtained. A typical formulation of the flushed W682 dispersion is given in Table III. The laboratory procedure was further demonstrated with a continuous centrifuge (''Sharpies" model ) as shown in the sketch. The mixture from the mixing tank was gravi tation-fed into the ''Sharpies" centrifuge which rotated at a pre-set speed. Two streams of liquids came out of the centrifuge continously--one was primarily water, and another the flushed dispersion. The advantages of the "mlxing/centrifugatlon'' flushing process are: It is simple and economical (see next section) It removes residual water and coarse agglomerates in a single efficient operation. It thus elllminates the need of vacuum distillation and grinding. In comparison to the "FB" process, there won't be any foaming problem which could occur during the vacuum distillation (in the "FB" process). Potentially it can fractionate the particle size and therefore regulate the dispersion quality. - 19 - DUP03001 3903 The requirements for applying the "mixing/centri~ fugation" flushing technology are: The majority of pigment particles must be small and deflocculated. The dispersion should be reasonably fluid. C. ECONOMICS A preliminary study of the manufacturing costs of "FB and "mixing/centrifugation" flushing processes was made by J. P. Herring, Engineering Department, at my request. The results were summarized in a letter (I. C* Chue to P. N. Jones, April 21, 197I5 see Appendix I). For comparison, the mill base manufacturing costs by the various existing processes are included in the following table. MILL BASE MANUFACTURING COSTS BY VARIOUS PROCESSES Process Cost f$/lb. of Pigment Processed) *(1) "47" Process At 12$?! Pigment Cone O.63 *(2) Two-Roll Mill (a) W & P Cost (b) Two-Roll Milling Cost (c) "49" Process *(3) "PB" Flushing Process At 23# Pigment 1.32 (0.40) (0.55) (0.37) 0.52 (*) "Mixing/Centrifugation" Flushing Process O.36-O - 20 - DUP03001 3904 MILL BASE MANUFACTURING COSTS BY VARIOUS PROCESSES fCONT'D.) Process Cost ($/lb. of Pigment Processed) *(5) Conventional Flushing Process (a) W & P Cost (at 30# Pigment) (b) "49" Process 0,77 (0.40) (0.37) It is to be noted that the divisional figures were averaged over all mill bases (prepared by the same process; for example "47" process) that went into various end products. The actual cost for manufacturing a particular high quality dispersion from a given pigment can be significantly higher than the average figure depending on the pigment concentra tion in the mill base and the number of passes through the mill etc. Recognizing the possible variation from individual to individual cases, we can still obtain an overall picture of the cost associated with each process. It is seen that the flushing processes (Cases 3. 4 & 5) on the average compare favorably against the conventional processes (Cases 1 & 2). Among the flushing processes, the "mixing/centrifugation" process has an estimated economical edge over the other two. In our opinion, there should also be some cost savings at the pigment manufacture end. However, Pigments Department Indicated that the cost of presscake (on dry basis) would be essentially the same to the dry bagged pigment (Ref. 9). This means that the economics of "flushing" will be justified solely on savings of the cost of mill base manu facturing and/or technical advantages. We will be reviewing our program closely in light of these two factors (see later sections). Divisional average cost figures (as of 1970) compiled by E. T. Brewer. - 21 - DUP030013905 D. EVALUATION (V.S. Two-Roll Milled Control) A mill base is an intermediate product. Its usefulness has to be judged in terms of its stability in the end products (i.e., paints) and its performance In the final films as well as those of the mill base itself. An Indepth evaluation of the flushed transoxide yellow) (W682) dispersion (Coded 538E-II5) was conducted In comparison to a two-roll milled control. The results are given in Table IV from which the following summary regarding the implementation of flushed transoxides In the organosol lacquers may be made. The flushed W682 dispersion (538E-115) contained coarser particle size (O.050.20p). The dispersion quality of the flushed mill base was therefore not as good as the two-roll milled control (in terms of transparency and cleanliness appearance). However, the flushed dispersion showed better stability in the organosol lacquers than the tworoll milled control. In the final films, both dispersions were flocculated: In the solid color films, the flushed dis persion flocculated to a much lesser degree; in the metallic films, the degrees of floccu lation, as exhibited by the electronmicrographies (Figures 5A and 5B), were equiva lent but the two-roll milled control gave slightly better two-tone characteristics. The organosol lacquer prepared from tworoll milled W682 (912-G-60886) dispersion showed aging effect, while that from the flushed dispersion (538E-II5) was stable over the same aging period (1 week). For glamorous metallic finishes, a strong side tone is a major quality-determining factor. In this regard, the organosol - 22 - DUP03001 3906 lacquer based on the flushed dispersion (i.e., W682) is nearly! equivalent to that of the two-roll milled control. This--in combination with lower manu facturing cost (estimated in the pre vious section), better stability in lacquers and quality improvement poten tially achievable with the use of fresher presscakes--has led us to urge that the flushed transoxide yellow dispersion be seriously considered for implementation in the organosol lacquer (Ref. 10). Other findings and implications worth discussing from Table IV include; The dispersion quality ini the final film stage is far inferior to that of the mill base. This applies to both the flushed dispersion and the two-roll milled control--the later being deteriorated to a greater degree. Close examination of the flocculation pattern in the final film (Figures 4A and 4B) reveals that it resem bles very much the phase structure of the organosol film^-the discrete phase being smaller near the sur face and growing larger near the bottom. The pigment particles obviously are preferentially' associated with the discrete phase and take on a flocculation pattern similar to that of the phase structure. The sizes of the discrete phase in the film derived from the two-roll milled control dispersion (Fig. 4A) are significantly larger than those from the flushed dis persion. Since the two panels were prepared under otherwise identical conditions, we deduced that the large phase structure was caused by factors existing in the two-roll milled dispersion but absent in the flushed dispersion. At least two factors appeared to have some bearing on this problem (a) the disper sant (RCH-14639) molecular weight in the two-roll - 23 DUP03001 3907 milled dispersion was shown by GPC to be significantly lower (~ one half) than the molecular weight of the original dispersant and (b) the additives present in the pigment. Molecular weight decrease is inherent in the two-roll milling process; whereas contamination inherited from the pigment. The flushed dispersion avoided the former factor but not the latter and con sequently an improvement in degree of deflocculation quality was observed. In another experiment (Section II A(2)), the fine fraction of laboratory synthesized hydrous iron oxide pigment was used to avoid additives in the commercial pigment and a film of exceptional transparency was obtained. Of course, the sizes of the discrete phase in an organo sol film can also be influenced by formulation and application conditions. And in this regard, it may indirectly alleviate the dispersion problem in films in-so-far as the flocculation is phase-structure associated. But the basic solution in our opinion lies In the removal of adverse causes in the mill bases, as this would decrease the complexity of paint formu lation and broaden the formulation latitude. We would like to elucidate the effect of fundamental particle size on film appearance apart from the effect of pigment flocculation. This attempt has never been made in the past in our interpretation of film appear ance and its relationship to pigment distribution be cause the two are closely related and difficult to separate. Work on the iron oxide pigment distribution in organo sol films stimulated our thinking as to the individual contributions of the fundamental particle size and pig ment floccules to the overall film appearance. Electron photomicrographies shown in Figures 5A, 5B and 5C are the cross sections of metallic organosol films pigmented with a two-roll milled dispersion (Fig. 5A), a flushed dis persion (538E-115) from a commercial presscake (Fig. 5B) and a flushed dispersion of the fine size fraction of a laboratory synthesized iron oxide pigment (Fig. 5C). The fundamental particle sizes of the three dispersions range In the following order: - 24 - DUP030013908 Flushed dispersion--commercial presscakes (-v0.050.20u,) two-roll milled dispersion (~0.G1-0.05m.) > flushed dispersion - lab. synthesized pigment ( 0.01-0.02 or less) the darkness of the side tone (i.e., the glamorous effect) of the metallic films arrange in exactly the reverse order. On the other hand, the pigment distributions (or the average size of the floccules) in the three films (judged by the electro photomicrographies) failed to show any correctable trend--Figure 5A & 5B being equivalent and Figure 5C somewhat worse. This indicates that the fundamental particle Size has an independent contribution to the film appearance apart from the pigment flocculation--particularly to the two-tone characteristics. As long as the degree of floceula-* tion is moderate as in the cases represented by Figures 5 (A, B, & C), the fundamental particle size has a dominant effect on the two-tone characteristics which is believed to result from better transparency. The following is a plot of light scattering coefficient as function of particle size for iron oxide pigment constructed from published data iBef. Ill, FIGURE 11 DU P030013909 Where Qs * Relative effective cross section (of par ticle) for total scattering. }f * 77 d Y * * * /Nb D = Diameter of Particle A* = Wave length of light in the binder in which the particle is imbedded A * Wave length of light Nb Refractive index of the binder If we assume some reasonable values for A and Kb say A 0.471j j l and Nb 1.5, then /'will be equal to 10D (A 10D). This will enable us to relate the scatter ing coefficient directly to the particle size. It is seen that the maximum scattering for iron oxide pigment occurs between 0.15 to 0.20p. (10-4 cm) and only par ticles in the size range of O.Olp, - 0.04p. have neglig ible scattering. The results of our fundamental particle size studies are in good agreement with that anticipated from the theorectical treatment. And it explains why the funda mental particle size is a dominant factor for film trans parency and two-tone characteristics. The above conclusion points to a new direction for pig ment dispersion research and new opportunities in the field of glamorous finishes through control and regula tion of pigment fundamental particle size in combination with the dispersant technology. 2. SYNTHESIZED HYDROUS IRON OXIDE PIGMENT The color of hydrous iron oxide presscak.es has been known to change with time (Ref. 12). This is attributed to the growth of particle size in the water wet state. After the pig ments are properly dispersed in the organic phase, the color will be more stable. To avoid color change and particle growth, it is therefore desirable to flush fresh presscakes. - 26 DUP030013910 Attempts -were made to synthesize hydrous Iron oxide pigments in the laboratory, and flush them over Into the organic phase from freshly precipitated filter cakes. We repeated the art of hydrous iron oxide syntheses revealed in three U. S. patents: U.S. 2,335*760, E. I. du Pont de Nemours & Co., 7/31/1941 U.S. 3,398*115* Chemetron i Company, 8/20/1968 U.S. 3*565*656, Allied Chem. Company, 2/23/1971 The synthesized hydrous iron oxide pigments in all three cases showed broad range of size distribution and much coarser than what were claimed in the respective patents. Upon fractiona tion by centrifugation, the products based on U.S. 3*398,113 and U.S. 3,565,656 gave approximately 5-10# fines (estimated); and the product based on U.S. 2,335*760 contained only neglig ible amount of fines. The following are the results of our evaluation of the synthesized hydrous iron oxide pigment, 567E-28 & 34 (U.S. 3,398*113). Pigment surface carries a positive charge and a Zeta potential of ~25 MV. The flushing aid (Bis (2EH)hydrogen phosphate) and flushing process ("mixing/centrifugation" procedure) employed for the transoxide yellow presscake (precursor to W682) are also applicable to the laboratory syn thesized pigment (567E-28 and -34). The flushed dispersion derived from the fine fraction of 567E-34 exhibited a high degree of transparency and cleanliness without the milkiness appearance which was shown to be characteristic of the presence of high scattering particles. Figure 9 is an electron photo micrograph of Sample 567E-34. The loose fluffy agglo merates were formed during sample preparation where a drop of the diluted dispersion was laid on the carbon grit and dried. The single particles which formed the agglomerates were easily distinguishable and esti mated to be 0.02p or less. - 27 - DUP030013911 * The flashed dispersion is stable in the organosol lacquer. A single pigment tinting lacquer formulated from 567E-34 gave drawdowns (567E-31-3) of exceptional transparency. Electron photomicrographies of the cross section of the film (567E-31-3) are shown in Figures 10A and 10B. Again, as shown in the earlier films (Figs. 4A and 4-B), the pigmented region was smaller near the surface and grew larger towards the bottom. As already discussed in previous sections, this pattern resembles the phase structure of the organosol film and the pigment particles are associated with the discrete phase regions. The metallic organosol lacquer derived from 567E-34 has already been discussed earlier. The superior twotone characteristics (Figqre 12) is attributed to the sub-scattering fundamental particle size. The above work furnishes a lead to relatively troublefree iron oxide dispersions and superior glamorous (metallic) finishes provided that we have access to presscakes of sub scattering particle sizes or the know-how to make them effec tively. Such a possibility may come from Pigments Department's know-how on F-115-D--an auric brown transparent iron oxide pigment. (See next section). 3. AURIC BROWN TRANSPARENT IRON OXIDE (W-461) We have evaluated the presscake (F-113-P) for this pigment and found that it contained particles primarily in the sub-scattering region. The presscake could easily be flushed over into the organic phase (i.e., toluene) by the procedure described earlier and gave a dispersion of excellent appearance (i.e., transparency and cleanliness). An electron photomicro graph for the flushed dispersion is shown in Figure 13. Dis regarding the few coarser particles which form the foreground, the majority of the pigment particles forming the background of this electron photomicrograph are In the sub-scattering^ region barely resolvable at 38QQoix magnification. - 28 - DUP03001 3912 The major drawback of F-113-D is that it cannot be used in acrylic vehicles. We have mixed the dispersion with (a) organosol, (b) current acrylic lacquer (RC-909), (c) acrylic refinish lacquer (RC-3176), and (d) Centari refinish enamel (RC-3346). All gave poor results. The incompatibility of F-113-D with the agove vehicle lines is believed to be due to either the petroleum sulfonates or the amine compounds or both which have been added to the pigment during its manufacture (Ref. 13). However, if we could intercept the pigment at the precipitated state before the incompatible additives were introduced and applied materials compatible with the vehicle lines for which it is intended, this could provide a useful flushed dispersion with superb particle sizes for high quality finishes. We intend to touch base with pigments and pursue further along this line. B. FLUSHED ORGANIC PIGMENTS - Monastral Blue & Green 1. GENERAL CHARACTERIZATION The flushing and dispersion problems of organic pigments are quite different from those of iron oxides. Some of the common characteristics of Monastral (copper phthalocyanlne) blue and green presscafees are described below: Fundamental Particle Size - The fundamental particle sizes of a Monastral bide (W550) and green (W765) were determined by electron microscopy. W765 ranges pri marily from 0.05 to 0.10U (Figure 14); W55O has slightly smaller particles mostly in the range of 0.03-0.08y, (Figure 15). Flushability - Both W550 and W765 can be flushed into organic media without the use of any flushing aid. This would be true for many other organic presscakes as the pigments are basically organophilic. Some organic pigments (such as W552 and W537), however, have been surface treated to improve their disperse ibility. Such pigments usually have insufficient flushability and flushing aids;would be required for effec tive flushing. Suitable flushing aids are water-insolu ble cationic surfactants Such as tri octylmethyl ammo nium halides. - 29 - DUP030013913  Agglomerates - Are present in the presscakes of most of the organic pigments tested. For example, W552 (BT-463P) contains about 15-20# agglomerates. These agglomerates would not be broken up in a low shear flushing operation. Vigorous agitation or mechanical grinding would be needed to reduce the agglomeration. 2. CONSIDERATION OF FLUSHING PROCESS The "mixing/centrifugation" process developed for flushing of transparent iron oxidjes unsuitable for some of the organic pigments on account of the following: Many organic pigments (i.e., W765, W550, etc.) exist in the organic media (solvents or resins) in a some what flocculated state. As indicated earlier (Section II A(l) - (b), pigment flocculation and non-fluidity interferes with centrifugal separation. Effective centrifugal separation would require that: (a) Pigment particles well deflocculated (b) Dispersion fluidity less than 50 cp (estimated) (c) Dispersion density less than water (i.e., 1.0 cp) Organic pigments are quite expensive (ranging from $4~$12/lb. In contrast to ~$l/lb. for the Iron oxides). The agglomerates should be processed (i.e., deagglomerated) and fully utilized rather than removed (or discarded). A maximum of 5# pigment loss in the form of agglomerates during centrifugation may be tolerated without jeopardizing the economics of the centrifuga tion process. Other wise it would become economically unattractive, (unless the removed agglomerates can find uses in other grinding processes). - 30 - DUP03001 3914 Our current process concept for flushing of organic pigments is illustrated schematically below: Add Resins of Flushed Dispersion Various Dispersion for Various Lines Lines CP*O Flushing in Mixing Tank flushed pigment .Resin A Resin B Resin C *> tfancT Grinding > > > The flushing is to be carried out in a low shear large mixing tank (equipped with heat jacket, condenser and vacuum) into a commonly used organic solvent (i.e., toluene). To the flushed pigment are added the resins of various dispersion lines and the resultant mixtures subject to sand grinding to break up the agglomerates which are present in the presscakes. This concept enables the flushing to be carried out in large quantity and the flushed pigment goes into various dispersion lines. The economics of the flushing of organic pigments lies in (a) equal quality resulted from less grinding or (b) better utilization of pigment or (c) both. The above process has been demonstrated successfully with the presscake of W765 in laboratory scale operation. Pre liminary evaluation of the flushed W765 as a 508-line refinishes dispersion indicated that better dispersion quality could be obtained with equal amount of sand grinding or equal quality achieved with significantly less grinding. Most conventional sand ground dispersions of organic pigments used in the auto motive and refinishes required twp passes through the sand mill (as is the case for W765). Therefore, the key factor in comparing the flushing route v.s. i conventional dispersion - 31 - DUP030013915 process is whether the combined savings from better grindability and/or improved pigment utility can justify the additional cost of flushing. This situation, which is characteristic of organic pigments, is now under evaluation in the refinishes as well as the automotive areas. MacE 10/20/71 - 32 - DUP030013916 -33REFERENCES ( 1) S. Wu - ESR-70-21 (2) K. J. Brzozowskl - EX-70-22 ( 3) SURE PROGRAM - D. BarrIs, Engineering Department ( 4) S. Wu - Memoto Pile 211014, May 5, 1971 ( 5) F. M. Fowkes < - Ind. & Ind. Chem* Vol. 5& (12) 40-52 (1964) (6) D. W. Fuerstenau - Pure & Applied Chem., Vol. 24, 135-164, 1970 (7) A. S. Peck, L. H. Raby and M. E. Wadsworth, Trans. AIME, ]J2 301 (I966) ( 8) Communication with E. D. Brewer ( 9) Letter E. D. Brewer to T, A. Ashe, August 6, 1971 (10) Letter I. C. Chu to T. A. Ashe, September 3, 1971 (11) A. Brocks, Optiks 24, 550-566 (1964) (12) W. H. Edwards, J-l4, 10/15/1954 DUP03001 391 7 >P aCJj rt 4> rHH i--3I .<O6 EOh C CO O Hf3e +Cr3t CO co CO CO 4) O O 4) (U <1) S3 a >H <v -Jj 2 'u*3 <5-4 5 coo 2^ ^o S3 to o is* is s vo * CO in o\ ON CM tn CO CM -=t- tn in VO s H CO H fd p %C4 o o g Eh CO tn VO Cf rH k N K\ CM ts .=4 o LTV iH in JP i &< u t2 0$ g HMui raVH/ w d< s CM =4 OJ o\ H VO * +! o n o O o n rH KY CM CM rH rH O rH s H 4O0a-1J) J<4Od8) i -=4 CO I w a Eh Eh CO C3O w<U fi rH * rH CM CM CM * 4 . CM : IS tn m CT\ o <M CJ K\ tn tn sJ h VO tn * O Ch VO r) 4 CM (C K\ (T\ KN LH CM: rKHO CO rH rH o rH in VO VO VO is S K\ IS CO CO 0? *oo o CM CO o ON H CO LTi o KV VO VO1 tn1 m CO 1 1* t a Isr J* a K\ LTV S VO !S tS: I , ts a> *a O rH rH o f--Io <U S rH ITs0) M 0J VO >aH) O Idf v ca O COco 0) idfH a. Eh H O O Q O <u a Q X CM i i t Ca CO IS in IS ( H m rH CM On co m N LA cti Jh E-i 34 vo CO rH ctf 1 i I rH h0 id EH EH ( SH ma M ts EH a IC C /a y k 10/18/71 DUP030013918 #These pigm ent s u rfa c e te n s io n v a lu e s a re ta k e n fro m R eport EX-70-22 (K. J. B rz o z o w s k l). ICC/ayk 10/18/T1 O W W H C tH O rl 4-5 03 a> <d P$ rl Pi o 01 p rH <u H CD H CD i--i CD rH 0) rH CD rH CD rH CD rH CD H r>> P <9d .O cd 3 Q9$ X* a X> rl JQ rl ,o rl XH> rH CD CD CD CD CD bO bO bO hD fn O O O O O H H iH H O H rl rl H H rH rH rH rH T3 P -P P P P bO bO 60 bfl O O O O O CD CD CD CD sa ia t5 ss; ia ia Sa sa sa a> c CO d o<*3 xa)l xas> CO CO CQ CO CO CO CO CO CO 03 t O CD CD CD ID CD HD CD CD CD 3 525 is; rH tH tH iH iH in iH ' iH tH H H rs 0 1 bO H S 03 a rp3Hn< <3> P Vi a) O 1& Hh as 01 s *v. CO hO G W ~ " 5h a> o> in o O OJ VO VO H st S;O* o VO St rH in in <30 rH i--1 in O CO m st CM m CM -=t ^t i111i 1 i1 C H *a0 <D o ?-l cd II to d> pC P } 0) 0 0) Xi 1 A. P P. CO >< 3o rl cr\ U* KN CM O KN Vi -- W Xi CD G VS ai rH > CM P* ^ CO OQ S xs 3 C rH 8O o d >25 P c8 0 CD d 1 P xs >s CO CM t- 5 s O rH in II Pi 6 >s *l CO CM II g <D td >a II T3 P ?H 0} rH O 04 g ia d -p <sS G rH CD -- C3 O P*d O Pi O CO H W P 0 h CVoo <v a is; O EH WH | O Pi O 3 Vi Vi O U 3 CO G i--S1 P0i CD r-' OGPiPVCiO. Sw O st -=t 00 vo in 6- o oo St 00 in in CO CO in MOON CO CO CM KN st KN CO o CM N- t- KN CM st ON H in o On O CM tn m KN KN st KN -=t in rH* -O=jt OHCMNi O irHn CM rHH orrHH C*-- CM tH- VO ON st VO CM H CM d h 1CD) rl Pi O o KN ii KN KN ON ON C-- m KN vo co 11i is IS is IN K'S in 1 is CM in in vo UN E 1l is is in :o o MO in is i!2 co co ini is hO a XcSo d El Vi O G CD G w CD 0) Vi O CO >9 03 GI O rl p CoO 03 * roH a SH PM P>Ooi + Pi P>CoOi Ql X? >o + PC<O PC0h X> xo d O >o >* 'O T5 n>55o. CO d d >o + d d ra CO x> >=* =t st II >s & >* co > > II II O J >wo Vo DUP030013919 -36- TABLE IIX TYPICAL FLUSHING PROCEDURE FOR TRANSOXIDE YELLOW (W-68g) PRESSCAKE Aqueous Phase Hllton-Davis Exp 920-6$ Presscake Water 560.0 960.0 Oil Phase Toluene 680.0 Bis(2EH)Hydrogen Phosphate 17.0 (pre-dissolved)----------------------------------------------- Mix 20-30 minutes in a container with a propeller stirrer/decant H2O; approximately 12001300 g. of HgO can he removed by decantation. Add Toluene H-257 A-B dispersant 217.5 217.5 130*0 --------------- ------------- Centrifuged to remove residual Hs0 and pigment agglomerates. The recovered dispersion was deflocculated and stable on aging. The composition of the flushed dispersion (coded 538E-115) was as follows: Toluene H-257 H-5581 W-682 Bis(2EH)HP ICC/ayk 10/13/71 % by1 Weight 61,1 15.0 8.9 13.9 1.1 100.0 DUP03001 3920 -37- TABLE IV COMPARISON OP FLUSHED AND TWO-ROLL MILLED DISPERSION OP TRANSOXIDE YELLOW PIGMENT (W-682) Two-Rolled Milled Dispersion (912-G-60886) Flushed Dispersion References of (538e_115 ) Supporting Data (!) Mill Base (A ) Flocculation Tendency Deflocculated Deflocculated NB 538E-116 and NB 558E-193 ( B ) Particle Size ; Very fine, primarily in the range of 0.01-0.05m Coarser, primarily in the range of 0.05-0.20m Figure 3B and Figure 3C ( C ) Visual Appearance Transparent and clean Less transparent and milky at low angle of viewing NB 538E-II6 and NB 538E-193 ( D ) Mill Base Stability 0. K. O.K. NB 538E-195 (II) Organosol Lacquer - Solid Color (III) Organosol Lacquer - Metallic (A) Stability of Lacquer Unstable:(became flocculated on aging) Stable (no sign of flocculation) ( B ) Solid Color Drawdown Visual Appear- Visual Appear- ance ance Hazy Less hazy Film X-Section by Electron Microscopy Severely floc culated Film X-SectiOn by Electron Microscopy Flocculated (A ) Stability of Lacquer Separation of A1 Less separation flakes from color pigment NB 538E-195 NB 538E-195 Figure 4A and Figure 4B NB 538E-195 DUP03001 3921 -38- TABLE IV Page 2 COMPARISON OF FLUSHED AMD TWO-ROLL MILLED DISPERSION OP TRANSOXIDE YELLOW PIGMENT (W-682) " (Continued) (III) Organosol (Continued) Two-Rolled Milled Dispersion mmmmmmtMmmmmjmrnmijgmfmm, (912-0^0^86) Flushed Dispersion 1 '(5 ' 38E-115m)m (B ) SprayOut Panels Two-tone: Better twotone Two-tone Slightly less two-tone film X-Section by Electron Microscopy Pigment moder ately floccu lated; equiva lent overall appearance Effect of Aging Aged lacquer gave reduced two-tone 1 week) Film X-Section by Electron Microscopy Pigment modern ately floccu lated; equiva lent overall appearance Effect of Aging No significant effect on aging (*v 1 week) References of Supporting Data Panels 538E-14C and l4Dj Figure 6 (OMR data) Figure 5A and Figure 5B Panels 538E-20A and 2OB; Figure 7 and Figure 8 ICC/ayk 10/18/71 DUP030013922 ZETA POTENTIAL -39 DUP030013923 PRESS CAKES (BT4 PIGMENT AGGLOMERATES I l f oo) vo r+Xt<ao(pW~WoS>0!)) M is- X X. <S CO X. H 1o0 !Xor~,f DUP03001 3924 -4i FIGURE 3K T'RANSOXIDE YELLOW (W-682.); ER1SS0AKE: PROM .HXKC.OI DAVIS : V SHOWING BASIC' PARTICLE SIZE ...... AID SHAPE ..." I h ifeyp it *% 1 S-V ^ , , j*/ ...,,4. , ........ Mmmm........ .> ' f * it' *1 X 'Mmj ti *.*<* _S2 *"*' k * t?*x- +.', ' ^^ : 5.;.. dk^-*.J`"^ ; ... wi/t w fL-.,547'.Ac. V^, ,t$^L* - '-.. dfiLi*. j .yyw" ' # JY^Sfei.,.. jl- ;< ^ > d 4? isIi*t.:?'JL'I T' ;#,, it Jk 4 Fjt*"^ j mart*;, 1*1 .* .... >5!* , ..K4ii <,. **L-2k*AJiff4fi..*.id--8^ - J^- - s<*r * Ifap * nr1 Yw:Tk 4 I, . .... t '---^: DUP03001 3925 DU P03001 3926 DUP030013927 ~44~ FIGURB 4A (567E-51-4) CROSS SECTION OP ORGANOSOL LACQUER FILM PX&MINTED WITH TWO-ROLL MILLED w-68g d is p e r s io n SHOWING GROSS .FLOCCULATION AND ASSOCIATION OF PIGMENTS -WITH PHASE STRUCTURE JJ DUP03001 3928 iv j * z- -45- OL FILM PIG2 DISPERSION DERATE DUP03001 3929 FIGURE 5A CROSS SECTION OF METALLIC FILM BASED. ON TWO-ROLL DISPERSION (*-682} 912-0-60886 '' DUP03001 3930  PIQUEE 5B GROSS SECTION OP METALLIC FILM BASED ON BLUSHED DISPERSION (-682) 538E-115 DUP030013931 : -M-.. FIGURE 5C GROSS SECTION OF METALLIC FILM BASED ON FLUSHED DISPERSION OF LABORATORY SYNTHESIZED HYDROUS IRON OXIDE PIGMENT (567E-34 and 567E-37) t JJ DUP03001 3932 * *T 110 100 90- 8o 70i 60. 5-0 - -49- >. FIGURE' 6 LIGHTNESS OF METALLIC ORGANOSOL PANELS FROM VARIOUS ANGLES OF VIEWING (OMR DATA) 0 = 981-R-P731V538E-115 - pigmented with flushed W-682 dispersion & 98'1-R-2731;i/912-G-6o 886 - pigmented with two-roll milled W-682 dispersion L (R ightness o f R e fle c tio n ) 30 h 10 20 30 40 ~4o 60 Angle of Viewing (degree from normal direction) 70 DUP03001 3933 (Lightness Value} -50FIGURE 7 DUP030013934 1104 100 -51- FIGURE 8 LIGHTNESS VS. ANGLES OF VIEWING Effect of Aging of Organosol Metallic Lacquers Pigmented with Flushed W-682 Dispersion. 90' 80 0) 3 !--I 70- g w to <D C -p .c w 60 rt sn 50 40 A * Freshly prepared lacquer (567E-14D) @*1 week old lacquer (567E-20B) 30 20 0 10 20 30 40 50 e Angle of Viewing (from the normal position) 7b DUP03001 3935 FIGURE 9 ELECTRON:PHOTOMICROGRAPH: ,PE SAMPLE 56TE-34 SHOWING THE FUNDAMENTAL PARTICLE SIZE OF SYNTHESIZED HYDROUS IRON OXIDE PIGMENT i! i DUP030013936 -53- f ig u r e 10A CROSS SECTION OF 567E-31-3 {ORGANOSOL LACQUER PIGMENTED WITH FLUSHED DIS PERSION OF LABORATORY SYNTHESIZED HYDROUS IRON OXIDE PIGMENT) SHOWING TOP PORTION OF THE FILM. J I DUP03001 3937 *A DUP030013938 -55- FIGURE II See Page in body of report * DUP03001 3939 (Lightne ss V alue) -56FIGURE 12 LIGHTNESS OF METALLIC ORGANOSOL PANELS AS A FUNCTION OF ANGLES OF VIEWING {OMR DATA) 110 \ 100 90 80 70 60 50- HO 30 20 10 20 30 40 50 6( 70 Angle of Viewing (from the normal direction, DUP03001 3940 w0 DUP03001 3941 DUP03001 3942 FIGURE 15 PARTICLE SIZE OP BiT~383-D' (w-550} 1; DUP030013943 -60APPENDIX I Page 1 cc: W. M. Duffy, Flint J. M. Donatello, Flint R*" Vest, Flint A. F. Nugent, Flint D. M. Marsh, ESL T. A. Ashe, ESL Not Indexed Files 611331 Experimental Station April 21, 1971 F. N. JONES F & F DEPARTMENT EXPERIMENTAL STATION MANUFACTURING COSTS OF MILL BASES VEA "FLUSHING* A preliminary study of the manufacturing costs of mill bases via "flushing" was made by J. P. Herring, Engineering Dept., at my request. In accordance with the Initial leads developed in the "Pigment Flushing" program, two "flushing" processes received consideration. Case I was a low shear mixer flushing followed by decantation and vacuum distillation; Case II, a low shear mixer flushing followed by continuous centrifugation. A brief descrip tion of the basic procedures for the two cases and the estimated manufacturing costs are given in the following table. Description Manufacturing Proceduress Case I Case II (l) Charge presscake, organic liquids, transfer agent, etc. into the mixing tank. (2) Continue agitation until "flushing" is completed* (3) Stop agitation. Let stand until water foms a separate phase. (1) Same. (2) Same, (3) Pump mixture through a continuous centrifuge. Water & coarse particles are simul taneously separated from the pigment dispersion. DUP03001 3944 -61-3- APPENDIX I Page 3 The Experimental Station Special Services Lab has two continuous centrifuges of the type described. Arrangement can be made for temporary use of their equipment. A sample of flushed transoxide yellow dispersion obtained by the Case II process has been sent to Flint (J. G. King) for evaluation. Pending the results of King's evaluation, a course of action will be recommended. ICC/mcc 4/21/T1 I-Cheng Chu DUP03001 3945 Description Manufacturing Cost ($/lb., pigment): 20$ Return on Investment: Total: -62-2 Case I (4) Decant water. (5) Vacuum distill residual water. (6) Packaging. $0.20-0.35 $0.19 $0.39-0.54* APPBNDIX I Page 2 Case II (4) -- (5)----(6) Packaging $0.17-0.25 $0.19 $0.36-0.44* These manufacturing costs are to be compared with $1.30-1.60/lb. pigment by the two-roll milling procedure which is required for most high quality dispersions. It is assumed that the flushing operation can be carried out in the 1000-gal. mixing tank now available at the Parlin Plant. Hence, for Case X, no new investment will be needed. However, for Case II, new investment in the continuous centrifuge will be required. The reduced manufacturing cost for Case II is the result of labor savings due to the shorter batch cycle and reduced operator attention. ROI is assumed equal in both eases, since the new investment required in Case II Is approximately off-set by the reduced allocated investment. Case II Is particularly suitable for pigments such as iron oxides because centrifugation permits removal of the aggregates as well as residual HaO. {Presscakes of iron oxide pigments already contain significant amounts of aggregates.) The prices of small continuous centrifuges suitable for the described type of operation range from $10,000-15,000. These centrifuges would have a capacity of 5-25 gallons/mln. which would be adequate for manufacture of medium batches of mill bases. DUP030013946 63- APPENDIX II SUMMARY OF PRIOR ART ON "FLUSHING" U.S. g,140,745 - Da Pont Co., 12/20/1938 This patent relates to improved cellulose coating compositions and a process for direct transfer of pigment particles from a water suspension to a cellulose derivative vehicle. Process: W & P mixer followed by tray drying or vacuum drying until the residual H20 is less than 3$. Vehicle System: Cellulose nitrate. Transfer Agent: Presumably, dlbutyl phthalate, etc. Pigments: Chinese blue pigment pulp; maroon pigment pulp; lead chromate pulp; carbon black dry pigment, etc. U.S. 2.112,222 - InterChemlcal Company, 9/21/1933 This invention relates to a method of manufacturing pigments dispersed in oil and to the product thereof. The flushing of pigments from wet presscakes into organic media and the separation of water from the oil curd containing the pigments are improved by using a soluble emulsifier (i.e., soap) followed by a precipitant. Soluble Emulsifiers: Sodium or potassium soaps of oleic acid, stearic acid, recinolelc acid, etc. Precipitants: Barium, calcium, magnesium, aluminum, strontium and lead salts. Organic Media: Linseed varnish DUP03001 3947 -64 U.S. 2,112,222 - InterChemlcal Company, 9/21/1933 (Cont1 d.) Pigments: Para toner, lithol R Red, and Prussian Blue Process? The flushing Is carried out In a simple mixer; the final drying in a steam jacketed container. U.S. 2,335,760 - Du Pont Co.. 11/30/1943 This patent discloses the direct incorporation of a hydrous iron oxide pigment into a cellulose derivative vehicle by flushing. Process: Kneading in a M & P mixer followed by (a) sheeting the pigmented mass through a loosely set steel rolls and tray drying, (b) vacuum drying in the M & P mixer. Vehicle System: Cellulose nitrate derivations. Transfer Agents: Presumably blown castor oil and dibutyl phthalate. Pigment: Hydrous iron oxide presscake freshly prepared. U.S. 2.964.418 - National Lead Co., 12/15/1960 This patent discloses a process for flushing a basic cadnium salt of a fatty acid. Process: Vibratory mixing apparatus or heavy duty Baker-Perkins mixer followed by vacuum evaporation. Organic Phase: Plasticizers (such as dioctyl phthalate); mineral oils; polyester of adipic acid and glycol; soybean oil, etc. DU P030013948 65- .)U.S. 2,964,418 - National Lead Co,, 12/13/1960 (Cont'd Transfer Agents: Not mentioned specifically; increased temperature facilitated phase transfer; aliphatic alcohols also were used. Materials to be Flushed:_________ Cadnium salt of fatty acids such as cadnium caprylate and cadnium laurate, etc. U.S. 431*130 - Chemetron Corporation, 3/4/1969 Dry pigments of fine particle size and dispersibility are produced without agglomeration by dehydrating a presscake with a low boiling organic solvent and combining with a solid salt dispersing agent which is thermally stable under the con ditions of mixing but volatile when heated up, Process: 'i Presscake was washed with a low boiling solvent until most water Is removed from the pigment. The solvent presscake was mixed with a solid salt dispersing agent in a ball mill. The resulting mix ture was heated up to the decom position temperature of the dis persing agent and thereby leaving dry pigments in a non-agglomerated state. Thermal Deeompassable Dispersing Agents: Ammonium carbonate. Pigments: CPC blue & green; indanthrone; quinacridone, etc. DUP030013949 -66- U-S. 5,508,941 - Kogrs,, Co.,..4/2,8/lgJO This patent describes the flushing of nacreous pigment crystals such as crystals of lead hydrogen arsenate by using a polyethylene phosphate as flushing agent. A typical examples is given as follows: A thickened paste of lead hydrogen arsenate platelet crystals was stirred with a mixture consisting_of dioctyl phthalate and a small percentage of polyethylene phthalate as flushing agent. With gentle stirring at 8oC, all of the crystals were transferred to the dioctyl phthalate phase. After separation of free water, the residual entrained H2O was expelled by centrifugal separation at 2000G. Process: As described above, involves mixing/centrifugal separation of residual water. Organic Phase: Dibutyl phthalate, dioctyl phthalate, nitro cellulose solution, organic solvents--hydrocarbons, esters, ketones, alcohols, etc. Transfer Agents: A I R-0-(CH2CH20)-P-B It 0 Where R =* Alkyl, alkenyl. Alkyl phenyl radicals in which the alkyl portion has C4- Cis. N * 3-15 A & B are selected from group OH and RO(CH2CH20) ( o c h 2c h 2 ) -o r DUP030013950 -67- U.S. 3,508,943 - Koppers Co,, 4/28/1970 (Coat'd.) Applicable Piam.en.ts: Lead carbonate, lead hydrogen arsenate, lead hydrogen phosphate and bismuth oxychloride. British 1,212,346 - ici, 11/11/1970 This patent discloses an improved process for "flushing" of pigments from dispersions in aqueous media to non-aqueous media. The process consists of (a) ball milling the press- cake in water with a water-soluble salt of a water-insoluble carboxylic acid (e.g. potassium linoleate); (b) the aqueous dispersion from (a) being mixed with an alkyd resin; (c) the "flushing" being carried out in a heavy duty mixer with the assistance of a precipitating agent (i.e., acetic acid) which converts the water-soluble salt into a water-insoluble form depositing on the pigment particles; flushing and water sepa ration follow immediately; (d) the flushed dispersion still contain significant amount of water on pigment weight). The following patents are not directly related to "flushing" but contain comments and references on "flushing": U.S. 3.565,656 - Allied Chem. Corp.. 2/23/1971 The patent relates to a soft, easily dispersible iron oxide pigment obtained by taking i.e., precipitating freshly prepared ferrous/ferric hydroxide together with alum inum hydroxide. It also contains information concerning the art of flushing of iron oxides. U.S. 3*553,93.4. - This patent discloses the method for dehydrating nacreous pigment platelets. The process involves washing the presscakes with water-miscible solvent and collecting the platelet pigments by settling, the settling rate being increased by using current through the dispersion. It also contains a review of. flushing methods for the nacreous platelet pigments. DUP030013951 -68ABSTRACT The Important factors governing pigment flushability and guidelines for selection of flushing aids are discussed. A simple and economical process for flushing transoxide yellow pigment into organic phase is described. The flushed dispersion could be used in organosol lacquers. The influence of fundamental particle size on film transparency and metallic twb-tone was elucidated. DUP03001 3952 -69- DISTRIBUTION LIST p Copy No. Full Reports 1 2 3 4 5 6 7 8 9 10 11 12 13 -15 16 17 18 19 20 -23 24 25 26 27 28 R. S. Prengle, Wilm. G. I. Mulholland, Wilm. R. E. Pike, Wilm. E. T. Breuer, Wilm. E. H. Berg, Marshall Lab K. A. Saegebarth, Marshall Lab P. B. Allen, Marshall Lab J. R. Chalmers, Marshall Lab H. S. Wellman, Marshall Lab S. Hochberg, Marshall Lab W. S. Zimmt, Marshall Lab P. M. Hodgson, Marshall Lab Library, Marshall Lab (3) F. M. Gavin, Flint C. N. Swinehart, Flint D. M. Marsh, ESL W. M. Duffy, ESL File Room, ESL (3) J. B. Lyman, Marshall Lab F. N. Jones, ESL T. A. Ashe, ESL I. C. Chu, ESL Writer File First Parts W. D. Lawson, Wilm. J. A. Klacsmann, Wilm. ) In Turn J. C. Richards, Wilm. ) S. R. Miller, Wilm. ) D. L. Herndon, Wilm. ) K. F. Browning, Wilm.) In Turn A. J. Wells, Wilm. C. W. Stahl, Wilm. T. R. Matthews, Wilm. R. W. Laurrell, Wilm.. J. G. Metzger, Jr., Wilm, A. W. Andreseh, Marshall Lab C. E. De Boer, Marshall Lab D. M. Glenn, Marshall Lab C. H. Knop, Marshall Lab L. W. Crissey, Flint J. M. Donatello, Flint A. V. Scancella, Fairfield T. R. Louer, Wilm. A. F. Nugent, Flint C. W. Bulloch, Flint P. L. Mitchell, Parlin R. H. Vining, MRL DUP030013953