Document g25opo4Y8NaNQxDngwRDBKnbN

CONTAINS CONFIDENTIAL INFORMATION. RESPONSIBILITY OF RECIPIENT. WHEN NO LONGER NEEDED, DESTROY BY BURNING OR SHREDDING. Acc. Ho. CDP-ES-79-23 Issued 12/11/79 Copy No. Qjtg Distribution on last page. EXPERIMENTAL STATION RESEARCH AND DEVELOPMENT DIVISION TECHNICAL REPORT CHEMICALS, DYES AND PIGMENTS DEPARTMENT ________E. I. DU PONT PE NEMOURS & COMPANY DURABILITY TESTING Work Done By: Approved By: Patent Situation Approved By: Previous Related Reports: Project Code: Type Technical Work: Period Covered: Notebook Nos. Personnel: J. H. Braun K, K. Bhatia J. W. Heberling Date: 9/27/79 PTD-SA-78-2B 7053-184270/185272 IEB 8/1/78 - 9/30/79 (part time) E-17871, pp. 12,28,34,46,48,68,74 82,94,98,106,112,126 E-19615, p. 42 J. H, Braun, R. W. Blair ABSTRACT Durability of Ti02 pigments is measured by outdoor exposure of paint panels for chalk/fade performance. The testing takes up to 3 years, and new durable pigments are not commercialized until such data is in hand. Accelerated testing in a Weatherometer takes about 3 months, but has been considered unreliable and thought to have little correlation with outdoor results. Accelerated tests have thus been used only to a limited extent for laboratory guidance. Statistical analyses of more than 1100 paints, however, show a good correlation between outdoor and accelerated chalk/fade data. The accelerated data show larger experimental errors than outdoor results, but this can be compensated for by multiple panels. Accelerated exposures can thus be used to reduce lead time in pro ceeding with projects for commercialization of new grades and process improvements, Two supplementary techniques - a UV reactivity test and electronmicroscopy of leached silica shells - were developed for quick lab guidance and to diagnose production problems, N40791 2 CDP-ES-79-23 I. INTRODUCTION As a basis of business decisions and for product develop ment, we should have a precise, accurate and quick test for Ti02 durability* Available techniques fall short of these requirements because of: intrinsic limitations imposed by the manifestations of "durability", experimental difficulties in testing, and seeming failure of different durability measures to correlate. we can: detect substantial durability differences reliably by outdoor exposure but it takes up to 3 years less reliably and in 3 months in the Weatherometer, but considered not to correlate with outdoors, measure quickly, characteristics relevant to but not identical with degradation, ^ We cannot detect minor durability differences reliably and quickly enough for most business needs. IX. OBJECTIVES The study had three objectives: Review the intrinsic limits of chalk/fade testing in the light of current theory, Prove or disprove the value of accelerated tests for business decisions, and Develop methods for lab guidance. Ill. SUMMARY AND CONCLUSIONS Certain testing problems of "durability" are inherent in its concept, which allows neither its quantitative definition nor its precise measurement. DUP050059272 "3- CDP-ES-79-23 A comparison of outdoor and accelerated chalk/fade performance ratings of rutile pigments shows: Florida and Weatherometer chalk/fade ratings do correlate. Experimental errors of the individual measurements rather than lack of correlation account for most of the data scatter. Neither outdoor nor accelerated chalk/fade ratings repro duce well, one-third of all experimental data points differ from "true" by more than 3.1 and 4,4 chalk/fade units for Florida and Weatherometer exposures, respectively. Accelerated results can be used to shorten lead time for business decisions if experimental and statistical techniques are combined to anticipate exterior performance. Two supplementary techniques were developed for lab guidance and production assistances A UV reactivity test for quick, precise, though not infallible differentiation of similar pigments, and Electron microscopic examination of leached silica shells to pinpoint process failures, IV. PATENT SITUATION No filing action is contemplated on the contents of this report. V. PROGRAM No further work is contemplated. VI, PUBLICATION STATUS Not intended at this time but publication may be considered if it serves Company interests. VII. SPECIAL SAFETY PRECAUTIONS Dry white lead was handled exclusively in a hood with greater than usual attention to cleanliness and hygiene. Lead is now handled subject to special OSHA regulations, UV source and samples are in an opaque box to protect workers, VIII. ENVIRONMENTAL CONSIDERATIONS No environmental problems are expected from lab work. DUP050059273 4 CDP-ES-79-23 IX. CHEMICALS USED Titanium dioxide, ammonia, glycerine, white lead, hydrochloric acid, sulfuric acid. X. TOXICITY DATA Since 4/1/79 white lead (lead carbonate) is OSHA-regulated with permissible exposure limits of .050 mg/m3 lead averaged over 8 hrs. and an "Action Level" of .030 mg/m3. XI. SPECIAL PRECAUTIONS None other than safety. XII. ASSISTED BY / R. W. Blair; Experimental Station Analytical Lab. XIII. ACKNOWLEDGEMENTS D. A. Holtzen, CD&P Chestnut Run, suggested the use of lead salts for UV reactivity; H. W. Jacobson helped develop the test. L. A. Wierzbowski provided all exposure data for 1200 paints prepared and evaluated by numerous members of CD&P Department staff. We thank M. P. Morse, F&F Marshall Lab; V. L. Bacchetta, Engg. Dept. ESD; W. D. Ross for statistical advice and computations; and H. W. Jacobson, P. G. Schmidt, and L. S. Wilkens for useful discus sions. DUP050059274 5 XIV. TABLE OP CONTENTS CDP-ES-79-23 I. II. III. IV. V. VI. vii. VIII, IX, X. XI. XII. XIII. XIV. XV. XVI. XVII. Introduction ......... Objectives .... . ..... Summary and Conclusions . , . Patent Situation .......................... , Program ........... Publication Status ...... Special Safety Precautions , . Environmental Considerations . Chemicals Used ........ Toxicity Data ....... , Special Precautions . , . . . Assisted By ... ... ..................... Acknowledgements ...... . Table of Contents ...... Experimental and Discussion . A. Durability .... . . . . 1. Mechanism of Chalking 2. Chemistry of Chalking B, Exposure Testing ..... 1, Precision ...... 2. Correlation of Outdoor and Accelerated Exposures ...... 3. Chalk/Fade Prediction C. Supplementary Tests . . . 1. UV Reactivity . . . . a. Procedure .... b. Applications . * . 2. Silica Shells . . . . a. Procedure .... b. Applications . . . References ............................................... Indexing Terms ........ : 2 2 >2 3 3 3 3 m3 4 4 4 4 4 5 .6 6 6 8 10 10 12 15 15 15 20 m 20 . 24 24 24 28 28 DUP050059275 -6- CDP-ES-79-23 XV. EXPERIMENTAL AND DISCUSSION As a basis of business decisions and for product development, we should have a precise, accurate and quick test for Tio2 durability. Available techniques fail short of these requirements because of: intrinsic limitations imposed by the manifestations of "durability", experimental difficulties in testing, and seeming failure of different durability measures to correlate. A. Durability "Durability" is the continuity of decorative and protective performance of paint films and the effects of their components, for us the pigment, under the influence of weathering. Lack of outdoor durability manifests itself primarily as chalking of the Ti02 with concomitant erosion, gloss loss and discoloration. Very durable paint films erode without detectable accumulation of chalk. Considerable progress has been made in understanding chalk ing and is discussed below in terms of pioneering contributions. Emerging theories define the limits of performance measurements, showing that the degradation process lacks any singular accurate and quantitative definition. 1. Mechanisms of Chalking A morphological model of chalking was developed and elegantly confirmed by Kaempf et al of Bayer, A. G1. They described pro foundly variable characteristics of weathered paint films in terms of four combinations: A. Exposed particles of stabilized Ti02 resting on and protecting pedestals of photo-active binder (Fig. 1A). B. Stabilized rutile protecting and degrading photoinactive binder (Fig. IB). C. Holes in a film of photo-active binder degraded directly by UV and catalytically by unstabilized Ti02 (Fig. 1C). D. Unstabilized Ti02 degrading photo-inactive binder. The authors conclude that "Due to these differences in the surface structure of weathered paint films, the individual test methods used to evaluate weather resistance (measurements of chalking, gloss, weight loss) lead to differing, incomparable results". DUP050059276 -7FIGURE X , Exempt* A Exempt* B CDP-ES-79-23 Exempt* C Figure 1--Schematic representation of degradation processes during the vmotlfering of TiO.-pigmeuled binders. Sample A--Stabiliicd rutile pigment in binder of low light stability; Sample D-StabiliinJ rutile pigment in hinder of high light stability; Sample C--Unstubilitcd rutile Or anutase pigment in* binder of high light stability FIGURE 1 * from G Kaerapf et al J- Faint Techn. 46 (1974) 57 _ FIGURE 2 Reflection (Gloss) Degree of Chalking Atill, a, Cfhmevvrtutii und Kunlun# ullvr Vonturhe lut voi#c#e>i.ncn Vorsuchszeilraum von i) Jahmi iur KondiUon I. Exposure time, hrs FIGURE 2 Gloss Loss and Chalking of all Experiments over a Period of 3 Years for Condition I from W. Papenroth et al, DEFAZET 32 (1978) 103 DUP050059277 8 CDP-ES-79-23 Rate effects of weathering have been studied by Dunderdale et al of LaPorte2, They found that pigment volume concentration (PVC) increased during exposure at the paint Surface and enhanced the effectiveness of protection by TiC>2 The degradation sequence of photo-reactive binder has three phases: a period of increasing protection through increasing PVC (decreasing rate' of weight loss) is followed by a steady state (constant rate of weight loss) and then by rapidly diminishing pro tection (increasing rate of weight loss). High PVC improved dura bility of photo-active binders but diminished durability of photo inactive vehicle. 2. Chemistry of Chalking The chemistry of chalking was outlined by Voltz et al of Bayer, A. G.3 showing UV, water and oxygen to be the essential envi ronmental factors in the TiC>2 catalyzed degradation of binder. The authors detected Ti3+, hydroxyl radicals and peroxides and proposed the following reactions, supported by an abundance of data by numer ous other investigators: The sequence continues 1 - 2 - 3 - 1 - 2...., with an overall reaction: TiO- a2 + 2 ------- fiv -------- 0H* + H02*' and hydroxyl and peroxyl radicals oxidizing and degrading the binder. The chain of chalking events is cyclic with respect to Ti02 and can be disrupted at its surface by exclusion of either UV, water or oxygen. Lack of the latter causes photo-graying without signif icant degradation of vehicle. DUP050059278 -9- CDP-ES-79-23 Additional details and verifications of Ti02 solid state and surface involvement in chalking and in particular in photograying were provided by Gulley of CR&D4. His electron spin reson ance and microwave conductivity experiments distinguished readily between rutile and anatase but not among rutiles of different photo activity. Gulley interprets graying as direct photo-reduction and as a portion of the chalk cycle disrupted by the lack of oxygen. We can summarize these and many other contributions to the understanding of chalking by a sequence that shows the multiplicity of possible pathways: 1. The absorption of UV radiation by the paint film initiates events. Rates depend on the relative degrees of photo-activity of binder and of TiOj stabilization. a. In photo-inactive binder, TiOj acts as a catalyst for degradation of vehicle. The product of UV radiation acting on TiC>2 is Ti3+. b. Photo-active binder is degraded in two ways, directly by UV and catalytically by TiC>2. Shielding from UV or passivation of the Ti02 by coatings or by surface substitution of other ions for Ti can disrupt the degradation sequences. 2. Hydroxylation of TiC>2 is essential to its catalytic effect. UV radiation generates hydroxyl radicals. 3. Water is required for re-hydroxylation. Its lack disrupts or retards the sequence. 4. Oxygen re-oxidizes Ti3+ and forms peroxy radicals. Exclusion of oxygen disrupts the sequence and results in photo-graying. 5. Diffusivity of water and oxygen in conventional paint films is involved but rarely rate-1imiting. 6. Hydroxyl and peroxy radicals oxidize and degenerate the organic binder. The general outline of this sequence is well confirmed; details are tentative but not relevant to durability testing. Prom the fore going it follows that: Intrinsic limitations to testing are imposed by the complexity of events involved, which allows no single nor quantitative statement of performance. Experimental problems arise because differences in paint compo sition, particularly choice of binder and PVC, sensitize differ ent pathways of degradation and affect rates selectively. DUP050059279 - 10 CDP-ES-79-23 B. Exposure Testing Paints are exposed outdoors to "natural" weathering and "accelerated" in- a cabinet that combines high radiation and elevated temperatures with water sprays. Outdoor exposures are not entirely natural but accelerated by a factor of perhaps 3 over ordinary weathering at the same loca tion by angle and orientation of the panels. Other parameters include choice of location: desert (very high UV, low moisture), tropical (high UV, high moisture), and industrial (low UV, contamin ants) . Meteorological and seasonal variations add further compli cations, "Accelerated" exposures performed in the controlled environ ment of a laboratory were intended to be more reproducible but are not. Papenroth et al of Bayer6reported from an abundance of data - 2 series of more than 150 repeats of identical exposures of a single pigment in one paint formula in 2 Weatherometer arrangements that "discrepancies of 50% have to be accepted". Figure 2 illus trates this conclusion by some of their data. Our pigments were exposure-tested in Florida and sometimes Delaware and in' the Weatherometer until the notion arose that Florida and Weatherometer results do not correlate. 1. Precision The uncertainties of measurement, both outdoors and accel erated, are minimized by the former Pigments Department's instru mented and computerized rating system7, possibly the most refined one in use. All subsequent conclusions are based on exposures in one paint formula, a blue automotive refinish, Syntex 3833, chalk/ fade rated by Daiger and Madson's procedures. According to the authors, this method suffers a 5-point spread for the 95% confidence level in outdoor exposure. We have evaluated data for both outdoor (Florida) and accelerated (Weatherometer) exposures and find: Exposure _______________________ Standard Deviation of the Test Results for the Same Pigment Exposed in the Same Paint Formula, C/F Units Florida 3.1 Weather-o-Meter 4.4 Standard deviations (Table I) seem to increase with the chalk/ fade ratings, suggesting the use of the coefficient of variation/cr \ in \VJ place of the standard deviation (a) in further computations; but this trend is not firm and we continued with a. DUP050059280 - 11 - TABLE I CDP-BS-79-23 PRECISION OF C/F RATINGS Same Pigments and Formula Different Paintgrlnds and Exposure Series Code Pigment ______. .. Lot # Description C/P Ratings, Standard Deviations ____________ and Data Points, C/P Units; n Florida Weatherometer Mem Std. Data Mean Std, Data PeV. Points Dev, Points R-900 tt R-902 i t** tt R-960 It CLNC OR-65O EXPTL. it tt tt it U083/68U 4479 4995/674 4990/673 4995/721 4926 3048/661 6001/694 TC7302 UC7300 FX9017A FX9017B FX9022A FX9022B FX9022C Shipping Std. 10 C/F Std, Cand. Shipping Std. 32 C/F Std. Shipping Std. Competitive Competitive 9.1 15.8 19.1 20.5 17.0 19.7 32.6 33.0 21.3 26.3 51.0 43.3 33.7 26.0 33.3 7 3.9 2.2 7 1.0 1.0 3.44 1.0 .6 1.2 5.6 3.1 4.2 1.0 2.1 7 8.3 .8 13 I7.3 2.8 7 17.1 3.2 2 18.0 1.4 3 l4.0 3.6 3 17.7 .6 6.0 26.1 5.16 3 31.7 1.5 3 *29.7 1.5 3 *22.7 .6 3 *52.7 8.5 3 *53.0 2.6 3 *35.7 3.1 3 *9.3 5.5 3 *31.7 6.1 7 11 7 2 3 3 58 3 3 3 3 3 3 3 3 POOLED STANDARD DEVIATION 3.1 119 4.4 115 * P&P Exposure Cycle X-l DUP050059281 - 12 - CDP-ES-79-23 Results are based on evaluations of 119 experimental paints rated against 94 paints made up from 2 chalk/fade pigment standards, exposed in 47 series during 1965-1970 both in Florida and in the Atlas Sunshine Carbon Arc Weatherometer, Model XW, with a few f &f Weatherometer cycle X-l points added. 2. Correlation of Outdoor and Accelerated Exposures A comparison of outdoor and accelerated chalk/fade perform ance ratings of rutile pigments revealed: Florida and Weatherometer chalk/fade ratings correlate. Experimental errors of the individual measurements, rather than lack of correlation, account for most of the data scatter. This conclusion is based on evaluation of 1,048 experimental paints made up from production samples of most Ti-Pure grades, competitives and experimental pigments, rated against 83 paints from two chalk/ fade standards exposed in 50 series during 1965-1970 both in Florida and in the Atlas Sunshine Carbon Arc Weatherometer (Model XW). Included in the analyses were all exposure series 65625-68603 for which matching Weatherometer results were found. These data were not edited. The correlation is evident from the data plot (Figure 3). The correlation equation (C/F) Fia - . 4 +1.44 (C/F)Ac c - .02 (C/F)^ accounts for 72% of the data scatter with a residual standard devia tion of 5.8 C/F units. Scatter of 5.4 of these 5.8 C/F units results from the poor precision of the Outdoor and accelerated C/F measure, accounting for 86% of the data scatter, Details of the calculation are shown in Table II. The conclusion that outdoor and accelerated exposure correl ate well is consistent with F&F's broad experience but at variance with opinions held in the former Pigments Department % F&F experience with accelerated weathering in general can be characterized by the statement that "X-l cycle predictions of gloss and color retention of coating exposed in Florida at 45S have been accurate in nearly all instances, where the purpose of the exposure study was to evaluate modifications in the compo sition of the product... (for example) Comparisons of different pigmentations in a product line". % "Removed from F&F units because of lack of correlation with Florida exposure" were panels of an exposure series intended to correlate C/F in Florida with 3 F&F accelerated Weatherometer DUP050059282  DUP050059283 - 14 TABLE II PRECISION OF C/F TESTING Florida, apia Weatherometer, aA. cc Correlation Equation, Oq o t Accounted Data Scatter, Unaccounted Data Scatter, a cc)* ** CDP-ES-79-23 Standard Deviation, C/F Units 3.1 k.k 5.8 5.U 2.2 a, . , = \l<? (Fla, Acc) . |/ Fla ' Acc * o = |/a2 - (a2 + a2 ) Residual y Cor Fla Acc These equations are strictly true only for: =a + b <C/If)Ace and b-1 This condition is approximated in the relevant region of the data (for C/F <ko) DUP050059284 - 15 CDP-.ES "79 -23 Q cycles . Re-examination of these data does not support this conclusion. F&F exposure cycles X-l and X-39 correlated quite well with Florida (Figures 4 & 5). standard deviations cal culated from these data agree with the results reported above. F&F cycle X-41 failed to correlate, because the C/F standards themselves acted peculiarly, grossly distorting all numerical results. 3. Chalk/Fade Prediction From these analyses of correlation and experimental pre cision, we can conclude that Weatherometer data can be used to anticipate outdoor exposure results provided that appropriate sta tistical techniques and criteria are applied: Critical samples must differ by more than 6 C/F units. Multiple panels are exposed to make up for lower precision of accelerated results. An estimate of sample requirements for reliable (95% probability level) differentiation between pigments is given in Table III, based On Students "t" calculations with the standard deviation and correl ation equation Of Table I and Figure 3. C. Supplementary Tests 1. UV Reactivity Test Without hope for a fully satisfactory measure of durability itself, we must guide development work by partial solutions. Comparisons of rutile and anatase pigments show that the largest performance differences are attributable to the photoavailability of Ti3+ sites. We have, therefore, chosen to charac terize pigments by photoreactivity rather than hydroxylation. Discoloration by photo-graying itselflOH is slight, ranging from white to light gray. Direct measurements are therefore subject to Substantial experimental error. The addition of lead salts* as both indicator and reactant allows reduction to continue beyond Ti3+ and extends the scale to almost black, improving sensitivity of the method to a point where it distinguishes clearly between similar pigments. Representative data are shown in Table IV. We attribute the darkening to formation of metallic lead from the TiC>2 and UV catalyzed reduction of lead (II) and oxidation of glycerine (Table V). Suggested by E. A. Holtzen, CD&P, Marketing Division. DUP050059285 -* 16 - FIGURE It uoatharometer c/f rating CDP-ES-79-23 10*31 E6-JUL-7? FIG0R5 5 UEATHEROMETER C/F RATING $3 26 14~AUG-7 CORRELATION r FLORIDA VS.UEATHEROMETER DUP050059286 If two pigments differ hy A C/F units......... 12 10 a 6 5 It 3 - 17 TABLE III CDP-ES-79-23 C/F PREDICTION* it takes at least (n+m) exposures** to show that a difference exists with 95% probability of being right. At Florida (sec = 3.1) In Weatherometer (sec " it. 4) By Anticipation*** (sec = 5*8) 23 2 It 35 59 7 13 10 20 18 It 6 10 15 y* * Based on Students "t" calculation: n X m n +m t X s \ 2* "ith 120 D.F. ' 1-# s = 3.1; lt.it; 5.8 ** For n "experimentals" plus m "standards", with n - m *** Anticipation of Florida from Weatherometer results * > DUP050059287 18 TABLE IV CDP-ES-79-23 UV REACTIVITY AND CtiALK/FADE UV Reactivity rated by shades of gray (Munsell Color Scale): 10 white - no activity 0 = black - saturation activity Pigments R-960 ShSt. 6098 R-994 ShSt. 9749 R-931 ShSt. 6791 R-902 ShSt.' 9436 R-900 ShSt. 4646 R--1Q1 Lot 8554 R-1G0 8939 Cyclone Discharge Reactivity* 8.60** 9.2 8.9 7.8 5.00** 5.1 5.0 5.1 Representative Chalk/Fade Rating 32 31 28 20 10 8 *Standard Deviation jh .6 **Testing Standards DUP050059288 - 19 - TABLE V UV REACTIVITY TEST PROBABLE CHEMISTRY CDP-ES-79-23 VV Ti + hv --------> Ti3+ + (?) 2 + O" --- 10] 2Ti3+ + PB** -------- > 2Ti^+ + PB CHg OH.CiOH.CH^OH + [0] ' -> oxidation products On Exposure to Air: 2Fb + 0^ ........ > 2Fb + 2<f DUP050059289  - 20 CDP-ES-79-23 Dispersions without glycerine or without TiC>2 stay white. Exposure to air re-oxidizes the lead and bleaches the graying. The test responds only to the reduction step in the chain of chalking events. As a consequence, the measure should give false negatives by under-rating pigments that derive their protection from a disruption of the hydroxyl involvement in Chalk/Fade. The rela tive ratings of R-994 and R-960 are probably a case in point. But it is not likely that pigments with an abundance of Ti3+ sites are chalk resistant or pigments with few sites are not durable. The test does, therefore, provide valuable but not infallible guidance, a. Procedure For testing, .50 g pigment is dispersed with 2.50 g white lead and 1.25 g glycerine. The paste is pressed between 2 microscope cover glasses and exposed for 2 hrs. to UV light (GE bulb H-85A3, UV 85 W) at a distance of 4 inches. Chips of the most reactive pigments turn almost black; with least reactive TiO^ they stay nearly white. Graying is rated* against a Munsell scale of neutral grays; 10 White, 0 Black. Ratings are normalized against 2 standards used in every, series; R-960 at 8.60 and R-900 at 5.00. The standard deviation of the test is .6 reactivity units. Graying response to exposure time, Figure 6, (Table VI) shows that time can be used to optimize results for pigment type. Long exposures enhance differences between stabil ized pigments, brief exposures between unstabilized pigments. b. Applications The test is sensitive enough to distinguish between similar products. We can even detect the advantage of process modifications. We tested 3 series of D, P, Fields' sandmilled pigments, and found in agreement with exposure results (Table VII): Advantages for sandmilled over control pigments for all grades tested: R-960, R-902 and R-900. *Instrumental measurement is subject to errors caused by occasional white spots in the chips surrounding small air bubbles. DUP050059290 :;vi' t?' * DUP050059291 uu REACTIVITY RATING -UU PHOTOGRAYING 17871-112 15*12 22 TABLE VI CDP-ES-79-23 UV PHOTOGRAYING* Exposure Tine, Min. 15 30 45 60 90 . 120 180 240 00 ^1 R-100 S fa. S t d. ** .3 7.5 .5 6.5 + .3 5.9 .5 5.8 + . 5 5.6 + .6 4.4 .4 4.0 + .8 R-900 **Sh. Std, 9.1 .4 8.4 .7 8.0 .8 7.5 1.0 6.5 .7 5.7 .5 4.4 .4 4.4 . 6 R-960 **Sh. Std, 9.5 t .0 9.5 + .0 9.5 .0 9.5 t .0 9.5 .0 9.1 Hr .2 8.6 + .4 7.9 .7 R-994 **Sh. Std. 9.5 .0 9.5 .0 9.5 .0 9.5 .0 9.5 .0 9.5 .0 9.3 .3 9.1 . 2 * White 10; Black 0 ** Standard Deviation for 1 paste exposed and rated in 5 chips DUP050059292 - 23 - TABLE VII SANDMILLED PIGMENTS UV REACTIVITY CDP-ES-79-23 Grade R-960* Code 136-A-l 5 136-B-l 2 3 5 Type ... Conventional H Sandmilled First II II II Micronizer Steair^/Pigment 1 5 1 2 3 5 UV Reactivity Normalized to R-960: 8.60 R-900: 5.00 Students *-l-" ** 8.7 1 8.9 / S.8+.6 (n * 10) 9.6 1 ) 4.1 > 2.8 t99%=2-8 9,6 \ 9,6+,4 9.7 j (m - 20) 28d f 9.4 / R-902 135-A-l 5 135-B-l 5 Conventional n Sandmilled First, 1 5 1 5 8.4 I , 8,2 j 9.1 l 8.7 I S.3+.5 (n * 10) 1> 8.9+.6 j! (m m 10) 2.4 >2,1 SsiT2*1 18DF R-900 125-A-l 5 125-B-l 5 Conventional ti Sandmilled -- First f| 1 5 1 5 4.9 l 4.7 | 5.21 6, Of 4.S+.5 (n m 10) > J 5.6+.6 (m * 6) * 2.9 > 2.2 t95%=2,2 14DF * Reported previously * t= ~\/n s V n 4* in s = standard deviation n,m number of exposures of each sample = means DUP050059293 - 24 CDP-ES-79-23 Effects of micronizing intensity (steam/pigment ratio) on reactivity and/ by implication, durability were not detectable, i.e. less than sandmilled versus control or grades versus grades. Results were analyzed statistically and found better than 99 and 95% probable. The reactivity test has been used extensively in devel opment efforts for silica-free, durable pigments and zirconia treatments. 2. Silica Shells A new analytical tool reveals details of the silica shell that imparts durability to R-960. Aggregation effects, coating integrity, and structural details of Her coatings are made visible by a transmission electron micros copy of empty silica shells, left behind after the rutile has been leached out by hot concentrated acidic. The shape of the shells {Figure 7 A to C) documents aggre gation as it existed during silica precipitation. Damage and break age of shells illustrates effects of micronizing. a. Procedure For examination, 50 mg pigment is leached in 10 ml concentrated sulfuric acid at 195C for 5 hrs. The suspen sion is centrifuged repeatedly at 5000 rpm, first with con centrated, then with diluted acid, and finally with water. A drop of ammonia solution in the last rinse neutralizes acid residues. Empty shells are kept wet for direct appli cation to TEM grids. b. Applications Examination of leached silica shells has been used to show that: Glidden's RCL-6 has an Her silica coating. The silica shell of zirconia treated pigments contains some Zr02 in its glass structure because it disinte grates in concentrated acid. Iler coatings are X-ray amorphous. A serendipic R-960 gloss improvement at Johnsonville was related to the state of aggregation or flocculation during the Her treatment. Antimony treatment changes the nature of the pigment surface itself, protecting the particles from acid attack. DUP050059294 25 CDP-ES-79-23 DUP050059295 26 CDP-ES-79-23 * DUP050059296 DUP050059297 - 28 CDP-ES-7 9-2 3 XVI. REFERENCES 1. G. Kaempf et al, J. Paint Techn. (1974) 56. 2. J. Dunderdale et al, FATIPEC 1974, p. 69, 3. H. Volz et al, Farbe u. Lack, 82 (1976) 805, 4. J. E. Gulley, CRD-77-75. 5. Oil, Colours Chem. Assn, Australia, "Surface Coatings" NSWU Press, 1974. 6. W. Papenroth et al, DEFAZET 32 (1978) 102. 7. W. H. Daiger and W. H, Madson, J. Paint Techn. 39 (1967) 399. 8. F&F reports: TS-74-1 and TS-77-6, M. P. Morse "Comparison of Outdoor and Laboratory Weathering Tests for Determin ing the Durability of Coil Coatings" -II" TS-76-4, M. P. Morse "Reliability of Du Pont X-41 Cycle Exposure Tests for Predicting the Florida Durability of Automotive Coatings", 9. Exposure series 69601-12, L. A. Wierzbowski to T, S. Wollenberg, 6/24/71. 10. TFW-328-1 and 324-3. 11. Letter, R. H. Tait to H, S. Jarrett, 2/23/78. 12. H, Weber, FATIPEC XIV (1978) p. 697 XVII. INDEXING TERMS Titanium dioxide Durability Chalk/Fade Exposures Weatherometer DV Reactivity Silica Coatings DUP050059298 - 29 DISTRIBUTION CDP-ES-79-23 1. 2, 3. 4. 5. 6. 7. 8. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. J. H. Braun K. K. Bhatia W. J. Marshall, EM H. B, Clark/R. W. Bess R. A. Darby/A. S. Bjornson/ L. T. Frick/J. G, Ishikawa E. C. Broge/M. A. Toomey/ J. A. Blumberg J. M. Hustler L. N. Fisher/G. A, Hapka Central Report Index, ISD, C-3211 L. A. Wierzbowski, EM W. L. Kremer, EM G. E. Lynskey, JV D, A. Nelson, DeLisle W. J, Lawrence, Antioch D. Nguyen, EM T. B. Scarfe, DeLisle R. L. Heffelfinger, JV R. E. Edwards, JV D. U. Gwost, JV G. H. Senkler, Jr,, EM W. D. Ross H. R, Linton, EM S, V. R. Mastrangelo, EM H. W. Jacobson P. G. Schmidt A. Baidins J. H. Boughton L. A. Monson R. J. Bruehlman, EM 6. D. Gemmell w. E. Stevens, EM C. R. Buchanan, EM H. A. Wildt, EM J. DelPrete, EM 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 4.6, 47, 48. 49-53. A. J. Coombe, JV M. T. Doyle, JV M. R. Baloga, JV G E. Watkins, JV G. A. Ganley, Antioch J. E. Pesek, DeLisle H. J. Malasky, DeLisle D. L, Chandler, DeLisle J. G, Dickinson, Ch. Run V, L. Bacchetta, Louviers b. A. Holtzen, Ch. Run A. Allen, EM c. W, Anderson, J Lab CD&P Information Center, DUP050059299