Document 6wBRLoOmBao7roOeL3BpQoVX6

495986J CT-JL-9E-6 CHEMICALS CONTAINS CONFIDENTIAL INFORMATION. RESPONSIBILITY OF THE RECIPIENT, WHEN NO LONGER NEEDED, DESTROY BY BURNING OR SHREDDING l.----------------- ------------- .. . ......... --. Distribution on last page. CT-JL-92- 06 MARCH, 1990 copy No. l~J JACKSON LABORATORY RESEARCH & DEVELOPMENT, TECHNICAL REPORT CHEMICALS AND PIGMENTS DEPARTMENT E. I. DU PONT DE NEMOURS & COMPANY MAXIMUM LOW PVC TI02 HIDING POWER EFFECT OF PRIMARY PARTICLE SIZE AND AGGREGATION Work Done By: Report Written By: Approved By: previous Related Reports Project Code: Type Technical Work: Period Covered: Notebook Numbers: Personnel: January 1984 - December 1988 (Part Time) E32956, E39389 R . G. Fernald, Professor John Vander Sande, MIT Abstract Commercial Ti02 pigment hiding power potential has been reported to be -65% of theoretical. Our primary particle size and aggregation experimental programs (low PVC) suggest that our best commercial pigments scatter at about 80% of theoretical. Hiding power losses are -15% for primary particle size distribution and -5% for primary particle aggregation. N40672 I. INTRODUCTION Hiding power is the single most important attribute of Ti02 pigment. If this is true - A. What is the ultimate hiding power of Ti02? and B. How do commercial pigments (duPont and competitive) compare with the above? In order to measure these hiding power losses we need to understand the hiding power theory. The low pigment volume concentration`system was selected for study as it represents the simplest system from the standpoint of optical theory and modelling. The major Ti02 hiding power losses in low pvc systems are due to deviation from monosize and monodisperse particles. II. OBJECTIVES Optical theory predicts ~15% hiding power improvement for optimum monosize Ti02 compared to commercial pigment particle size distributions. Our program is to show experimentally (for the first time) that the theory is correct and -15% hiding power potential does exist. Ti02 aggregates are formed in the oxidation process by collision of primary particles. Hiding power is lost as multiple particle clumps are formed due to inefficient spacing of the Ti02 particles. Past work has suggested that hiding power could be improved -15% in low PVC systems by decreasing aggregate size from 4 to 1 particles/clump. Our program is to determine the actual aggregation of the Ti02 pigment to accurately assess the potential hiding power improvement. Optical theory predicts Ti02 scattering values significantly higher than our experimentally measured s values (corrected for width of distribution and aggregation). investigate the optical theory and suggest reasons for these differences. -1- DUP050055124 Ill, SUMMARY AND CONCLUSIONS Primary particle Size Measurements - Monosize o DuPont theoretical scattering calculations predict hiding power can be increased ~15% by narrowing the primary particle size distribution to monosize Ti02. o DuPont experimental classification program has demonstrated ~5% increased hiding power by narrowing the primary particle size distribution at constant mean diameter. o Extrapolation of experimental program geometric standard deviation data predict monosize Ti02 hiding power can be increased ~15%? thus, the experimental work is in good agreement with the theoretical calculations* o Work is currently directed towards narrowing the width of distribution in the oxidation process. Aggregation Measurements - Monodispersion o Apparent aggregation is 2-4 particles/clump (end-use) o Actual Ti02 aggregation is -1.6 particles/clump -1.6 particles/clump represents -5% hiding power loss or -1/3 of calculated theoretical hiding power - the remainder HP (7-10%) is due to Ti02 crowding. o Aggregate hiding power variables oo End-use system (random or charge stabilized) oo Pigment volume concentration oo Ti02 pigment (type and grinding energy) Ti02 Scattering Review The major difference between Tip2 scattering values by optical theory and corrected experimental measurements may be in our measurement procedures. Data are presented and discussed. -2- DUP050055125 IV . PATENT STATUS No plans to apply for patent coverage V. FUTURE PROGRAMS None by the author - Program suggestions can be found in the discussion VI. PUBLICATION STATUS No plans to publish VII. SPECIAL SAFETY CONSIDERATIONS No special problems outside of normal laboratory hazards VIII. ENVIRONMENTAL CONSIDERATIONS No special problems IX. ACKNOWLEDGMENTS This was a part time program conducted over a five year period. Many discussions, suggestions, and constructive criticisms were offered by the R&D community. Hopefully most of these have been incorporated into the report. -3 DUP050055126 TABLE OF CONTENTS PAGE NUMBER I. INTRODUCTION II. OBJECTIVES III. SUMMARY AND CONCLUSIONS IV. PATENT STATUS V. FUTURE PROGRAMS VI. PUBLICATION STATUS VII. SPECIAL SAFETY CONSIDERATIONS VIII. ENVIRONMENTAL CONSIDERATIONS IX. ACKNOWLEDGMENTS X. DISCUSSION A, INTRODUCTION B. PRIMARY TI02 PARTICLES 1. BACKGROUND 1 1 2 3 3 3 3 3 3 6 6 8 2. EXPERIMENTAL PROGRAM 3. PRIMARY TI02 PARTICLE SIZE MEASUREMENT C. AGGREGATION 1. BACKGROUND 13 2. EXPERIMENTAL PROGRAM a. DUPONT/MIT PROGRAM b. END-USE SYSTEM AGGREGATION c. VARIABLE PVC END-USE INVESTIGATION DUP050055127 d, AGGREGATE REDUCTION BY GRINDING e. IMPROVED AGGREGATE SIZE MEASUREMENTS D. AGGLOMERATION E. TI02 SCATTERING REVIEW REFERENCES DUP050055128 X. Discussion A, Introduction The greatest physical attribute of pigmentary titanium dioxide is its ability to scatter light and provide end-use product opacity/hiding power. This Ti02 scattering is dependent upon a number of physical variables including: o index of refraction (pigment and vehicle) o wavelength of light o particle size and distribution o particle shape o aggregation o dispersion o impurities (color) o end-use formulation - PVC, Ti02 loading, etc. For most of our studies, the majority of these variables are constant; thus, we can concentrate On the pigment variables * primary particle size, aggregation, etc. Let's start by defining some terms - you may personally use other definitions but this report is based upon the following definitions. Primary particle - smallest discrete particle or building block - we will see later that this does include both growth and collision twins Aggregate - clump of primary particles which is difficult to impossible to break up * once you have formed an aggregate you are stuck with it Agglomerate - clump of lightly bonded aggregates which is shear sensitive - the grinding operation in the Ti02 process is primarily a deagglomeration process -6- DUP050055129 We do know some of the pigmentary properties are controlled or influenced by particle size. Figure DPF-.l presents a particle size distribution of R-900 08 code (high gloss) product showing general quality parameters as a function of particle size. The distribution is an aggregate distribution for the most part. For comparison let's look at the three most important distributions for oxidation base Ti02 -primary particles, aggregates, and agglomerates (Figure DPF-2). The oxidation base is highly agglomerated as it is produced. High shear wet or dry grinding will produce an aggregate distribution with a median diameter of approximately 0.3-0.4 micrometers. The primary particle size distribution is measured by transmission electron microscopy with a median diameter of about 0.16 to 0.24 micrometers. Oxidation base slurry is really an agglomerate distribution composed of aggregates which in turn are composed of primary particles. The mean,size of this agglomeration distribution has varied widely among the oxidation lines. (Ref, 1, 2, 3) Figure DPF-3 presents some additional particle terminology. Primary particles range is shape as well as size, some primary particles result from growth or collision twins such as example A. Bonding of primary particles to form aggregates also occurs. Example B shows typical sintered and chemically bonded primary particles, obviously there is a fair amount of confusion between A and B as the particles overlap. We have used a tilting stage on the electron microscope to make these distinctions. To make life even tougher -most aggregates are Composed of more than two primary particles. We try to deal with this by resolving how many scattering centers we have present; that is, how many primary particles are actually in the aggregate. Table DPF-1 presents common particle size measurement methods for primary particles, aggregates, and agglomerates, A number of marketing needs originate in the oxidation area and are particle size related. A list of these needs is presented in Table DPF-2. The major needs are higher CBU (smaller size), increased hiding power, and improved gloss. As the individual oxidation lines are slightly different from each other (Table DPF-3), a single solution to these problems in the oxidation area does not exist. A more practical solution is to understand the basic fundamentals of each problem and then work back through the process. What is the ultimate hiding power of Ti02? Theoretical studies twenty-five years ago suggested that commercial Ti02 pigment has -30-40% hiding power potential in low pigment -7- DUP050055130 volume concentration systems (enamel, plastics, etc.). (Ref.4) The improved hiding power is split about evenly between: o Monodisperse Ti02 (no aggregation) and 0 Monosize Ti02 (all optimum size particles) We have worked part time over the last few years trying to better understand not only what is theoretically possible but also what is commercially possible; that is, how can we improve our understanding and also demonstrate experimentally improved hiding power in low pigment volume applications. Primary Ti02 Particles 1. Background Let's first consider the role of primary particles. Table DPP-4 and 5 present some Ti02 primary particle facts and observations. Figure DPP-4 presents Bill Ross' chart showing relative hiding power as a function of wavelength, and Ti02 primary particle size. An optimum particle size exists for maximum hiding power? that is, all particles produced smaller or larger than this optimum size will decrease the hiding power. Small particles scatter more efficiently in the blue; this is the basis for our CBD test which measures the blue to red ratio and predicts particle size. We are primarily interested in the green curve as the green response more nearly duplicates what the eye sees. Approximately 0.21 micrometers diameter rutile particles are about optimum for maximum HP. A relationship exists between CBD and primary particle size diameter as explained above. The best fit for primary particle size as measured by transmission electron microscopy and CBU for oxidation bases is presented in Figure DPF-5. Approximately -15% higher hiding power has been calculated for optimum monosize Ti02 particles (W. D. Ross - Ref. 3). No experimental verification of this higher hiding power for more narrow distributions was ever demonstrated. Figure DPF-6 presents graphically the salient points on Ross' chart Figure DPF-4. What we have is a distribution of primary particles and what we want is for all the particles to be the same size and optimum -8- DUP050055131 diameter. The 15% HP improvement is the theoretical improvement and is not necessarily attainable, part of our program is to define what is possible and/or reasonable. The demonstration of significantly higher hiding power by decreasing the primary particle size width of distribution is necessary to justify R&D development efforts in this area but ------ production of monosize Ti02 particles is thought to range from difficult to impossible from our current Ti02 agglomeration process. The next two charts (Figures DPF-7 and 8) show the difficulty in the production of monosize Ti02 particles. The first figure on DPF-7 shows that commercial pigment is not monosize and presents distributions for three different mean particle size pigments. Increasing particle diameter increases the width of distribution. This-.is characteristic of an agglomeration process. The second figure on DPF-7 shows that the width of distribution is commonly expressed as the geometric standard deviation and is calculated by either the ratio of d50/dl6 or d84/d50. Figure DPF-8 presents geometric standard deviation as a function of primary particle size for a group of duPont and competitive pigments, both chloride and sulfate processes. Regression analysis shows a linear best fit. Figure DPF-8 also presents an extrapolated version of the same plot. This work predicts a geometric standard deviation of -1.0 at 0 micrometers. A GSD of 1.0 means monosize particles. This suggests that any finite size particle will exhibit a distribution of particles and the width of distribution will increase as the diameter increases. Significant differences in width of distribution at a given diameter were observed suggesting that some level of hiding power improvement might be attainable. Experimental Program A laboratory program was undertaken to demonstrate higher hiding power by narrowing the particle size distributions with centrifugal sedimentation techniques. Vinyl film testing was selected for measurement due to its low Ti02 sample end-use requirements. The normal test loading was further reduced by 80%, Table DPF-6 presents early lab tests to show that vinyl TS could be used to measure hiding power, A weight reduction of Ti02 in the formulation showed a comparable loss in vinyl TS. A reduction of Ti02 in the standard formulation yielded DUP050055132 similar relative TS values to the control. R--101 was selected for study due to its superior performance in vinyl film. Since we were primarily interested in single particle size width of distribution effects, the R-101 was remicronized in the laboratory at 6 S/P to minimize aggregate coarse tail. The vinyl TS increased for 108 to 114. This is our actual starting point from which to measure increased hiding power as a function of decreased width of distribution. Initial centrifugation efforts produced smaller and possibly more narrow distributions. Highspot hiding power work showed no significant improvement for the fine fractions. We believe we replaced the inefficient oversize particles with inefficient undersize particles. Attempts to remove the inefficient undersize particles (<0.2 micrometers) by sedimentation field flow fractionation were unsuccessful. Multiple cut centrifugation was conducted to develop hiding power data as a function of particle size and width of distribution. If centrifugation is successful in making primary particle size cuts from a finished R101 product, then by definition the width of distribution of the fractions would be more narrow than the starting material. Vinyl TS data are presented for the multiple cut centrifugation work as a function of CBU in Figure DPF-9. The optimum CBU appears to be -16.5 which is in good agreement with past work. Hiding power increased as the Ti02 size increased until a maximum was reached; then hiding power decreased as particle size further increased. This is the familiar HP chart that Bill Ross developed from theoretical Mie scattering (Figure DPF-4) only now developed from experimental data. The maximum HP demonstrated was 119 TS which is 5% higher than our double micronized starting material. While the hiding power improvement is attributed to a more narrow distribution, centrifuge fractions were submitted to MIT for primary particle size distributions (TEM). Additional centrifuge fractionation work was conducted using a centrifuged fraction as the starting material. No additional hiding power improvement was demonstrated. Evidently the centrifuge classification procedure had reached its limit. Another technique for varying the width of distribution at constant particle size (CBU) is by blending small and large particle size pigments. -10- DUP050055133 In this case the blending of small and large particle size commercial products will have a wider particle size distribution than the commercial TI02 product with the equivalent mean size. The low PVC hiding power will also be poorer due to the wider width of distribution -see Figure DPF-10. MIT projected area measurements were converted to diameters and then means, medians, GSD, etc. calculated. A plot of low pvc hiding power vs. geometric standard deviation is presented in Figure DPF-11. All samples plotted have diameters ranging from 0.155 to 0.165 micrometers. To increase the range of GSD's we included a sample (equivalent diameter) that was prepared by blending high and low CBU products (large width of distribution - 1.730. As vinyl TS of the plant R-101 was increased 6% by removal of aggregates/agglomerates (double micron!zing), the blended plant product was adjusted accordingly (104 to 110%). Figure DPF-12 presents extrapolation work as well as a summary. Approximately 5% increased hiding power has been demonstrated at equal median particle size by centrifugation. Centrifuge techniques narrowed the primary particle size distribution. Extrapolation of the linear best fit line predicts that -15% high hiding power can be attained with a monosize Ti02 pigment (GSD = 1.0). This work is in excellent agreement with Bill Ross' early scattering calculations. Table DPF-7 presents additional data from the centrifuge series. Noteworthy items include: o multiple centrifugation produced samples with c b u 's ranging from -12 to 22 from a feed CBU of 14 o TEM aggregation measurements show starting material and centrifuge fractions were equal A summary of this primary particle size work follows: o DuPont theoretical scattering calculations predict hiding power can be increased -15% by narrowing the primary particle size distribution to monosize Ti02. o DuPont experimental classification program has -11- DUP050055134 demonstrated ~5% increased hiding power by narrowing the primary particle size distribution at constant mean diameter. o Extrapolation of experimental program geometric standard deviation data predict monosize Ti02 hiding power can be increased -15%; thus, the experimental work is in good agreement with the theoretical calculations. o Work is currently directed towards narrowing the width of distribution in the oxidation process. Primary Ti02 particle Size Measurement Historic TiQ2 measurements yield primary particle size distributions with increasing width as primary particle size increases. These distributions generate excessive specific surface area compared to actual sample measurements; thus, the actual primary particle size must be larger. This decreased surface area and increased particle size can occur via fusion and sintering. Other aggregation mechanisms do occur such as small neck sintering and chemical bonding but these mechanisms would not greatly affect the specific surface area. Enamel and plastic hiding power (low PVC systems) are controlled by primary particle size, distribution width, and aggregation. We need to develop additional information before we can accurately predict future hiding power improvements and potential. A good first effort would be to measure primary particle size distributions and determine whether the sintering process is constant as a function of CBU and oxidation line. Of equal importance would be to measure the best competitive products for comparison. It is possible that all oxidation products are similar and no short term hiding power improvements are predicted/possible. If this is true we need to know it. It is also possible that we have historically optimized the current chloride process for low PVC hiding power and if we do learn to inhibit fusion/aggregation we will have to produce larger particles for optimum scattering. Experimental evidence for measurement of the sintered particle size distributions does exist (Figure DPF-13). This work by Buchanan, Hench, and Fields shows the classic TEM primary particle size distributions (CRB count) and the sintered -12- DUP050055135 "scattering center" counts by Hench. With the development of smarter image analysis hardware and software, these counting objectives become more realistic. Aggregation 1, Background Ti02 aggregates are formed in the oxidation process by collision of the primary particles. Bonding is controlled by Ti02 surface chemistry as well as the individual oxidation line conditions. The effect of aggregation on product quality is primarily associated with Ti02 scattering (low PVC hiding power, CBU, etc.) although the coarse tail does effect gloss in both emulsion and solvent systems. Past theoretical and experimental work by WD Ross and RJ Bruehlman (Ref. 5, 6) suggested low PVC hiding power could be improved -15% by reducing Ti02 aggregate size from -4 to 1 particles/clump (monodispersion) - see Figure DPF-14. Our program was to determine the actual aggregation of our oxidation bases and products to accurately assess the potential hiding power improvement. Bill Ross made theoretical calculations on the effect of multiple particle attachment (Ref, 5). He found a hiding power loss for 0.2 micrometer particles of ~4% per particle attachment or -12% for a 4 particle/clump aggregate -(Table DPF-8). While these calculations were for a linear geometry, Ross stated that they would also be true for other common geometries. This hiding power loss occurs due to particle to particle interference. For maximum scattering we desire particle separations of -1/4 wavelength or about 1/2 the diameter of the Ti02 particles? thus, we do not want the Ti02 crowded together. Ross' work also showed that aggregation played a minor scattering role as the Ti02 diameter increased. At 0.3 micrometers diameter aggregation had essentially no effect on scattering. Experimental work by Hennessy and Paulson in 1968 showed a relationship of hiding power with reduction of mean agglomerate size - Figure DPF-15. The size reduction was controlled by grinding. I would interpret this minimum size by grinding to represent sample aggregate size. Extrapolation of aggregate size to mean crystal diameter would represent a reduction in aggregation; thus, the -13- DUP050055136 hiding power improvement is estimated at -15%. The major problem with this rationale is that most particle size measurement techniques are relative and no two techniques should be used on the same data plot. A summary chart on aggregation types and measurement techniques is presented on Table DPF-9. A new instrument has been developed for measurement of Ti02 aggregate size - the X-ray scanning disc centrifuge. A comprehensive report on this instrument was issued (Ref. 7) by the author. Our evaluation work suggests that fine particle size measurement (both accuracy and precision) is improved vs. gravitational equipment such as the Sedigraph. A direct comparison of XSDC and Sedigraph particle size distributions for R-900 is shown in Figure DPF-16. The XSDC produces a smaller and more narrow distribution than does the Sedigraph. The number of 0.2 micrometer primary particles possible in the median equivalent spherical diameter of R-900 is 13 particles/clump for the Sedigraph and 4 particles/clump for the XSDC. The equivalent spherical diameter from the XSDC is much more realistic than the Sedigraph when compared to transmission electron microscopy studies and past historical aggregation measurements. Figure DPF-17 shows these relative particle size distributions. Experimental Program a, Dupont/MIT Program Much aggregation counting work has been conducted by Professor John Vander Sande, MIT and us through the years. Figure DPF-18 presents a flow chart of how the aggregate measurements were made. Figure DPP-19 and 20 show relationships between researchers and methods. While differences do exist, these techniques must be considered relative rather than absolute. In an effort to document this aggregate counting work, typical oxidation line counts and oxidation line ranking are presented in Figure DPP-21 and Tables DPP-10 and 11. As can be seen, no measurements have been made for the more recent deagglomerated fin flue oxidation line developments. Aggregation among oxidation lines was originally thought to be constant with aggregate a function of primary size (Sedigraph median size equal to 2.25X primary particle size). While this is still probably true within an -14- DUP050055137 oxidation line, significant aggregation differences do exist at constant CBU among the various lines. The best lines were the high pressure Antioch and JVII lines and the worst line was EMU. EMIX appeared not only larger but also much more highly sintered. The addition of co ox. P205 significantly reduced EMII aggregation. The examination of both primary particle size and aggregation suggests that CBU is a function of both primary and aggregate size. At constant CBU, the oxidation lines with larger aggregates will have smaller primary particles while the least aggregated lines will have larger primary particles. There is little past evidence to show CBU is a function of primary particle size and aggregate size within a single oxidation line. This may mean that the aggregate bond strength is too great to decrease aggregation by conventional grinding. The one possible exception is Mike Baloga's recent work on the media mill where significant CBU increases have been noted -ostensibly from reduced aggregation. Table DPF-12 presents typical oxidation line and finished product measurements while Table DPF-13 presents all of our aggregation measurements made during these investigations. A few observations -Table DPF-14 shows that oxidation lines with more sinter bonding have larger aggregates. It should also follow that aggregate reduction should be more difficult due to sinter bonding. Table DPF-15 shows the effect of surface chemistry (co ox. A1203) on aggregation. Aggregate size is constant even though the bond mechanism changes from chemical to sinter bridges. This suggests that aggregation may occur independent of bond type or in other words mother nature strikes again. Table DPF-16 and Figure DPF-22 show the aggregation changes due to sandmilling and micronizing. Sandmilling reduces aggregate size (as well as micronizing) and when combined with micronizing appears to minimize the oxidation base aggregate differences. An attempt was made to correlate MIT aggregate counts with Sedigraph clump size data. Table DPF-17 presents aggregate means and medians by both number and weight for a progression of samples through the Ti02 process and typical end-uses. The major observations are: o aggregation is a coarse tail problem rather than a primary aggregation distribution problem (low median by number) -15- DUP050055138 o End-use systems vary in aggregation (same pigment) - possible reasons include system shear, end-use surface compatibility and PVC (Ti02 crowding) Seven competitive R-900 products were evaluated vs. DeLisle, Johnsonville, and Antioch R-900 code 08 products. DuPont aggregate counts show the best competitive R-900 products were equal to the best duPont products (Table DPF-18). A series of variable energy micronizing tests were conducted on JV1 R-101DD (Table DPF-19). Aggregate measurements Show no significant differences for the micronized products. This suggests grinding harder does not reduce aggregation (TEM). On the other hand vinyl tinting strength does improve as grinding energy increases and has been correlated to the reduction in oversize coarse fraction (%>0.6 micrometers) - Table DPF20. If agglomerate/aggregate coarse tail is variable and aggregate measurements do not see it, it may be that coarse tail clumps >0,6 micrometers are too large for the TEM mounting process. This would also suggest that the coarse tail is possibly a second distribution (bimodal) as the aggregation data are not at all influenced by the coarse tail. b, End-Use System Aggregation If aggregation is a primary player in determining end-use hiding power, then it becomes obvious that we must resolve how may particles/clump our pigments contain in the end-use. Historic counts indicate 4 particles/clump while our work suggests 2-4 particles/clump. Initial end-use evaluations are presented in Table DPF-21. Observations: o product quality (gloss) plays a minor role on aggregation unless it is poor o significant aggregation differences are observed depending upon pigment volume concentration aggregation increases as PVC increases It is logical that as the concentration of Ti02 increases that the apparent aggregation increases due to Ti02 crowding. The implications of this work are that Ti02 aggregate size is smaller than measured in past studies -16- DUP050055139 and may be only 1-2 particles/aggregate. Potential hiding power improvement from a strict aggregate size reduction would appear to be less also. Additional work is required to substantiate this position. Table DPP-22 presents a graphic summary of our aggregation work thus far. The five oxidation lines investigated are different and range from 3-6 particles/clump. Following wet and dry finishing, the aggregate size decreases about 30% to 2-4 partic1es/clump. The ranking remains unchanged; that is, the larger aggregate bases exhibit larger aggregate finished products. The main area of concern is not the dry product but the end-use system as it is here where the Ti02 scattering is measured. End-uses studied ranged from about 1.5 to 3 particles/clump (different systems). Limited work suggests that PVC may play a major role in measured aggregation. c. Variable PVC End-Use Investigation A computer program was written to produce random particles in two dimensions as a function of time. The pigment volume concentration was calculated from the number of particles, particle diameter, and X-Y area. Computer down loads are presented in Figure DPF-23. Aggregation was measured using similar techniques to those used for Ti02 electron micrographs. The increased particle attachment can be readily observed as PVC increases (Ti02 crowding). We label this aggregation as "perceived" aggregation as all Computer simulated particles were single particles (monodisperse). A plot of mean aggregate size vs. PVC is shown on Figure DPF-24. This work suggests that a major pari: of our end-use aggregation could be "perceived" (Ti02 crowding) and not due to the actual pigment aggregation. The next step to resolve this issue was to investigate the various end-uses at variable PVC Ti02 additions. Even though a given end-use was designed for a specific PVC, our work would look at a range of PVC's. The vinyl end-use system results are presented in Figure DPF-25. Aggregation increased with concentration. Similar slopes (computer simulation and vinyl system) suggest the vinyl system is randomly aggregated. This is reasonable as the vinyl test is a high shear viscous fast freeze system. The actual aggregate size of the pigment is the aggregation at 0 PVC or approximately 1.6 -17- DUP050055140 particles/clump. The hiding power loss for this aggregation would be less than 5%. Figure DPF-26 presents similar end-use aggregate measurements for an automotive enamel. The actual pigment aggregation is again about 1.6 particles/clump. The higher aggregate sizes measured previously in the standard auto formulation were due to Ti02 crowding. The difference in slope between the computer simulation and the auto enamel suggests that the auto enamel is not randomly aggregated but the system is probably charge stabilized; that is, Ti02 crowding has been minimized. The third system, emulsion enamel, is presented in Figure DPF-27. This time the Ti02 PVC is significantly higher around 2.5 particles/clump. We believe that this increase is probably due to poor paint grinding (Ti02 dispersion) as the industry uses a Cowles grinder rather than sandmills. A similar slope to the automotive enamel is observed. This suggests that the emulsion paint is also charge stabilized. A summary of our low PVC aggregation studies follow: o Apparent aggregation is 2-4 particles/clump o Actual aggregation is -1.6 particles/clump 1.6 particles/clump represents -5% HP loss or -1/3 of theoretical hiding power calculated -remainder (7-10%) is due to Ti02 crowding o Hiding power variables oo End-use system (random or charge stabilized) oo Pigment volume concentration oo Ti02 pigment (type and grinding energy) d. Aggregate Reduction by Grinding We discussed earlier our inability to reduce aggregate size by variable energy micronizing (fluid energy mill). Aggregate grinding might occur if the grind time was increased to some unreasonable level. Ball milling was selected as it is used in the sulfate process to grind Ti02 agglomerates/aggregates. Figures DPF-28 and 29 -18- DUP050055141 present ball milling particle size data for both an oxidation base and a finished R-900 product. Significant particle size reduction occurred in 12 weeks and it appeared only a matter of time before the aggregate structure was size reduced to that of single crystals. Aggregate size reduction appeared too uniform and prompted samples for analytical tests. After 16 weeks of ballmilling the oxidation base sample contained 38% Si02 and A1203. The size reduction of the grinding media was the real cause of the observed oxidation base particle size reduction. This would also have to be considered as a potential problem for sand/media milling experiments. e. Improved Aggregate Size Measurements Oxidation base agglomerate and aggregate sizes are currently measured in aqueous dispersions on the Sedigraph/XSDC. Agglomerate size is measured on the "as is" (no shear sample), while aggregate size is measured following ultrasonic grinding. "Aggregate" size has been previously defined as a multi-particle structure which is difficult to impossible to break apart. For convenience, this structure has been further defined as the Ti02 particle clump remaining following rigorous steam micronizing. The oxidation base mean aggregate size (ultrasonic grinding) is always larger than the post micronizer aggregate size due to lower grinding energy. This difference is due to the presence of hard agglomerates that are not broken down in ultrasonic grinding, but are reduced by micronizing. The aggregate differences are oxidation line dependent; that is, the more agglomerated oxidation lines exhibit the largest aggregate size (constant CBU) following ultrasonic grinding. The measurement of oxidation base "true aggregate" size requires development of a high energy laboratory mill that simulates micronizer grind conditions. In addition to mean aggregate size, this mill would also be useful for measurement of the oxidation base coarse tail that controls both alkyd and emulsion gloss. A vibrating mill employing small glass spheres has been investigated (dry impact). This mill was originally developed by Tioxide for the preparation of Ti02 samples for electron microscopy. EMII, JVI, and JVII oxidation bases were milled using a 20 second grind time. These data are shown in Figures DPF-30, 31 and 32. vs. -19- DUP050055142 Sedigraph distributions for both oxidation base and finished R-900 products at three micronizer grind intensities (S/p - 2, 4, and 6). In each case, the vibrating mill overground the oxidation base samples. This is the first time we have observed a grind technique superior to fluid energy milling* Grind time was investigated from 10 to 80 seconds. These data (JVI r --101 base) are shown in Figure DPF-33, It would appear that a grind time could be selected to match a given coarse tail or mean aggregate size. Whether this will be constant for all the oxidation bases is not known. A 10-second mill time was selected for evaluation of EMII, DEL, and JVII bases. The oxidation base coarse tails are plotted vs. their respective coarse tails for laboratory R-900 (S/P = 6) products (Figure DPF-34). A 10-second grind time appears to be getting close to matching the R-900 micronized product coarse tails. The aggregate mean sizes for the vibrating mill samples (oxidation bases) were identical to the lab R-900 (S/P = 6) product samples. We appear to have a laboratory grind technique capable of measuring both the finished product aggregate mean size and coarse tail starting with oxidation base samples. Agglomeration Ti02 agglomeration has been extensively addressed in our past oxidation base studies but a few words are in order. Ti02 agglomeration forms in the oxidation base and continues through wet and dry finishing. Ti02 agglomeration in oxidation is currently not totally inhibitable but is certainly reducible by grinding in finishing. It is true that flocculation, a form of agglomeration, can be a severe end-use problem. Flocculation is formed from particles already size reduced thus flocculation is concerned with Ti02 stabilization rather than size reduction. Our goal is to size reduce or remove all agglomerates as the agglomerates are equally as bad as aggregates in the end-use. Figure DPF-35 presents the effect of finished product coarse tail (agglomerates) on film properties (gloss). Figure DPF-36 overlays the oxidation base agglomerate distribution onto the finished product aggregate distribution. The difference in these distributions (marked) represents the agglomerate fraction in the oxidation base that must be inhibited, removed, or size reduced. We have shown in our previous oxidation base studies that agglomeration is reduced by increasing CBU and decreasing oxidation base turbulence. See Figures DPF-37 and 38, -20- DUP050055143 Sandmilling is also a viable candidate for oxidation base deagglomeration (Ref. 8). Representative data is presented in Table DPF-23. To produce high quality finished products for today's markets, we must reduce our oxidation base agglomeration. Recent studies suggest that our current oxidation base target is: 25% >p.6 micrometer coarse tail maximum at 12.5 CBU and 20+ tons/hour A Ti02 agglomeration summary follows: o Finished product coarse tail controls end-use performance [film properties (gloss) and HP (vinyl TS)] o Finished product coarse tail is a function of oxidation base coarse tail o The origin of the coarse tail is unknown - it may be the coarse tail of the growth distribution or a separate distribution such as wall scale o Oxidation base/finished product coarse tail is reduced by: oo increased CBU oo decreased agglomeration (turbulence, etc.) oo surface chemistry (coox. P or Si) Ti02 Scattering Review The greatest physical attribute of pigmentary titanium dioxide is its ability to scatter light and provide end-use product opacity and hiding power. This Ti02 scattering is dependent upon a number of variables including: o Index of refractions o Wavelength of illumination o Ti02 particle sise While we cannot do much to alter the first two variables, we need to understand their basic relationships. Lets's start with the index of refraction. -21- DUP050055144 Ti02 hiding power is proportional to MA2 where (Ref. 9) M = ([Np/Nv]A2 - l)/([Np/Nv]A2 +2) Lorentz-Lorenz Expression and Np ~ pigment refractive index Nv = vehicle refractive index An approximation for Ti02 hiding power is proportional to 0.16(Np-Nv)]A2 This approximation formula is limited to Nv =5 -1.5 and Np = 1.5 to 2.75 (low PVc applications) Optimum Ti02 particle size can be estimated from the above relationships. d(optimum) = wavelength/([2]A{1/2)*Nv*HP*Pi The above refractive index relationships were programed in QuickBasic. The program INDXREF1.BAS and sample output are appended (ATT - l, 2, & 3). Hiding power (arbitrary scale) for both Anatase and Rutile Ti02 as a function of vehicle refractive index is presented in Figure DPF-39. Ti02 scattering (hiding power) increases as the vehicle refractive index decreases. Rutile has approximately 25% more hiding power than Anatase at 1.5 vehicle refractive index. Figures DPF-40 and 41 present plots of Rutile and Anatase hiding power (both calculated and approximated from above equations) at variable vehicle refractive index. While not identical, both relationships would appear to be fine for calculating relative scattering differences. A plot of optimum diameter for Rutile Ti02 at variable vehicle refractive index is presented in Figure DPF-42. Optimum diameters for Rutile Ti02 at 1.5 vehicle refractive index are -0.16, 0.18, and 0,20 for blue, green, and red wavelengths. Figure DPF-43 compares Rutile and Anatase (green wavelength). Ross Mie scattering calculations do not seem to be in agreement (Figure DPF-44); in fact, Ross's work does not show a diameter dependence on vehicle refractive index. We have very little input into these hiding power variables as end-use system design determines the vehicle refractive index and we do not know how to significantly increase the refractive index of Ti02 above that of Rutile. The refractive index is also dependent upon the wavelength of illumination and on crystal axis orientation for anisotropic pigments. The use of average -22- DUP050055145 refractive index vs. ordinary and extra-ordinary refractive indices for Rutile is acceptable for scattering calculations. Refractive index is higher for the short wavelengths {blue) than for the longer wavelengths (red). WD Ross calculated Hie scattering relationships as a function of Ti02 particle size and wavelength (blue, green, and red) (Ref-10), This work was programed for convenience in w d r is c a t .b a s and PRTSCATl.BAS. Sample output from these two programs is appended, A plot of these relationships is presented in Figure DPF-45. This is the familiar Ross chart presented earlier as Figure DPF-4. The undertone (B/R) can be calculated from Ross' work. Figures DPF-46 and 47 present undertone as a function of Ti02 particle size diameter. This relationship is pretty much linear over the pigmentary size range (0.16 to 0.24 micrometers). This undertone relationship is for monosize Ti02 particles at low pigment volume concentration; Let's compare this response with actual experimental reflectance measurements. Figure DPF-48 presents the primary particle size-CBU relationship for oxidation bases (TEM counting -mean data). This relationship is for commercial Ti02 not monosize Ti02. We next determined the relationship between the undertone (B/R) and CBU for the Ti02 CBU test in silicone oil (Figure DPF-49). Figure DPF-50 uses these above relationships and presents a plot of undertone vs. particle diameter for the experimental data. This plot looks identical (similar slope) to the original plot (DPF-47) from Ross' Mie calculations. The difference in absolute undertone might be explained by the addition of carbon black in the experimental CBU test or due to the aggregation in the commercial pigments. The above is reasonable evidence that the CBU test is measuring primary particle size differences in oxidation bases. We are next interested in determining the Ross Mie scattering for Rutile Ti02 as a function of vehicle refractive index and primary particle size distribution. This work was reported by Ross in CPJL-87-08. Figure DPF-51 presents monosize Ti02 scattering as a function of diameter for a single vehicle refractive index. Figure d p f -52 presents similar data for different widths of distribution. The absolute decrease in Ti02 scattering is readily observed as the width of the primary particle size distribution increases (monosize to commercial distributions). At optimum particle size this difference has been calculated to be -15% hiding power using actual TEm particle size data. Monosize Rutile Ti02 scattering is presented in Figure DPF-53.as a function of particle diameter for a number of vehicle refractive indices (Ross -CP-JL-87-08). Figure DPF-54 presents Ti02 -23- DUP050055146 scattering as a function of vehicle refractive index. These data were obtained from Figure DPF-53. Additional Ross data were found in WP-ESI-75-2 report (Figure DPF-55) and converted to Figure DPF56. Scattering data from Figures DPF-54 and 56 were combined and are presented in Figure DPF-57. Why spend so much time determining the theoretical Ti02 S values as a function of vehicle refractive index? WD Ross proposed that the observed scattering loss at high pigment volume concentration (Ti02 crowding) might simply be due to the change in the paint film refractive index. This reasoning also would explain the increase in scattering observed above the critical PVC (decreased paint film refractive index due to entraped air). The paint film refractive index is a function of both the vehicle refractive and the pigment refractive index thus increases at high Ti02 loadings. Again these 5 value calculations are for monosize Ti02. The calculated loss in relative hiding power is about 15% comparing monosize particles with actual commercial pigment particle size data (TIM data). Ross and others measured paint film refractive index for Rutile Ti02 as a function of pigment volume concentration (Figure DPF-58) (Ref. 10), The Ross/Armstrong and Kawabata data appear reasonably uniform. My estimate of the line of best fit is noted on Figure DPF-58 and replotted in Figure DPF-59. We can now convert pigment volume concentration to medium refractive index and use the medium refractive index in the Mie scattering equations to predict Ti02 scattering at higher PVC's. Figure DPF-60 presents calculated Ti02 scattering as a function of PVC for monosize Rutile Ti02 particles as a function of particle diameter. This curve looks similar to the experimental Ti02 S curve determined in coatings. Figure DPF-61 presents Figure DPF-60 data at 0,2 micrometers particle diameter in addition to experimental data developed by Ross and Bruehlman. The slopes appear to be right in line but the absolute S values do not; the measured S values are significantly lower than calculated. Can we explain these differences? Monosize Ti02 scatters -15% greater than a commercial Ti02 product. The observed scattering difference is significantly larger than 15%. The next biggest influence on scattering is probably Ti02 aggregation. Figure DPF62 presents Figure DPF-61 with the addition of aggregated commercial Ti02 pigments at 2, 3, and 4 particles/clump. Scattering was reduced 4% for each additional particle/clump in keeping with Bill Ross' aggregate calculations. The theoretical scattering is still too high even when aggregated to 4 particles/clurap. This means that the experimental scattering measurements are not as high as they should be (or the Mie scattering calculations are too high), A number of reasons suggest that the experimental data are probably too low. Our day -24- DUP050055147 to day hiding power measurements are low because we measure specular green reflectance instead of diffuse reflectance and make no corrections for sub-surface reflection. The Ross data set did measure diffuse reflectance and transmittance but neglected sub surface reflection. our dilemma is summarized on Table DPF-24. Ross and Bruehlman measured commercial Ti02 pigments at 0.52 MA2/GM with a vehicle refractive index of 1.514 (low pigment volume concentration). The calculated Mie scattering S value for monosize Ti02 is 0.785 MA2/GM. The calculated loss in S value for the commercial pigment is 0.265 which suggests that scattering can be increased ~ 51% ([0.265/0.52]*100) above current levels or - 34% ([0.265/0,785]*100 of theoretical. We feel confident that the loss in hiding power for commercial products is -15% for width of primary particle size distribution and -5% loss for aggregation thus we have an impasse. This coupled with the observation that we have not significantly increased low PVC hiding power in coatings in the last 25 years suggests that a problem exists in our measurement techniques. Sophisticated Ti02 S value determinations were made by Cairns and Spooner in the 196O's at Chestnut Run (Ref. 11). This work was part of their color matching program. Their work showed a significantly higher experimental S value than those derived today. The work utilized mylar drawdowns for diffuse reflectance and transmittance measurements as well as corrections for sub surface reflection. This work is presented on Table DPF-25. Cairns and Spooner measured the Ti02 S value at 0.70 for a 1.48 vehicle refractive index (low pigment volume concentration). The calculated Mie scattering S value was 0.885. The difference in S values is -21% (loss from theoretical) and compares favorably with our estimated S value losses for width of distribution and aggregation (-20%). If these data are correct then the obvious conclusion is that there is no significant HP loss in today's commercial pigments that has not been accounted for in low PVC systems. Where do we go from here? Current Ti02 products should be measured at low PVC by the methods of Cairns and Spooner and then compared to Mie scattering S value calculations. If we can duplicate Cairns and Spooner's original work, we can feel confident that we understand Ti02 scattering at low pigment volume concentration. If this work is successful then Ti02 scattering work at higher PVC's can be followed-up using Ross' proposal for utilization of paint film refractive index. Ti02 crowding (particle separation as a function of surface treatment and geometry) will also have to be addressed. -25- DUP050055148 XX. REFERENCES: X. DP Fields, CS Daw, Ti02 Oxidation Base Characterization/ Optimization, CDP-EM-79-5 2. DP. Fields, Ti02 Oxidation Base Optimization, CP-EM-81-8 3. DP Fields, Reactor Discharge Slurry Products, CP-EM-84-ll 4. WD Ross, Theoretical Computation of Bight Scattering Power, Journal of Paint Technology, Vol, 43, No. 563, Dec. 1971 5. WD Ross - Unpublished Calculations 6. RJ Bruehlraan, Effect of Aggregation on the Hiding Power of Rutile Pigments, WP-ESI-65-13 7. DP Fields,. Ti02 Particle Size Measurement, CT-JL-91-16 8. DP Fields, Emulsion Gloss - Particle Size Studies and Laboratory Sandmilling Studies, CP-JL-90-3 9. PB Mitton, Chap.III-D-c, Pigment Handbook, Wiley and Sons,1973 10. WD Ross, Light Scattering by Titanium Dioxide pigment, CP-JL87-08 11. DL Spooner, Determination of Absorption and Scattering Coefficients for 5 White Pigments using an Improved Method, WG-66-74 -26- DUP050055149 i h d i h m a s AoNtanoaiac^ -27- DUP050055150 T I0 2 A G G R E G A T E P A R T IC L E S IZ E D IS T R IB U T IO N D IA M E T E R -- M IC R O M E T E R S FIGURE DPF-2 SIZE - MICRONS A0N3003Ud 3AIAV13U -28 DUP050055151 FIGURE DPF-3 PRIMARY PARTICLE BONDING A - PRIMARY PARTICLES HEAVY FUSION OR SINTERING B AGGREGATES SMALL NECK SINTERING CHEMICAL BONDING AGGREGATE FORMATION SCHEMATIC SINTER NECK -29- DUP050055152 TABLE DPF-1 TIQ2 PARTICLE SIZE DETERMINATIONS PRIMARY PARTICLE SIZE DISTRIBUTIONS MEAN SIZE -TEM -SEM -CARBON BLACK UNDERTONE (CBU) -SURFACE AREA -SEDIMENTATION (SONICATED) FINES SEDIGRAPH, x s d c , s f f f AGGREGATE SIZE DISTRIBUTION -TEM -SEM -SEDIMENTATION (SONICATED) SEDIGRAPH, XSDC, SFFF AGGLOMERATION DISTRIBUTION MEAN SIZE -OPTICAL MICROSCOPE -SEDIMENTATION (AS IS) SEDIGRAPH -BULK DENSITY -OIL ABSORPTION -HG POROSITY -30- DUP050055153 DECREASE OXID BASE AGGLOMERATION AND COARSE T A IL AT HIGH PRODUCTION RATE H 0Z w8 8S as sB *Ift QQK UXH ft X j ft < H QX2Q <N H fffttt I HX o 1N 1o | w wgj 2, i^cmU^ >ftv___g_g_S HwOft WM W0 ft HO >! fftt fut o ooo 0 .B8 GLOSS IS CONTROLLED BY PARTICLE S IZ E {COARSE T A IL ) IMPROVED GLOSS -31- DUP050055154 TABLE DPF-3 OXIDATION FACTS o OXIDATION LINES ARE NOT EQUIVALENT o OXIDATION BASE QUALITY VARIES WITHIN A LINE o MAJOR OXIDATION LINE OPERATING DESIGN VARIABLES ` o PRESSURE o TURBULENCE o COOLING RATE o MAJOR OXIDATION BASE QUALITY PARAMETERS o PARTICLE SIZE (CBU) / AGGREGATION O CLUMPING AGGLOMERATION \ COARSE TAIL o SURFACE CHEMISTRY -32- DUP050055155 TABLE DPF-4 PRIMARY PARTICLES ULTIMATE BUILDING BLOCKS - SCATTER LIGHT TO PROVIDE MAXIMUM END-USE HIDING POWER OPTIMUM MEAN SIZE o UNDERTONE o HIDING POWER, TINTING STRENGTH oo REFRACTIVE INDEX oo WAVELENGTH OF LIGHT oo PARTICLE SIZE OO PIGMENT VOLUME CONCENTRATION WIDTH OF DISTRIBUTION AS ABOVE -33- DUP050055156 TABLE DPF-5 PRIMARY TI02 PARTICLES - OXIDATION BASE STUDIES O MEASUREMENT - UNDERTONE (CBU), TEM, SURFACE AREA o MANY CRYSTALLOGRAPHIC TWINS D LARGE HUMBER OF HIGHLY SINTERED MULTIPLE PARTICLES O RUTILE WITH SOME ANATASE AND POSSIBLY BROOKITE o VARIABLE SHAPE - SMOOTH TO ANGULAR O WIDTH OF DISTRIBUTION (GSD) INCREASES WITH INCREASING MEAN SIZE O ALL OXIDATION LINE HAVE SIMILAR MEAN SIZE AND WIDTH OF DISTRIBUTION FOR A GIVEN CBU (ASSUMPTION) O 1% CO-OXIDIZED AL203 oo APPROX. 50% OF AL203 IS ON PARTICLE SURFACE OO NON-UNIFORM SURFACE (AL203 THICKNESS) O VARIABLE SURFACE SI02 (UP TO 25% OF TOTAL SURFACE) -34- DUP050055157 CO <0 o cc Q FIGURE DPF-4 DIAMETER, MICRONS ONIQIH 3AllV13y -35- DUP050055158 -36- FIGURE DPF-5 CARBON BLOCK UNDERTONE (CBO) xi DUP050055159 HIDING POWER IMPROVEMENT (LOW PVC SYSTEMS) PREDICTED HIDING POWER IMPROVEMENT - 15% io op -J lit U. N O 55 lO I- z C30 z5 _UJi 2 WD ip *9- * So > Ui oE DC l S 2 P* co a: xO Ui uj O CO uj 2 -J o n cc* <p fi-Sco 2xw P<H 229 j-Cff Q. O OiS CO _Uli 0 1 g Ui N CoO z o 2 2 P 2 P Q. O. < a -37- DUP050055160 PRIMARY PARTICLE S IZ E DISTRIBUTIONS WIDTH OF DISTRIBUTIO DUP050055161 -PARTICLE DIAMETER MICRONS ms a g e FIGURE DPF-8 -39- DUP050055162 TABLE DPF-6 MAXIMUM LOW PVC HIDING POKER VINYL FILM EVALUATION - VARIABLE TI02 LOADING HP <TS> AS f(LOADING) % TI02 LOADING 100 (SID) 90 95 99 101 105 110 REL. VINYL TS 100 89 94 98 100 104 109 REDUCED TI02 LOADING VINYL LOADING (CMS) PRODUCT A 5' 2 1 PRODUCT B 5 2 1 REL. VINYL TS 109 107 106 113 112 113 -40- DUP050055163 -41 MAXIMUM LOM PVC H ID IN G POWER PARTICLE SIZE DUP050055164 -42- PIGMENT BLENDS VARIABLE WIDTH OF D IS T R IB U T IO N DELTA UNDERTONE DUP050055165 -43- MAXIMUM LOW PVC HP VS GSO (050/016) (DIAMETERS CALCULATED FROM PROJECTED AREAS) GEOMETRIC STANDARD DEVIATION (D 50/016) BY HUMBER DUP050055166 -44- MAXIMUM LOW PVC HP VS BSD (0 5 0 /0 1 6 ) (DIAMETERS CALCULATED FROM PROJECTED AREAS) GEOMETRIC STANDARD D EVIATIO N (0 5 0 /0 1 6 ) DUP050055167 p* CO * in CO to p* in 40 10 o o o O . ** 10 O CD N N 01 0* in H N IN N H H rl in in M N ts to to to to to tO N to N in ts rl to * -i M H m H r10t to p* t* 03 CO H H wCQ I P* EM O o O O o *O MH e BBB TABLE DPF-7 -45- DUP050055168 FIGURE DPF-13 s M 9dm SJCz W0U3VW XHDH* -46- DUP050055169 CO 2 LU h* z O CO S 0 O> CL UJ z 0 s o o z CO UJ H2: d UJ z 2 LU CM >o O DC CL D UJ CO Z C UJ & UJ CO I oo z CL o 0s D X o O * U) 1 I jh* UJ 2 sUJ z a. i z UJ I CL 0 o s 0 UJ Io 5 UJ z Q. FIGURE DPF-14 -47- DUP050055170 RELATIVE HIDING POWER CALCULATIONS VARIABLE T I0 2 AGGREGATION DC UJ H LU 5< 5 z<oo oDC s O to CVI 00 O 0) O) CO CSI o 0. X LU > _J UJ DC CO oUJ _l O CO LU --I o UJ _l tz CO UI o ih* DC p DC UJ UzJ o o UJ DC X DC =o> u. (0 oCO DC o -48- DUP050055171 FIGURE DPF-15 ENAMEL HIDING POWER VS. MEAN PARTICLE SIZE (HENNESSY AND PAULSON - 1968) 1)0 100 0 0 El ----- --u. 0 0.1 0^2 OJ 0*0 ' bis 0.6 oT o'.l KM HMffctt SIR * 0.9 hm t i iding w w * -49- DUP050055172 TABLE DPF-9 STOKES LAW (X-RAY SCANNING DISK CENTRIFUGE) BETTER AGREEMENT WITH TEM X < UJ X 0 XQ Ul Z Y-% UJK az 2 a UJ e iz 0 0 UJ X g0 UJ _j g H" o CC LU sX 0 O< , Ul X 0 X Ul X 0 x UJ0 X? CL *** a<x 0 sX 0 00 r-5 0 UJ 0 X 0 3lit 3s 1 z UJ UJ 1X a11" UJ 8 %& 8 1 1CC o UJ Q z o m ze> D X f0 co UJ 0 0Q zO X X o ffl xs UJ h- 1 0 UJ Z 0 X o H z UJ 2 UJ X P o X CL x 2 x _j C<L x 0N. 0 UJ 0 0 UJ 0 h X X >* CO UJ 2 0 Ul s z o Q 0 Ul P o UJ z2 XP ox UJ tb- 0 UJ z Ul o s 0H UJ sc 2 UJ e C *- 0 B$ CD < UJ UJ LU XX I0 0 0 -50- DUP050055173 -51- DUP050055174 DIAMETER - MICROMETERS FIGURE DPF-17 EQUIVAIENT SI >- cl O O soo oc SO O Z *5 Ph 30 CQ UJ frru-Jj $2 <0 UJ N <5 < OC h- UJ 3g g2 g <H ctz goUJ o? 0i OJ x 1 oc z Q<. OC o 5 sUoJ -52- s DUP050055175 to LU HffCaaaIO Q =003 UJ w o <CDc aLU: CD <CDc -53- DUP050055176 FIGURE D P F -ig AGGREGATE COUNTS TEH- MEAN ARI AGGREGATE SIZE (PARTICLES/CLUMP) V \ \ \o \ V \ \ \ \ \ \o o * \0 \ \ \ V* *\ \* X 3c \ V <y> oo - N. t> LT\ fO 04 X <i> in J_ X 04 (dWfnO/SHlOUUVd) 3ZIS 31V93`d99V INOd fid NV3W -54- DUP050055177 (dwmo/s3TDuavd) azis Bivsaaosv nvsw --55-- AGGREGATE COUNTING MEAN ARI AGGREGATE SIZE (PARTICLES/CLUMP) DUP050055178 GGREGATION MEASUREMENTS -DUPONT COUNTS PART/CLUMP DUP050055179 TABLE DPF-10 AGGREGATION OXIDATION LINE RANKING SAMPLE DUPONT MEASUREMENTS KBRR-MCGEE ANTIOCH DELISLE JVII JVII W/CO-OX P EMI JVI W/CO-OX P JVI EMU EMU S/CO-OX P,SI 2.71 +/- 0.41 3.03 +/- 0.64 3.19 +/- 0.51 3.30 +/- 0.64 3.60 4/- 0.48 3.76 +/" 0.12 3.87 +/" 0.55 3.90 +/- 0.84 4.01 +/- 1.23 4.20 4/- 1.22 ARI MEASUREMENTS 2.94 4.38 4/- 0.38 5.91 4.95 4/- 0.22 6.10 4/- 0.49 6.28 4/- 1.08 6.11 4/- 1.13 FINISHED PRODUCTS ANTIOCH JVII JVI EMU DELISLE COMPETITIVES 2.14 +/" 0.18 2.15 4/- 0.55 2.51 +/- 0.38 2.58 4/- 0.36 2.78 2.88 4/- 0.46 2.66 4/- 0.13 2.62 4/- 0.24 3.63 3.97 4/- 0.73 2.40 2.91 4/- 0,60 -57- DUP050055180 b, QOi 1 I o DUP050055181 8 w ?S SS' o 8 &K * mk m CM s*s 8 NR IN M K o fe M SSS2SgSSgS8?8 aHU{ O'w Rf in ^.K W N t0t0t>i R CM S' * 8' S RSSin&RB* fO 1010 tOtOtOkOM'tOtfl m a! _ a. 3 I -- orr3^rnC cr x x33 XZZO.3 88yy 8"o C- a 35 S|8 ooxir aXirj w*-+* nol S gig gwM ssj OR is glWMtO VO<HTt OirtCa M 2.3L HsgrfYYH w ro r r j o o o V rv rv is is " a*iOsU9gD<h ^(^9> y.u (frofttf*cr S S w 3 3 5*h m 5 S JK S 3 2^Li2QI2l2llQ!2QCCi | iu t v S S ! ftsg lOvOtj S8& ? li lliilsi W s s8i*: 3 3 ii (V to ^ K) A fN CD 9 O *44ts4l104 ^r( in la4K4 O4j<tHiOt!4rNtNN KaSS&SRSSBBSBS AGGREGATE COUNTS -59- DUP050055182 . 3 MEAN ACC S IX 4 MEAN ACC S IX mm 3 85188 nnvn S*rr3o Rto.w3o s ro 5 833S85SSS3$3SS3SSSi35R5 $RS3&R$i SR inwMwww^n^iOTinrttWNTWWttw n n n mio n ^ rotRnnnnnnt'nr) n n 2 DESCRIPTION AGGREGATE COUNTS TABLE DPP-12 CONT'D 1 CODE UuK>ON in CO S ox *s.c /v St m______ UOtflUHCA :--v3--mt-f-lo-Ke.Lw2ci-ii SiXCL*fd.B. cs^souQOfzcOdVuouOsosCgtoOteOuosPugs t i 3a b sa Iff! OCflECEOS QQ CD CD CD 5322 5s oo TV a I as a aoao Su55u> Spxxixl p s* !S is sgai ioo M eo*j ON Sc 553 5? Ho4 Qa a v ? Illinmrfims u --iL-iu j SS uiujuiu bj w ujuiui biui lb to l u l u u | TV7T S7TS&SS& u K&&&&&&& OK-O4 C<vO4 Q4 w4 CO *1 Mi UJ Ui <c ui UJ O"' asJsasB! tn SH 5 WWSS*4* 3S33^3?8E?8S88fe88SS33S8a&8SRRRRSRRRRR8SS -60- DUP050055183 0 SAMPLE TYPE i 3 tEAN ACC S IX 4 HEM ACC S IX 0*0 ISN fv CD 5*8 X *SKSS8!5 SS8&S Van V (4 N^NNNNN NNNNf) 3 a SSKSSgSSBS^B&g.sfeitR.sasssssiaR r e ?r 8g s k b s s s m s s s b &k s s NR>NNni<iwfnnt>i -h ^c mc m mn rmNNnNN 2 DESCRIPTION AGGREGATE COUNTS TABLE DPP-12 CONT'D 1 CODE n IIUOiUOI H0.H0. 33^ aa?a aa*cocona3s 8 1 0.0.0.0.0 0. ii HH U-U. 1 8CXXJ8O235U5SOS2SOS2SOS mm 55 H M 1 o_nI o x n o a ssissTi: Kttft usmmzhUw1**U3 m m (fi0i0T Jii 40 Ifi i>01> tfiCP 0* 5! WUJU4 s?; toS SS Ct d_ l BBTiL;8^< gS5*Sw8 .IlsSSlN inms |H 4 %*44 T^4l v4* 44*41 liiSii IS YYYIIS .W?tT a mm mm mm gssgg iaaacia 8 2X88&SS8SSi^XRX&*88g|28miSS32SS&3SSSRa38^S5g?!R -61- DUP050055184 0 SAMPLE TYPE n 5S ft w CM CM w 1 !S N(0 (9 St si & gPB* 3..8'&'f!s:;?*fc.S8:883$gR SB*fiK'S'S N *H rt < 1-4 gg'RS'g! KINN NN I cStltswa.. NwNo^CiaDiI^Uy. g-gj cn tntn ssss tntnw npgofoiKt3 dddS ddd* b flfagg*gto*a*tg*a*B*a*o*<M>'nVj!2Pm9t!jPrflOM' S&$!g j ss CM a535S5-55*SS St-t n 8PIsH*aaliiiifffff 1 ddi~S<S <c <K<saeo:p:a!B:Kae i KttKUl 5-^w SSSSSS e. c-.ce. uSt n C3D 3CD3CD Ift r J s go s TABLE DPF-12 CONT'D a S $ ft 8 ffl ? S 5 S 5 V5 8?8 S 8 8 3 8 8 S? 8 S S 2 S3 382 8 -62-_ DUP050055185 $ s8 tom cr* CO c % 1% 8 K a St % So s feK $ c mm to ? 3.'S.B* imo it 8r's*v8xiu8>* 58 31? 8 58 2 823838388 tl> or 88 S**I MH tnui -e Sgg- 9 ,8833 88 aaaSaaa2 a as&J 8 N jidd* -ari aI d;v*SS S eo o ox x b . SXX""Ixa"ZSiia.iEi.H8 gS` r_a.a.a.o.a. . .tr -'ggS3? iiiS 3&XJI KI oSi 4 Mu -0S4*g, 04 Mj il a fc-o Sg lwt O aiSSS5fe*MMVEQ S S2U L Mf ...S..SiIs. wm g &M&i2cgg&{ Is 66! tST T TV ftSS Ttv IliilsiPfcg| U1U*J*Q tf& TABLE DPF-13 AGGREGATE COUNTS I X 8 a *<*>!* 3 g g -63- DUP050055186 w s CVsOK S3 S 8i ass <rw 3 ? v R0* 0S S*O 3 W. 8if) . SSS nnn 38SSSSRRSSSSS!SSRSS V(s*i'tritMNVnM*i<5w sN sCMsN*sC.fSerjtso*tsaals K>* SrOatO&lOSMaIDSMS MS MS. MifSYW&SMSN' M M 1 TABLE D P F -1 3 CONT'D AGGREGATE COUNTS nomoyOHSii) S8=g58; i|g||i auuSSui til ^ ~-- * 5*72 KftfeS J IP? Pi K *61?* as non 3 555 XlOI*Om tm B IS [S 33 s on ass I|0N)WWNU). ggsXsX R ^aoo --- "<$ HS5 5 5&* Ql I I $ w4IS l*1<TlT<6QOOTf 0> *4 v< r4 4 jS ffl v* v H H vl H H JEHxrfHBT ?*v 3 j2 in un I 595 o5j u4s0o pppopo^o>clsDcbiPoooonQDD0aac0 01 * &?&BS88fe8&33S3383&88J^RRfcRKf88B8SB*&*i:8&*te# -64- DUP050055187 n H a a! w rn r4C.M fM H S n nmnwm-n n SfO SNVNS BNN*8WBM*8M8 M38W$*.N1M8.3N&*KN W3 RM SfNSt<K*m3MS*nBMRt R3??^SSf?^!p viv(NNrtHI v 4rs|NN o si g CL CL >:>: Cl CL cNl !cSl t! cnc/* H C**OwM*lltOaWH^ WN<^Hv.,j CL w4 03 CB CD CD 03 CD: o 25555153' p N as ua Uat-4.1cU9-l4I CL. CL yy cioi 3 XV 86dTHt o. o. c l c l a. a. X X X XX X 888881 uouogu U t*frSai ii 55 U-J 8* ,rtp|5 Mgg -ieBS= UQ og ** g 2 3< 888588 AGGREGATE COUNTS U> Jl Jl Ul J> Jl U> jj)' a 8&&&5 IsBSiBiBi $$ $ s s BBe SBcacBBBBiSB I 83 MB mm.m CQQQ ^ru*Jicieoisaieai: ais ctc S o< (A I 0 8g!84<gH>gH8v<SN8M845v|84 s*1a*H*s"aa4s4va4'Ms*&4 <s *s<g.*4a**R-<R4s<*4a*4^*4&*HR4R4 ^rt4jS4R<H *S*4 .R<**S4 wHP:*!jHSM R*4 *t4s4s* !?*4?in -65- DUP050055188 wHto IM 8 li S R SB 5 S eJ s &s b &s CMC4M 5S 8&R5ff$as?GI6*l* SSSRSBfe8SESSRfc8B8s:RfetiSSfeSSSSSRf! cn rJrgriroNcviNMNMCJM 44.<h 4h *n h w < TABLE D P F -1 3 CONT AGCRE fCtTE COUNTS fsps WH :k:*S:S:So g wwwwcn^fgS. j Jg iJSSs g B ft jg i sgBfieaaaB^ w40>4 WWMW |liifiis:i;BS|i35Si33illl? rSs ?ss? So? "P:s::ii8gs8s|B|gg||isgii N ^35533 HpSu^K^aa ccB:o:ii :ococ ttKi PONO] Uc sUFSIhUEhJlii fififiBI BIEfifilBSIB*S P\(|l ffigill' rfltli RIRfcRg gfjB8lfr8KgBg sta&ssB aSaEcpcGccsh i 6 ts sslaSSSSS fegssgssxga&s v(944*-tU4.rt.4V<5<494w< s^KRjtRjs^RRs msggsBS&.g.s.g < *< *V * --' --* O' XX* -66- DUP050055189 9 Pi c-J ra stosts'a IyW w Ca' O* fri&lDM IOINNNN NCSIN OT CD.cn 888g s eoe K'# I fIt a,ti a,iUuJ >flu >omu >ma. y>a. N 3iu ^uji4jhqu0v; <(Dr-*^6T-0MU-)_Or4- * <d id ooa *9* S55SbU$* o9 5HJI|p| SSS8tt:S<i< N<fMWO> 4 SS8&SSSSSSssssisissssi 4-t wlw4*4< ^ W N W M NNNMNN MNN -67- DUP050055190 TABLE DPF-14 ti si ea e *i *2; as!j &* s 2=g & a fc! i t M 1 s fc*o . oi CO s03 acc tS3 IsSe SS ee " _"J Sf= ca _ 5 Htn VO CO _J ys --WST> OLU ste CO Q {3 as Safe m c> 'w fei ` tn ud sg *j* 13 2; g 33 5 -68- DUP050055191 TABLE DPF-15 AGGREGATE COUNTS ANTIOCH BASE INVESTIGATION STD BASE NO CO-OX A1.2D3 NO "Q" NO CO-OX AL2O3 WITH "Q" BONDING GEL BRIDGES SINTER SINTER MEAN* M A,3 3,9 MEDIAN BY ff. 2.2 MEDIAN by vrr 6.6 1,6 6,4 2.0 5,6 *TQTAL if QF PARTICLES TOTAL # OF CLUMPS -69- DUP050055192 a i-- ill o _l 9Q --*--I 0UJ a: p w a. z o_ <coc ocn _s e_: io on on LL. -i o c m in sr m c\i c m o rn AGGREGATE SIZE MEASUREMENTS VARIABLE OXIDATION BASES Q * OQQC- 9 in CO o in O c m c m in in rn o cni cn Q LU _l _--J1 UJ oo ?H in K. r-H CM sz c -=r cr in <> cr a cq z CO CUOJ PQ cr 00 00 o m < St J3- in t*C in o in *--4 --< CM .*--4 *-- D_ c o o uj z--:4 LU s in UJ z*--1 _l UJ X zo -* 1 --I o o >""5 9> s CO LU 9 sC > M cc *sc -70- DUP050055193 -71- AGGREGATE COUNTS MEAN - OXIDATION BASE (PARTICLES/CLUMP) DUP050055194 TABLE DPF-17 CLUMP SUMMARY OXIDATION BASE AS IS (SEDIGRAPH) SONICATED (SEDIGRAPH) PETRONATE (TEM) R-900 MZ'D PRODUCT (TEM) MEAN - 4.4 2.6 PARTICLES PERCLUMP MEDIAN BY NUMBER MEDIAN ELWgGHT - 16.5 - 7,5 1.6 6.4 1.2 3.1 END USE SYSTEMS EMULSION GLOSS PAINT (R-900-78 GLOSS) AUTOMOTIVE GLOSS PAINT (R-900-78 GLOSS) VINYL PLASTICS (R-lOO) 3.4 2.5 2,0 1.9 1.2 1.2 4.2 2.8 2.1 AGGREGATION IS A COARSE TAIL PROBLEM AND NOT A PRIMARY AGGREGATION PROBLEM (LOW MEDIAN BY NUMBER) -72- DUP050055195 TABLE DPF-18 HIDING POWER - AGGREGATION SERIES II COMMERCIAL HIGH GLOSS R-900.PRODUCTS a u t o . JWSW a g g r ;'---VINYL pr o d u c t * . mm &m u m. m j iz l is . o i * TIOXIDE RHD-6X S 82 78 102 13.5 3.0 109 -0.016 TIOXIDE RTC-90 c 79 76 102 13.5 4.2 107 -0.007 K-M CR-800 c 77 74 103 15.7 2,8 110 0.000 NJZ RF-30 c 75 71 99 11.7 2.4 106 -0.017 GLIDDEN RC1-9 c 78 71 102 13,3 2.5 107 -0.016 AM CYAN 0R-600 s 79 74 98 11.3 2.6 99 -0.028 NL 2090 s 78 72 99 14,2 2.8 102 -0.014 DU P - ANTIOCH R-900 c 78 76 103 12,5 2,8 102 -0.024 DU P - JV R-900 c 81 79 102 12.8 2.9 105 -0,026 DU P - DeL R-900 c 79 77 101 13.5 2,4 103 -0.022 SIGNIFICANT DIFFERENCES IN OA, HYDROUS A1203, SURFACE AREA, ETC. **FR0M R-101 STD -73- DUP050055196 TABLE DPF-19 HIDING POWER - AGGREGATION STUDIES SERIES III JV R-101 DP - VARIABLE LAB MZ ENERGY SAMPLE 1 2 3 A 5 6 7 8 NO MZ'ING INCREASING GRINDING ENERGY v S/P=12 m 11.9 12.2 12.2 12.1 12.2 12.1 12.6 12,0 AGGREGATE SIZE*.. A,3 2.5 2.9 1.5 2.9 2.7 2.2 2.5 VINYL TS UT 91.5 1.005 103 1.009 107 1.011 108 1.010 109 1.015 no 1,016 in 1.017 111 1.013 *IN VINYL -74- DUP050055197 TABLE DPF-20 HIDING POWER - AGGREGATION STUDIES JV R-1Q1 WORK ASSUMPTIONS ORIGINAL - IMPROVE HP BY DECREASING AGGREGATE SIZE (4-1) m. - IMPROVE HP BY GETTING RID OF COARSE OVERSIZE ASSUME CLUMPS >0.5v DO NOT SCATTER LIGHT VINYL TS 91.5 103.0 107.0 108.0 109.0 111.0 SERIES..III. NORMALIZE DATA FOR 111 MAXIMUM HP % >Q.5u (SEDIGRAPH) 25.0 12.0 10.0 8.0 7.0 5.5 VINYL TS 91.5 103,0 107.0 108.0 109.0 111,0 % >0.5p 25,0 12.0 10.0 8,0 7.0 5,5 NORMALIZED 2 >0.5w 19.5 6.5 4.5 2,5 1.5 0.0 PREDICTED VINYL TS (MAX - NORMAL GRIT) 91.5 104.5 106.5 108.5 109.5 111.0 -75- DUP050055198 TABLE DPP-21 AGGREGATE COUNTS - PAINT FILMS (DUPONT) AUTOMOTIVE GLOSS (30J) '15 FVC 56 GLOSS 67 GLOSS 78 GLOSS EMULSION GLOSS (TFW-ISO) '22.8 FVC 52 GLOSS 69 GLOSS 7S GLOSS FACTUAL ARI COUNT KEAN CLUMP 8IZE 2.0 2,0 1.9 1.7 5.5 2.8 2.7 2.5 EST. ARI CLUMP SIZE 3.0 2.9 2.6 2.5* 7.6 4.5 4.4 3.4* AGGREGATE COUNTS - PLASTIC FILMS (DUPONT) MEDIUM HARD POLYVINYLCHLORIDE ('1 PVC5 R-900. COMPETITIVE PRODUCTS KEAN CLUMP SIZE TIOXIDE RHD6X TC 9635 1.6 TZOXISE RTC-90 TC 9533 1.8 KERR-MCGEE CR-BOO TC 9548 1.5 GLIDDEN RCL-9 TC 9482 1.5 N L 2690 TC 9621 1.7 Jt-900 4193 1.6 EST. ARI CLUMP SIZE 2.0 2.6 1.7 1.7 2.3 2.0 VARIABLE FVC R-900 4193 0.5 FVC R-900 4193 fl 1.0 FVC R-900 4193 9 FVC 1.5 1.5 2.4 1.7 1,7 3.9 -76- DUP050055199 (1 PVC) TABLE DPF-22 CO H Z LU 2 LU cc 0 z Z) CO H < LU o o LU 2 < Ui 0 h" LU DC 0 UI CL N2 55 3 UI O ft co UI ^UccI Q ofc CD " < ( m- co II CO CM N O W CM CM *~ Ua.I > H <0 UI <3 S < D D O CC CL CM h Z 0 o UI Ui I- X CO o1z CL CC tZ oCM h 0O 1 > CL CO is. CM _J UI zO O UI s <c CD 2 l DC z< UI w Ui UI CO > oUi CO oD CO s i -J zo D 2 UI UJ < oCM VINYL FILM -77- DUP050055200 FIGURE DPF-23 COMPUTER SIMULATION OF "RANDOM AGGREGATION" 2 DIMENSIONS (RANDOM X-Y GENERATOR) CALCULATED PVC 2.8 v: % * v . I v1 9 9 0 5.7 !." * l.`*. * / ' > 8.6 11.3 13.7 25.6 48.2 DUP050055201 -79- 2 3) cc III H- RANDOM AGGREGATION AS FIPIGMENT VOLUME CONCENTRATION) CALCULATED PVC DUP050055202 FIGURE DPF-25 DUP050055203 -81- S|w po 0- op 2 i-O o OUJ o <w T I0 2 AGGREGATION AS PIPIGMENT VOLUME CONCENTRATION) CALCULATED PVC DUP050055204 zo 5 _l iSz <0gO Utf O tlOUJ 3c a -j S =5 2. 0 2=J OUIU- fZ o FIGURE D PF-27 CALCULATED PVC -82- DUP050055205 FIGURE D PF-28 EQUIVALENT SPHERICAL JN33U34 SSVW 3AUVTDWrO -83- DUP050055206 PARTICLE SIZE DISTRIBUTION I * it 1N3DH3J SSVW i/UiVYWffiD -84- DUP050055207 FIGURE DPF-30 Mmol sui nsmtunoM 'SMaMtNn*gMiQN ____ . _ b mp tmv uoun M\0 r.T'W' OmftrAfAY.t'ot Wmmi4VI*9 OATt FIGURE DPF-31 UKI lOCHtmCAtlpei <fZZffU-tf-Z OanMT Vs b tidoto jf%.9mTKW ArapMaa $t<Jt\ .... MJmcu vu DtsmtunON PrnmUfZWt yet OATf ____ t iw w at u a*__3L. DUP050055208 FIGURE DPF-32 rAsncu stzi wrr*i*unow SAMnitoctmncAnoN Jt3X*SU-tF-S _______. ____________ Omit f-*r V* UQu-/*i0 TK* . , ornliWi' flee Vmcomt ep taMnuii IMv,....... . .. . .................................... OATf ___ t imw u t u m . . -c -86- DUP050055209 sMn|iocNTincATiON 0wr vu l FIGURE DPF-33 rMitai >tzi MmitunoM VXHUnKS KILL - VMUUtX CUMS TOC * ft, -87- DUP050055210 FIGURE DPF-34 %> 0 .6 MICRON - OXIDATION BASE (VIBRATING M ILL - 10 SEC) UJ ID OC < o u UJ CO z< O CD M Z <(0 MO -4 I-- fto- < ID H >X z-t o (0 d> a3E H-J '< ID H Z W ID H 10 < IX CC < mo Mu o a o 0 0) 1 XX. *A 0) ? 2 in a> ID ID -88- DUP050055211 FIGURE DPF-35 R-900 FINISHED PRODUCT SIZE DISTRIBUTIONS -89- DUP050055212 FIGURE DPF-36 PARTICLE SIZE DISTRIBUTIONS OXIDATION BASE, R-900 PRODUCT FREQUENCY -90- DUP050055213 FIGURE DPF-37 OXIDATION BASE PARTICLE SIZE EM LINE II VARIABLE CBU PARTICLE SIZE DISTRIBUTIONS SEDIGRAPH DATA VARIABLE OXIDATION LINE, CBU -91- DUP050055214 -92- OXIDATION BASE EVALUATION -BATE T /H DUP050055215 TABLE DPF-23 OXIDATION BASE WET GRINDING SANDHILL TIME (KIN) 0 5 15 COARSE TAIL <%>0,6 MICRONS) DEAGGLOM BASE AGGLOM BASE 32 43 20 29 13 23 -93- DUP050055216 -94- CALCULATED DUP050055217 RUTILE TI02 HIDING POWE EFFECT OF REFRACTIVE INDEX -95* I DUP050055218 96- VEHICLE REFRACTIVE INDEX DUP050055219 OPTIMUM RUTILE DIAMETE VARIABLE VEHICLE REFRACTIVE IND -97- DUP050055220 uw<h m m ii CO * i ft. Q W g wO J*. me d iu m r e f r a c t iv e in d e x -98- DUP050055221 FIGURE DPF-44 VEHICLE REFRACTIVE INDEX -99- CD H IV CM DUP050055222 MIE SCATTERING CALCULATIONS 1.514 MEDIUM RI co EH + t--t c m m > I tz PS K tt! Dc O I CO oCO in -sf PS W s> t-3 0} -100- DUP050055223 PARTICLE DIAMETER - MICROMETERS hW g :h hj H-1 Yo-- <u j O> O'CO CK U ZD QUO DZg (0 2 COLL CD w ck r fs U> (M FIGURE D P F-46 - | ------------------------ .6 0 0 -IT -06 t r i ii i i i i i i i i i rii i T--l---- CM caBivircnvo a/a) -aNoiasaNn -101- DUP050055224 PARTICLE DIAMETER - MICROMETERS o J-- CO CKH- N "J* UJ1 I fa fa Q O < Ll I I-- cj > CKCOU lx) Q QUO H Z*-< '* fa DEoI COCK coil. C3 w O' 8 N r(1 I'rrrr IT! i rp t 1 n i l i ii i ii m iTT-rr LD LD LD nnN (QElVnnDTVD d/a) oj 3N0d30Nn -102- DUP050055225 PARTICLE DIAMETER VS. CBU CDPF BEST F IT S ) PARTICLE DIAMETER - MICROMETERS pri i i i i t i i|i 00 00 0 ID CM ii(iiii 0 naa -103*- T-rr U) rrr 00 0 0 DUP050055226 (d/a ) 3N01cd30Nn -104- DUP050055227 7 a/a 7 3Noia30Nn -105- FIGURE D P F-50 PARTICLE DIAMETER v o \ UNDERTONE ( TEN DIAM ETER, CBU UT ) PARTICLE DIAMETER - MICROMETERS <s DUP050055228 C A T T E R IN G BY S P H E R E S OF D IA M E T E R CO 1 E LO Ll I QC LU oc in uJmf j 9 d s -106- DUP050055229 z o z> CD QC O QC O N in & Wo g m UJ (O in fic O E a * co CN MEDIAN DIAMETER wrf J9d s -107- O DUP050055230 05 \ Cn . fO *' 5 si s< ^ Vi fie .V o i E a % * D IA M E T E R CO in i fa s M g U H uirt J9d s o -108- DUP050055231 -109- SCATTERING AS F(REFRACTIVE IND 0.20 DIAMETER RUTILE PARTICLE REFRACTIVE INDEX DUP050055232 AJ O in in fa S Ui 0: r. i ? >u l > 3 * D IA M E T E R UlT1 J9d $ -110- DUP050055233 SCATTERING AS F( REFRACTIVE IND 0.20 DIAMETER RUTILE PARTICLE -S M 2/G M lD^ON iH *"H fN H eo ^OO>qi>tt>iD *H O O O O O -111- id \o ia ID ID 00 REFRACTIVE INDEX DUP050055234 SCATTERING AS F(REFRACTIVE IND 0.20 DIAMETER RUTILE PARTICLE -112- CO lO f- lO < CD % a o 10 in m in in co co W ow 0 DUP050055235 -113- FIGURE D P F-58 8o* 05 JUN u > CL DUP050055236 -114- ENAMEL REFRACTIVE INDEX VS. P ROSS DATA REFRACTIVE INDEX o oq to iq co DUP050055237 -116- -T I0 2 SCATTERING DATA ROSS DATA 0 .2 0 DIANE' PIGMENT VOLUME DISTRIBUTION i DUP050055239  o cm dm cm m J<imADSC DSaCDK 2Z~Kv-< HU fri H K wH jwc m m < t<s o<s Q 05 05 Ok (M Ok U [4 UBKfaHUNM* HBQ||| M OIOEESX 2UECEG oooooo E (5 U o u o n!!I mi?? T I0 2 SCATTERING DATA- ROSS DATA 0 .2 0 DIAMETER PIGMENT VOLUME DISTRIBUTION -117- DUP050055240 TABLE DPF-24 ROSS/BRUEHLMAN (R&D TI02) - 1969 TI02 S VALUE (MEASURED) -.0.52 MA2/GM HIE SCAT MAX S FOR 1.514 REFRACTIVE INDEX - 0,785 MA2/GM 0.785 -0.520 0.265 ' 0.265/0.785*100 - 34% LOSS IN S (HP) KNOWN - 15% HP LOSS FOR COMMERCIAL PRIMARY PARTICLE SIZE DISTRIBUTION (VS. OPTIMUM MONOSIZE) - 5% HP LOSS FOR AGGREGATION .UNKNOWN - 14% REMAINING UNACCOUNTED FOR NOTE: NO SIGNIFICANT HP INCREASE HAS BEEN OBSERVED DURING THE LAST 25 YEARS (LOW PVC SYSTEMS) -118- DUP050055241 TABLE DPF-25 SPOONER/CAIRNS (TECHNICAL SERVICE COLORS) - 1975 TI02 S VALUE (MEASURED) - 0.70 M*2/GM HIE SCAT MAX S POR 1,48 REFRACTIVE INDEX - 0.885 MA2/GM 0.885 . -0.700 0.185 0.185/0.885*100 * 21% LOSS IN S (HP) KNOWN - 15% HP LOSS FOR COMMERCIAL PRIMARY PARTICLE SIZE DISTRIBUTION (VS. OPTIMUM MONOSIZE) - 5% HP LOSS FOR AGGREGATION 20% OR ESSENTIALLY ALL HP ACCOUNTED FOR -119 DUP050055242 'iron o f r ef r ac t ic s c al c o iaiio is as IfRDfT 'BDEX OF REFRACTIOH CiLCUttnOK* urn: b e e PEE *SBGLE [1] OE BMIPLE [2] aiCOUTMHS' BPDItt SELECT CASE 21 CASE 1 SOTO SDIC CASE 2 GOTO KILT APPENDIX ATT-1 "Ttjnx/?EF>, B/fj " SBG: DTO *PISG5T BOH OF REFMCTIOK'i IF UPm VEHICLE BDEX OF REFRACTIOH'; IV 1*(IP/IV) B*P-W I * (A * 2 - 1} I {A * 2 + 2J IAPPRDX * .4 * B IfRBT PIOBT BDEX OF IEFRACIIOK, IP * IP IfRBT VEHICLE BDEX OF REFRACTION, IV * KV br u it b comm = ; I IfRIHT KAPPROX - *; r ppr o x mm i ssmo (0i) * *?* 2. mm 1 SQUARED (APPROX) = *; IAPPR0X 2 mm omm DIAIEIER FOR BUIE LIGHT - (0. HCROES)1; ; FI1(.45 1 1000 / (4.143 * IV * I)) / 1000; ECHOES* BEE OPTDtUK DIAKTZR Ft* GREEK LIGHT - (0.56 ECROHS)*; ; FIX(.56 * 1000 / (4.443 * IV * I))/ 1000;' ECROHS1 IfRBT OPTO DIAXETER FOR RED LKBT - (0.55 ECROHS)';* ; FIX(.59 1000 / (4.443 * IV * I)) / 1000; * ECROHS* IfRIHT; IfRIHT UPOT "HUH AHOBER CAmiTIOR? - I/l*; TR$ IF D$ 'I* TBEH GOTO SIHG ELSE EKD BOLT: IfRIHT CHRS(27); CERS(IO); CHR$(115); CSK$(50); CSR$(4S); CHR$(72) EDTH BRIE 160 BRUT PIOtEHT*, VEHICLE', , V, 1*, 1*2', 1`2', *D Opt', "D opt*, *0 Opt', *IDIP M`2" IfRIHT 'BDEX*, 'BOB', , CALC*, *APPR0X*, 'CALC', 'APPROX*, BLUE', 'OH*, RED', 'APPROX VS. CALC* LPRBT ! IfRBT SPOT *PlGES7 BDEX OF REFRACTIOH START*; IPS! BPDT PlGKEIil BDEX OF REFRACTIOH ETO\ HPE! *mmmm Bd ex o f r ef r ac t io n bc r eh eh t *; ip ii BPDT VEHICLE BDEX OF REFRACTIOH START*; 195! BPDT TESaE BDEX OF REFRACTIOH ERD*; IVEI BPUI *VEECLE BDEX OT REFRACTIOH BCRZKEHT'; R7I! FOR IP! = IPS! TO HPE! SIB El! FOR IV! = HVS! TO KVE! STEP HVI! A! - (HP! /IV!) B! IP! - IV! E * (A! * 2 -1) / (A! \2 + 2) BAPPROX! = .4 B1 I! * (FIX(.45 / (4.443 * IV! * E) * 1000)) /1000 T! = (FB(.56 /(4.443 * IV! * I!) 1000)) / 1000 E! * (FII(.59 / (4.443 ` IV! * I!) 1000)) / 1000 BEIT IP!, IV!, , FIX(M! 1000) / 1000, FB(IBPR0X! * 1000) / 1000, FB(E * 2 * 1000) / 1000, FIX(IAPPROl! * 2 * 1000) / 1000, 1!, I!, I!, ((FB(E * 2 * 10 XEXTIV1 BRUT 1EXT IP! -120-* DUP050055243 non o f tmmcM aumms naow non o f r ef r ic t io k , ip 2.75 TEEICLE non OF REFRACTOR, KV = 1.5 I OUXUTED * .4401145 X APPROX * .5 8 SQDSURED (C1LC) * .1939649 1 SQUARED (APPROX) * .25 flpnXOK DIMETER FOR BIDE LIGHT * (0.45 COOKS) ORTHO DIMETER FOR GREEK LIGHT (0.54 COOKS) OPTO DUKE! FOR RED LICET - (0.59 COOKS) .151 COOKS .10 COOKS ,201 COOKS APPENDIX ATT-2 Ou t p u t f &<>m "nJbxteh, b a j -121- DUP050055244 INDEX OF REFRACTION CALCULATIONS PIGKEKT BDEX TEHICLE BDEX 2.55 1 2.55 14 2.55 1.2 2.55 1.3 2.55 1.4 2.55 1.5 2.55 1.6 2.55 1.7 2.5 1 2.5 1.1 2.5 1.2 2.6 1.3 2.6 1.4 2.6 1.5 2.6 1.6 2.5 1.7 2.65 1 2.65 1.1 2.65 1.2 2.65 1.3 2.65 1.4 2.65 1.5 2.65 1.6 2.65 1.7 2.7 1 2.7 14 2.7 1.2 2.7 1.3 2.7 1.4 2,7 1.5 2.7 1.6 2,7 1.7 2.75 1 2.75 14 2.75 1.2 2.75 1.1 2.75 1.4 2.75 1.5 2.75 1.6 2,75 1.7 E cue .647 .593 .539 .466 .435 .386 .339 .294 .657 .604 .551 .499 .449 .4 .353 ' .306 .667 .615 .563 .512 .462 .414 .367 .322 .677 .626 .575 .524 .475 .427 .381 .336 .686 .636 .586 .536 .487 .44 .394 .35 K UPRQX 1*2 cue .62 .418 .579 .351 .539 .291 .499 .237 .459 .189 .419 .149 .379 .115 .339 .086 .639 .432 .599 .365 .559 .304 ,519 .249 .479 .201 .439 .16 .399 .124 459 .095 .659 .445 .619 .378 .579 ,317 .539 .262 .499 .214 .459 .171 .419 .135 .379 ,104 .679 .458 .639 .392 .599 .33 .559 .275 .519 .226 .479 .182 .439 .145 .399 .113 .699 .47 .659 .404 .619 .343 .579 .288 .539 .238 .499 .193 ,459 .155 .419 .122 APPENDIX ATT-3 toTflUT FtO\ vZ*tbXjtZFl-ftAS 1*2 wm .384 .336 .291 .249 .211 ,176 .144 .115 .409 .359 .313 .27 .23 .193 .159 .129 .435 .384 .336 .291 .249 .211 .176 .144 .462 .409 .359 ,313 .27 .23 .193 .159 .489 .435 .384 .336 .291 .249 .211 .176 D opt BLUE .156 .155 .156 ,159 .165 .174 .186 .202 .154 .152 .152 .155 .16 .168 ,179 .193 .151 .149 .149 ,151 .156 .163 .172 .184 .149 .147 .146 .148 .152 .157 .166 .176 .147 .144 .143 .145 .148 .153 .16 .17 Dopt D opt m RED .194 . .193 .194 .199 .206 .217 .232 .252 ,191 .189 .19 .193 .2 .209 .222 .24 .188 .186 .186 .189 .194 .202 .214 .229 .186 .182 .182 .184 .189 .196 .206 .22 .183 .18 .179 ,18 .184 .19 .199 .211 .205 .203 .205 .209 .217 .229 .244 .265 .201 .199 .2 .204 .211 .221 .234 .253 .198 496 .196 .199 .205 .213 .225 .241 .196 .192 .192 .194 .199 .207 .217 .232 .193 .189 .188 .19 .194 .201 .21 .223 8DIFI*2 HMVS. cue 8.133971 4.271504 0 -5,061291 -11,64021 18.12081 -25.21739 -33.7209! 5.324074 1.643836 -2.960526 -8.433735 -14.42786 -20.625 28.22581 -35,7894? 2,247191 -1.587302 -5.993691 -11,0687 -16.35514 -23.39181 -30.37037 -38.46154 -.8733624 -4.336735 -8.787879 -13.81818 -19.46903 -26.37363 -33,10345 -40,70797 4,042553 -7.673267 -11,95335 16.66667 -22.26891 29.01554 -36.12903 -44.26229 -122- DUP050055245 am SBUcr Tisnoi mas s es = \ U im mmm furnm size * *, xi/ _d po ! pxmctE s u e m-m *\12/ lmn UWWH-MS' 01/*.073 02/ = .1533 OH * .392 al * 1553.1826/ al * 1.(2 C3/* .22591 iipmB(mBLDEDmnLEiiK*saWB , sb$ nm *ira Em BATA ?IU JttB - SUT/G , S6$ l*m toEODXlXPlBIlB-SCAT/R \ SK$ OPZH SB$ Me OOTPOT AS /I ul*U IP XX/ < Dll SSf GOTO I U XX/ <* D3| JHEN 6010 XX IF U| > 03/ IBS GOTO Ml X: 00 BILE XX/<01/ s/*xx/A3*a/ m/i,XA/,s/ U|`U|IU| urn >*i2i mm m XX; DO M XX/ <=03/ S/ = til * (Eff(([(ia(Ai/ / 02/)| / .53] * 2] * -.5)] TOTE /l, AX/, S/ XX/ * XX/ + XI/ IT AX/ >* 12/ GOTO AAAA LOOP XXX: do ranxx/ >03/ SI * 01/1*1 am /i, xx/, s/ al*at ml IF XX/ >* X2/ GCttO XXXX MOP XXXX: CMSE/1 `Irani nttVZLEHGTO - 550m1 01/ = .1 02/ = .21 03/ * .4 til * 375.3732 C2/ E 1 al * .19103 OPEN SG$ K ESnm AS 12 ui xz I? XX/ < 01/ KB GOTO B IF XX/ <= D3/ TBEK GOTO BB IP XX/> 03/ Kffl G0T0 BBB B* DO SHUEIX/ < 01/ s/*xx/`3*ci/ ms 12, ul.sl XA/*AA/+X1| APPENDIX ATT-4 -123- *UMlsctt7T (* DUP050055246 17 AAf >- A2f GOTO SBBB m Jgj D0*IBiI2 AA/ D3f SI * C2# * (EXP((((L0G(AA| / 02/)} / .53) * 2) * -.5)) KITE /2, AA/, Si All IP All >* A2/ GOTO SBBB MOP BBBs D0 7HU2ii/> 03# S|*CJ|/AA| ware #2, Aif, sf Ail *AA! f All IP iA/ >= A2| GOTO BBBB MOP CLOSE }2 'LPBM? ffiVILHIOT - eoon' Dll = .107 D2| = .2217 mi * .420 CL! * 265.0502 C2I = .065 01 = .17681 OPEX SK$ PCS OOTPOT AS |3 Ail * A| If AA| < Dll flEK SOTO C IP AAl <= 03/ THEE GOTO CC IF M# > D3/ THEN GOTO CCC C: DO SHIM AA| < 01/ Si = AA/ * 3 ` Cll tm 13, AAi, SI AAl * AAl 1 All IP AA| >-12} GOTO CCCC MOP CC: DO SHOE AAl < 03/ Si * / * |E((([MG(AA| / 02#)) / ,53) ` 2) ` -.5)) OKIE |3, AA/, S| AAf = AA/ + All IF AA| >* A2/ GOTO CCCC MOP CCC; DO 0002 AAl > D3| S| = C3/ / AA/ ware #3, aa/, $| AAl * AA/ * 41/ IF AA| >= A2/ GOTO CCCC MM CCCC: CLOSE 13 2HD -124- DUP050055247 CLS MM MU Mill Mil Mtlf urn *ed toss h e saimuc c h o t ao t s - dp? pk x ms kr isc at .w s, phsc ht i.ms * IMX? HP0T USKE 0? BIDE SCATTERDiG DMA FEE *, SS1$ nm nwtE o? rara saraiKG Dm f il e *,ss2$ HPOI VIE OF RED SaTOMG m PILE *, SS3$ DPEH SS1$ FOR DIM IS ll (mat $s# m min is to OPEE SS3{ POR DOW *S |3 CLS LPHWSSS LPRIST URDU * BLUE - 445ns ; *; * GREEN - 550na j BED - 60ta' URItt IHHT 'PIETICIi'; ; CilCDIATE)*; * *j PimOE';. ; 'OinJIATED'; * ; PMTICLE,i * IfHJT DIMBTER'i *i 'SOOTtlHG"; *; *DIAXTSt*;' ; 'SCATTERING*;' 'DIAffiTER'; APPENDIX ATT-5 ''M7Sc*cr/, Bfis li *j CiimWED' *i SCmERHG' WOT'--* ; * ; *-------- DO 8EILE (EOT E0F(1)) u f o t li, mal mrt 11, VEDE2/ DBOT #2, VAL0E3/ ih pd i 12, mon ism u, val ues * ism 13, mml traur o sdig pm$; v al uei/; v al ue2/; v al o ej /; wmh mmi; mmt LOOP UEHT CSR$(12) -125- DUP050055248  BOSS EE SCHJffiDtG OLCDIATIOHS - DPT BOOMS KRlSOT.ffiS, KKCin.BlS BL5E - 445m mm OLCDUTED warn SCATTERING 0.000 0.0000 0.010 0.0016 0.020 0.0125 0.030 0.0422 0.000 0.1000 0.050 0.1954 0.0(0 0.3375 0.070 0.53(2 0.080 0.7630 0.090 0.9778 0.100 1.1705 0.110 1.3315 0.120 1.45(0 0.130 1.(435 0.140 1.59(4 0.150 1.(186 0.1(0 1.(147 0.170 1.5895 0.180 1.5473 0.190 1.4925 0.200 1.4284 0.210 1.3582 0.220 1.2842 0.230 1.2086 0.240 1.1329 0.250 1.0583 0.2(0 0.9858 0.270 0.9159 0.280 0.8492 0.290 0.7859 0.300 0.7530 0.310 0.7287 0.320 0.70(0 0.330 0.(846 0.340 0.6644 0.350 0.(455 0.3(0 0.(275 0.370 0.(105 0.380 0.5945 0.390 0.5793 0.400 0.5648 0.410 0.5510 0.420 0.5379 0.430 0.5254 0.440 0.5134 0.450 0.5020 0.4(0 0.4911 0.470 0,4807 0.48Q 0.4705 0.490 0.4(10 0.500 0.4518 (BEER 550m flKIClE OlOTATED CIAXETER samm 0.000 0.010 0.020 0.030 0.040 0.050 0.0(0 0.070 0.080 0.090 o.ioo 0.110 0.120 0.130 0.140 ' 0.150 0.1(0 0.170 0.180 0.190 0.200 0.210 0.220 0.230 0.240 0.250 0.2(0 0.270 0.280 0.290 0.300 0.310 0.320 0.330 0.340 0.350 0.3(0 0.370 0,380 0.390 0.400 0.410 0.420 0.430 0.440 0.450 0.4(0 0.470 0.480 0.490 0.500 0.0000 0.0004 0.0030 0.0101 0.0240 0.04(9 0.0811 0.1288 0.1922 0.2735 0.3754 0.4751 0.5727 0.5(41 0.74(3 0.8175 0.87(7 0.9236 0.9586 0.9823 0.9958 1.0000 0.99(2 0.9854 0.9(88 0.9473 0.9220 0.8937 0.8(30 0.8307 0.7974 0,7(34 0.7292 0.(951 0.5(15 0.(285 0.59(2 0.5(49 0.5347 0.5055 0.4776 0.4(59 0.4548 0.4443 0.43(2 0.4245 0.4153 0.40(4 0.3980 0.3899 0.3821 RED* (00m P1OTCLE C81C0UTED DIAXETER sanERBc 0.000 0.010 0.020 0.030 0.040 0.050 . 0.060 0.070 0.080 0.090 0.100 0.110 0.120 0.130 0.140 0.150 0.1(0 0.170 0.180 0.190 0.200 0.210 0.220 0.230 0.240 0.250 0.2(0 0.270 0.280 0.290 0.309 0.310 0.320 0.330 0.340 0.350 0.3(0 0.370 0.380 0.390 0.400 0.410 0.420 0.430 0.440 0.450 0.4(0 0.470 0.480 0.490 0.500 0.0000 0.0003 0.0021 0.0072 0.0170 0.0331 0.0571 0.0909 0.1357 0.1932 0.2(51 0.3488 0.4294 0.5076 0.5807 0.64(8 0.7045 0.7531 0.7925 0.8227 0.8444 0.8580 0.8(43 0.8(42 0.8583 0.8476 0.8328 0.8146 0.7936 0.7704 0.7455 0.7194 0.(924 0.(650 0.(374 0.(098 0.5825 0.5556 0.5292 0.5035 0.4765 0.4544 0.4311 0.4112 0.4018 0.3929 0.3844 0,37(2 0.3(84 0.3(08 0.353( -126- APFENDIX . ATT-6 tOfTpar F&o ^ ''4/pbCctrA P/?K&Tm DUP050055249 0.510 0.4(10 0.520 0.4144 0.510 0.4202 0.540 0.4184 0.550 0.4107 0.500 0.4014 0.570 0.1901 0.580 0.1895 0.590 0.1829 0.510 0.2740 0.520 0.2074 0.520 0.1004 0.540 0.2518 0.550 0.2472 0.500 0.2411 0.570 0.2251 0.580 0.2294 0.590 0.2228 0.510 0.2407 0.520 0.2400 0.520 0.2230 0.540 0.3274 0.550 0.3215 0.500 0.3157 0.570 0.3102 0.580 0.3048 0.590 0.2997 -127- DUP050055250 JACKSON LABORATORY DUPONT CHEMICALS RESEARCH AND DEVELOPMENT DIVISION TECHNICAL REPORT (CT-JL-92-00 DISTRIBUTION LIST COPY 1. D. P. FIELDS, SR. RESEARCH ASSOCIATE, AUTHOR 2. M. H. LEWIS, DEVELOPMENT & TECH SERVICE MANAGER 3. A. M. DOYLE, RESEARCH DIRECTOR 4. C. E. LORENZ, RESEARCH VICE PRESIDENT 5. A. H. REID, RESEARCH SUPERVISOR 6. R. L. FREED, SR. RESEARCH SUPERVISOR 7. J. M. STEED, RESEARCH MANAGER 8. P. A. TOOLEY, RESEARCH SUPERVISOR 9. T. W. BOAZ, PRODUCT MANAGER, COATINGS 10. J. P. PRITCHARD, PRODUCT MANAGER, PLASTICS 11. R. A. JOHNS, PRODUCT MANAGER, PAPER CR CR BBJV EM JL DELISLE BBMP B- ABSTRACT ONLY 12 . D. H. REILLEY, VICE PRESIDENT - WP&MP 13 . J. P. JESSON, LABORATORY DIRECTOR 14 . G. E, LYNSKEY, MANUFACTURING DIRECTOR 15 , D. J, GOULD, GROUP COUNCIL, CHEMICALS BJL BM- 16- 17. CENTRAL REPORT INDEX 18- 22. JACKSON LABORATORY FILES 23, T. M. TAYLOR (EDGE MOOR FILES) 24. J. F. MURRILL. TECHNICAL INDEX 25, S. V. R. MASTRANGELO 26. J. H. BRAUN 27. R. E. MARGANSKI 28. R. A. GONZALEZ 29. R. J. BUCHACEK 30. M. R. BALOGA 31. D. P. SCHUSSLER 32. R. L. GOROWARA 33. B, W. SULLIVAN 34, A. BAIDINS 37. C, R. BETTLER 38. D.,A. HOLTZEN 39. B . TECLE 40. J.lD. LEE BMP JL EM EM EM CR JL JL CR JV DELISLE EM JL JL CR CR CR EXP ST 357252 -128- DUP050055251