Determination of Metallic Zinc Content of Inorganic and Organic Zinc-Rich Primers by Differential Scanning Calorimetry Dwight G. Weldon and Bruce M. CarI--KTA-Tator, Inc.*

INTRODUCTION he protective properties of zinc have been well k n o w n since the mid nineteenth century, w h e n hot dipped galvanizing was developed in France. 1 In the 1940s, the feasibility of applying a zinc paint (socalled "cold galvanizing") was demonstrated in Australia w h e n such a product was used on the 250 mile long Morgan-Wyalla Pipeline. The coating applied to the Australian pipeline was cured by baking. Since that time, considerable effort has been expended by paint m a n u facturers to develop zinc rich primers which cure under ambient conditions. There are two broad categories of zinc rich primers currently in use: inorganic and organic zinc rich prim ers. The vehicle in inorganic zinc-rich primers is a silicate, often tetraethylortho silicate. This binder cures by reaction with atmospheric moisture. The binder in organic zinc-rich primers is usually an epoxy resin, although other resins are sometimes used. A separate hardener, or activator component, is used with the epoxy zinc rich primers in order to achieve cure of the low molecular weight epoxy resins. By far, the largest ingredient of a zinc rich primer is p o w d e r e d zinc, or zinc dust. Zinc loadings of 80 to 90% by weight are common. To provide superior corrosion resistance, the zinc content of such primers must exceed the critical pigment volume concentration (CPVC), thus assuring good electrical contact between the zinc particles and the steel substrate. W h e n zinc is in contact with steel, it serves as the anode in an electrochemical cell. The zinc will sacrificially corrode and protect the underlying steel. Figures 1 and 2 are SEM microphotographs showing the high loading of the spherical zinc dust pigment in dried coating films. As the zinc cot rodes, so-called "white rust" forms (primarily basic zinc carbonate and zinc hydroxide) which fills in the pore structure of the coating, and, together with the remain ing zinc, acts as a barrier coating. 2 Since the primary protective mechanism of a zinc-rich primer is dependent on the electrical contact between the metallic zinc particles and the underlying steel, it is desirable to have an analytical method which can deter-

T

"115 Technology Dr., Pittsburgh, PA 15275.

several methods exist for determinA l t h o u g h ing the metallic zinc content of the zinc dust pigment in inorganic and organic zinc rich primers, no generally recognized method exists for the analysis of the metallic zinc content in the dried films of such coatings. This paper demonstrates a rapid and precise method, based on

di//erential sca i g c Iori, et , /or m ki g this determination. mine the metallic zinc content of a dried film. Although several methods exist for analyzing virgin zinc dust, including combustion, 3 oxidation reduction titrations, 4-6 and hydrogen evolution] no standard method exists for the analysis of metallic zinc in dried coating fihns. This paper demonstrates a simple, rapid method of deter mining the metallic zinc content in pigment and dried paint films by measuring the apparent heat of fusion of the sample from 415 428~ using differential scanning calorimetry (DSC), and compares the results with those obtained by the hydrogen evolution method referenced by ASTM 7 as the referee method.

EXPERIMENTAL A Perkin Elmer Model DSC6 differential scanning calo rimeter, operated under a nitrogen purge and cooled with a Neslab RTE 140 refrigerated circulator was used. The sample was weighed with a Perkin Elmer Model AD4 microbalance and placed in a standard a l u m i n u m weighing pan, which was then crimped. The calibration standard was high purity zinc foil (Perkin Elmer Part #023190036). The sulfuric acid and ferrous sulfate used in the hydrogen evolution method were reagent grade, obtained from Fisher Scientific. The platinum sheet, approximately 25 m m x 25 ram, was also obtained from Fisher Scien tific.

Vol. 69, No. 868, May 1997

45

D.G. Weldon and B.M. Carl

Figure I--SEM microphotograph at 600X of virgin pigment for Porganic zinc-rich primer,

SEMmicrophotographs were obtained with an RJ Lee scanning electron microscope. The coatings used in this study consisted of an ethyl silicate inorganic zinc-rich primer and an epoxy polyamide zinc-rich primer; both of which are standard commercial products. The manufacturer's product data sheets listed a percent total zinc (not metallic zinc) in the dry film of 86% + 2% for the inorganic zinc-rich primer, and 85% for the epoxy zinc-rich primer. The mixing of coating samples and the preparation of dried films were performed with the aid of an Ohaus model E400D top loading balance, accurate to +lmg. In b oth cases, the zinc dust pigment was supplied in a separate container, and subsequently combined at a specific ratio with the liquid component. Two liquid components were provided with the epoxy zinc-rich primer: one contained the epoxy resin, while the other contained the polyamide hardener. Prior to analyzing the dried films, the zinc dust pigmerit was analyzed as received. This was performed by both the ASTM hydrogen evolution method recomm e n d e d as a referee method7 and also by DSC. The hydrogen evolution method has been described elsewhom. 7 Briefly, it involves reacting a known mass of pigment, typically one gram, with 1:1 sulfuric acid, and measuring the volume of evolved hydrogen by water displacement. A small sheet of platinum is used to catalyze the reaction: ferrous sulfate is used to coagulate the sample to reducg floating of it through the apparatus: and the water is saturated with hydrogen gas prior to introducing it into the displacement tube, to minimize absorption of the evolved hydrogen. After thoroughly shaking the container of dry pigment, a small amount of pigment was removed for analy-

46

Journal of Coatings Technology

I Figure 2--SEM microphotograph at dOOX of inorganic ~no-rich primer, showing spherical zinc dust pigm en t emb edde d in silicate m atrix, sis by DSC. A spatula was used to transfer generally between four and eight milligrams of sample into a standard aluminum DSC pan. An aluminum lid was then placed on the pan, and crimped. The crimped alum i n u m pan containing the sample was then placed on the sample cell of the DSC, while an empty crimped aluminum pan was placed on the reference cell. The DSC was previously calibrated using zinc foil reference material. Both the calibration standard and samples were analyzed in a single dynamic step ranging from 370 to 440~ at 10~ per minute, under a nitrogen purge. The analysis produced a very straight baseline, with a relatively sharp endothermic peak near 419~ due to the molting of the metallic zinc (see Figures 3 and 4). The area under the endothermic transition, in J/& was measured by using the peak area software provided with the Perkin Elmer DSC6. The percentage of metallic zinc in the sample was then calculated by dividing the sample's apparent heat of fusion (the area under the endothermic transition, in J/g), by the value for pure zinc (108 J/g). The two pigment samples were re-analyzed at least five times each to obtain an average and standard deviation. One of the problems involved in evaluating any analytical method is the preparation of standards. Since dry film standards of this type do not exist commercially, they were prepared in the laboratory. This was performed by prg-weighing an appropriate amount of zinc dust pigment onto weighing paper, and then using a 10cc disposable syringe to quickly weigh an appropriate amount of liquid component (or pre-mixed liquid componentin the case of the epoxy zinc) into a 100cc plastic beaker. The pre-weighed pigment was then rapidly stirred into the liquid component. A small portion (approximately two grams) of the liquid paint was then poured in a bead across the top of a Ere-weighed thin

Determination of Metallic Zinc Content Table l--Analysis of Virgin Zinc Dust Pigment by Hydrogen Evolution Method and by g s c Sample

Metallic Zn, H2

Metallic Zn, DSC

Pigment, inorganic ............................... 96.5% Pigment, epoxy ..................................... 94.2%

% Recover~

98.3% 93.6%

Stnd. Dev., DSC

102% 99%

2.3 b 2.1 ~

(a) % recovery or the DSC Method Relative to the Hydrogen Evolution Method (b) Based on five determinations (c) Based on seven determinetions.

gauge 75 m m x 125 m m cold rolled steel coupon, and drawn down to obtain a wet film thickness of approximately 75 125 microns. Within approximately five sec ands of performing the drawdown, the panel was weighed again to determine the weight of wet paint actually applied to the panel. Various panels were prepared in the previously men tioned manner, and allowed to air dry for at least four days at laboratory ambient conditions of approximately 21 to 24~ and 30 to 50% relative humidity. The panels were then weighed to obtain the weight of the dried paint films. By knowing the percentage of pigment in the wet paint, the weight of wet paint applied to the panels, and the corresponding weight of dried paint, the percent pigment in the dried paint could be calculated. The percent metallic zinc in the dried paint was then calculated by using the values obtained by the hydrogen evolution method for the virgin zinc dust pigment. Drawdowns were made at the mix ratios specified by the paint manufacturers and also by using approximately half of the recommended pigment. Additional draw downs were made by substituting zinc oxide for a portion of the zinc dust pigment. This was done to demonstrate the specificity of the DSC method towards metal lic zinc. Once the standards had been prepared as described earlier, an X acto knife was used to remove the dried coating down to bare steel, and these portions were briefly ground in a mortar and pestal. They were then analyzed in triplicate using the same DSC technique as

described previously for the virgin pigment, and also by the hydrogen evolution method.

RESULTS AND DISCUSSION The results of the analysis of the two virgin zinc dust pigment samples are shown in Table 1. Good agreement was obtained between the hydrogen evolution method and the DSC method. The DSC method produced results that were 102% and 99% those of the hydrogen evolution method, with a standard deviation of approxi mately 2%. An analysis by DSC required approximately 10 rain, versus 1-1/2 to 2 hr by hydrogen evolution. The results from the dried films are shown in Table 2. The average percent recovery for all six samples (epoxy zinc and inorganic zinc, each at three mix ratios) was 105% by the DSC method, based on the calculated percent pigment in the dry film, and using the percent metallic zinc in the virgin pigment obtained by the hy drogen evolution method. Therefore, the error of approximately 5% cannot be attributed to the DSC method, but rather is the combined error of all aspects of the experimental protocol, including the analysis of the virgin pigment by hydrogen evolution, the preparation of the "standard" dry film specimens, and the DSC analy sis. Indeed, there will be a natural high bias in the results due in part to the evaporation of solvent during the mixing and drawing-down of the coating standards, estimated at approximately one percent by conducting

h~

I -5

e4 ~d

25aotsq c.54~

45-

I ' ~ ' " ]~ ~

,

.

aTo

I

..

~j

" ~66 " ~

...

" ~6

' ~ 6 " d~

,~

&~

~

" 45

"~

' "~

4~

" 4~;

Figure 3--DSC heating curve of virgin zinc dust pigment.

]~" ]~' ~"

~

~

~

:4Q ~4~[''~ "

'4~" ] ~ " '~

t~per~

' "

P

t

Figure 4--DSC heating curve of cured film of epoxy zinc-rich primer, corresponding t o sample A in Table 2. Vol. 69, No. 868, M a y 1997

"c

P 47

D.G. Weldon

and B.M. Carl

Table 2--Analysis of Dried Coating Films Sample

Type

A ............................ Epoxy B ............................. Epoxy C ............................ Epoxy D ............................ Inorganic E ............................. Inorganic F ............................. Inorganic

Zn ~ c a l c u l a l e d ~

Zn ~ by DSC d

Recovery

Zn~ b y H2

Recovery

77.6% 66.3% 64 ..5%b 78.4% 67.2% 70.9% ~

81.0% 72.0% 68.1% 82.3% 69.1% 75.2%

104% 108% 105% 105% 103% 106%

N/A s N/A ~ N/A s 83.3%

N/A s N/A ~ N/A s 106% 107% 106%

72.0% 75.0%

(a) Using metallic Zn (Zne) content of virgin pigment from H2 method. (b) Also contained 11.8%ZnO in dried film (c) Also contained 74% ZnO in dried film. (d) Average of three determinations Standard deviations from 1.0 to 2.7. (e) Not applicable to epoxy zinc due to excessive carryover of particles into a9169

weight loss measurements. Furthermore, at least some hydrogen gas would have been absorbed in the tubing or the hydrogen "saturated" water of the hydrogen evolution apparatus, which would bias the results by generating a low value for the calculated metallic zinc content. The results obtained by the h y d r o g e n evolution method for the dried films of inorganic zinc-rich primer were very close to those obtained by the DSC method, but required approximately 2-1/2 hr to complete. However, the hydrogen evolution method was not applicable to the dried films of epoxy zinc rich primer. Gas evolu tion was slower, and significant quantities of sample were repeatedly carried into the transfer tube of the apparatus. It was estimated that recoveries were 50 to 60%. Both the DSC method and the hydrogen evolution method were specific to metallic zinc, and were not influenced by the presence of zinc oxide. This was, of course, expected, but was important to verify, since zinc oxide would be an expected impurity in the zinc dust.

CONCLUSIONS An accurate and precise method for determining the metallic zinc content has been demonstrated for cured inorganic and organic zinc rich primer films. The method is also applicable to the virgin zinc dust pigment. The method measures the apparent heat of fusion of the sample, and compares the measured value to the known value of 108 J / g for pure zinc. This provides a simple and more basic way of determining the percentage of metallic zinc, and is unaffected by the presence of zinc oxidation products. Several different types of samples were analyzed, including those where some of the zinc dust pigment was replaced with zinc oxide, and percent recoveries were within approximately five percent of the calculated metallic zinc values for the dried coating films. Since this value is affected by the entire experimental protocol, the error of the actual DSC method is less than five percent. When applied to two samples of virgin pigment, the DSC method had recoveries of 99 and 102%, compared to the values obtained by the ASTM hydrogen evolution referee method. The method has several advantages over the hydrogen evolution method. One major advantage of the method is its applicability to dried paint films, rather 48

Journal of Coatings T e c h n o l o g y

than just the virgin zinc powder. The scope of the hydro gen evolution method as described in ASTM D 521 is for the analysis of zinc-dust pigment, not dried coating films. Although the work performed in this study shows that the hydrogen evolution method produces results of corn parable accuracy to the DSC method when applied to an inorganic zinc-rich primer, the method did not work when applied to an epoxy zinc rich primer. In contrast, the DSC method produced results of similar accuracy, whether applied to an inorganic or an organic zinc-rich primer. A second advantage of the DSC method is speed. Determinations conducted using the hydrogen evolution method took approximately 1-1/2 hr for the zinc dust pigment samples, and approximately 2 1/2 hr for dried coating films. In contrast, a single DSC analysis was conducted in approximately 10 rain, including the weighing and actual running of the sample. Therefore, an analysis could be conducted in triplicate by the DSC method in only a fraction of the time that it takes to do a single analysis by the hydrogen evolution method. A third advantage of the DSC method is its small sample size. Whereas the hydrogen evolution method requires approximately one gram of sample, the DSC method requires only a few milligrams. Although sample size is certainly not a problem when the analysis is being performed on batches of pigment, it can be a major factor if the purpose of the investigation is related to problems or deficiencies encountered in field applied coatings. In the latter case, it is not uncommon for the analyst to be provided with only a few square centimeters of sample, which may only have 20 to 50 microns of zinc rich primer on the backside. During the course of these experiments, it was found that the accuracy of the results improved if the calorim eter was calibrated daily with high purity zinc foil, and that the calibration standard be used only one time. Reagent grade zinc granules or zinc powder were of insufficient purity to properly calibrate the instrument, and using a zinc foil standard more than one time also resulted in inaccurate results, due to oxidation of the zinc at the high temperature in the DSC, coupled with the alloying effects of zinc metal with the aluminum pans. Other factors associated with obtaining accurate re sults included very gentle tapping of the pan once the sample was added in order to distribute it relatively

Determination of Metallic Zinc Content e v e n l y over the b o t t o m of the pan, a n d careful place m e n t of the p a n lid. A thin, e v e n d i s t r i b u t i o n of material p r o v i d e s better t h e r m a l c o n t a c t w i t h the b o t t o m of the pan, a n d the lid m u s t be carefully placed o n the p a n in o r d e r to a v o i d e x p u l s i o n of the fine p o w d e r d u r i n g crimping.

ACKNOWLEDGMENT The a u t h o r s w o u l d like to t h a n k Rick H u n t l e y for his h e l p f u l c o m m e n t s , a n d Jackie Sheerer for her p a t i e n t help in the p r e p a r a t i o n of this m a n u s c r i p t .

References (1) Munger, C.G., J. Protective Coat. Linings, June (1989). (2) Feliu, S., Baraias, R., Bastidas, J.M., and Morcillo, M., "Mecha nism of Cathodic Protection of Zinc-Rich Paints by Electrochemical Impedance Spectroscopy. I Galvanic Stage," JOURNALOFCOAT INGSTBCHNOLOGu61, No. 775, 63 (1989). (3) Fresenius, ZeitschriJ~Ffir Analische Chemie, 17, 1978. (4) Drewsen, Zeitschrift F~r Analische Chemie, 19, 1880. (5) Topf, Zei~schrijt FfirAnalische Chemie, 16, 1887. (6) ASTM 521-90, "Standard Methods for Chemical Analysis of Zinc Dust (Metallic Zinc Powder)," American Society for Testing and Materials, Philadelphia, PA, 1993. (7) Wilson, LA., Proceedings ASTEA, American Society for Testing and Materials, Philadelphia, PA, Vol. 18, Part II, 1918, p. 220.

DWIGHT G. WELDON received a B.S. Degree in Chemistry from the University of Michigan and an M.S. Degree in Chemistry from Michigan Technological University. He also did doctoral level research in Raman spectroscopy at the University of Kentucky. Prior to joining KTA-Tator, inc. as Laboratory Director in 1982, Mr. Weldon had been a Formulations Chemist for both Cook Paint and Varnish and lnmont Corporation. in addition to directing numerous projects involving the physical and analytical testing of paints and coatings, he specializes in performing failure analysis investigations of coatings. Mr. Weidon is a member of the Federation of Societies for Coatings Technology, the American Chemical Society, the Steel Structures Painting Council, and the Society for Analytical Chemists of Pittsburgh. BRUCE M. CARL received a B.A. Degree in Chemistry from Lycoming College. He has been a Chemical Technician with KTATator, inc. since 1988 where he has performed numerous analytical and physical tests, including, ASTM, NACE, and Federal Test Methods on a variety of paints and coatings.

Vol. 69, No. 868, May 1997

49