Corrosion inhibition by naturally occurringsubstances—I. The effect of Hibiscus subdariffa (karkade) extract on the dissolution of Al and Zn

Corrosion Science, 1972, Vol. 12, pp. 897 to 904. Pergamon Press. Printed in Great Britain

CORROSION INHIBITION BY NATURALLY OCCURRING SUBSTANCES--I. THE EFFECT OF H I B I S C U S S U B D A R I F F A (KARKADE) EXTRACT ON THE DISSOLUTION OF A1 AND Zn* A. A. EL HOSARY, R. M . SALEH a n d A. M . SHAMS EL DiN Laboratory of Electrochemistry and Corrosion, National Research Centre, Dokki, Cairo, Egypt

Abstract--The thermometric, the weight-loss and the galvanostatic polarization techniques were used to establish the inhibition of the dissolution of AI and Zn in HCI and NaOH by different concentrations of aqueous extract of Hibiscus subdariffa (Karkade). The extent of corrosion inhibition as measured by the three techniques is comparable. The results indicated that the additive acts by way of adsorption on both cathodic and anodic corrosion areas. Curves representing the variation of the reaction number (R.N.), in thermometric experiments, and the decrease in weight as a function of the concentration of the additive, are invariably sigmoid in nature. When present in enough amounts, the additive decreases the dissolution rate by as much as 85 per cent of the value recorded in its absence. The two main constituents of the aqueous extracts of Hibiscus subdariffa, namely the organic acids and the colouring materials were separated and tested independently for surface activity. Both constituents were effective in retarding the dissolution of the two metals; but the activity of the colouring portion was considerably higher than that of the organic acids. R,~sum6---On a us6 de thermom6trie, de gravim6trie et de polarisation intensiostatique pour examiner l'action inhibitrice d'extraits aqueux d'Hibiscus subdariffa (Karkade) sur la dissolution de Al et de Zn dans HCI et NaOH. Ces trois techniques ont conduit gt des r6sultats concordants. Ceux-ci indiquent que radditif se comporte comme un inhibiteur d'adsorption anodique et cathodique. Les courbes de variation du hombre de r6action (NR) (essais thermom6triques) et de pertes de poids en fonction de la concentration de l'additif sont invariablement sigmoides. La vitesse de dissolution peut ~.tre r6duite fi 15% de sa valeur par des additions suffisantes de I'additif. Les deux constituants essentiels des extraits aqueux d'Hibiscus subdariffa: acides organiques et mati6res colorantes, ont 6t6 s6par6s et 6prouv6s ind6pendamment. Tous deux se sont av6r~,s efficaces, mats ~t cet 6gard, la partie colorante I'emporte nettement sur la partie acides organiques. Zusammenfasstmg--Die thermometrische, die Gewichtsverlust- und die galvanostatische Polarisationsmethode wurden verwendet, um die Hemmung der Aufl6sung yon Al und Zn in HCI und NaOH durch verschiedene Konzentrationen eines w~.ssrigen Extrakts von Hibiscus subdariffa (Karkade) zu untersuchen. Das Ausmass der Korrosionshemmung, gemessen nach den drei Methoden, ist vergleichbar. Die Ergebnisse zeigten, dass der Zusatz infolge Adsorption sowohl an katodischen als auch anodischen Korrosionsgebieten seine Wirkung entfaltet. Kurven, welche die ,~nderung der Reaktionszahl bet thermometrischen Versuchen und die Gewichtsabnahme als Funktion der Konzentration des Zusatzes darstellen, sind stets S-Kurven. Wenn der Zusatz in ausreichenden Mengen vorhanden ist, setzt der Zusatz die Aufl/Ssungsgeschwindigkeit bis zu 85% des Wertes herab, der in seiner Abwesenheit festgestellt wird. Die beiden Hauptbestandteile der w~ssrigen Extrakte yon Hibiscus subdariffa, niimlich die organischen Siiuren und die f~.rbenden Substanzen, wurden getrennt und unabh/ingig auf Oberfl/ichenaktivitat gepriift. Beide Bestandteile bewirkten eine Verz/Sgerung der Aufl/Ssung der zwei Metalle, aber die Aktivit~.t der Farbsubstanzen war betrichtlich h/Sher als die der organischen S~,uren. *Manuscript received 17 March 1972. 897

898

A . A . EL HOSARY,R. M. SALEH and A. M. SHAMSEL D ~ INTRODUCTION

NATURAL products of plants or animal origin exhibit in general high surface activity. This is well demonstrated in polarography where small additions of gelatin, camphor or albumin are commonly used as maximum-suppressors. 1 In larger amounts these substances, as well as many others, nfight cause the distortion or the complete elimination of the polarographic wave. ~,2 This is related to their adsorption on the surface of the DME, ~ where they interfere with homogeneous electron or heterogeneous proton transfer processes. 4 Since corrosion is basically an electrochemical reaction, it is expected that the same substances will exert a retarding effect on the corrosion rate. Information is in fact available on the application of one natural product or the other as a restrainer in pickling, descaling or polishing processes, s A systematic evaluation of these materials has not, however, been undertaken. It is the object of this series to direct the attention to the high inhibiting properties of aqueous extracts of some of the common plants. The use of these extracts offers two main advantages. First, they can be easily obtained from cheap plants and/or by-products (peels, seeds, etc.) remaining in the canning industry. Second, since these extracts always contain more than one organic product, corrosion inhibition is fortified through synergism. The present paper reports on the retardation of the dissolution of A1 and Zn in HCI and N a O H by the aqueous extract of Hibiscus subdariffa (Karkade or Roselle). Three independent techniques were used to evaluate the inhibition efficiency of the extract. These were: (a), the thermometric ;B (b), the weight-loss and; (c), the galvanostatic polarization methods. The results obtained from the three procedures were comparable and showed that in the presence of enough of the additive--the attack on the two metals could be lowered to ca. 15 per cent of the value measured in retardant-free media. EXPERIMENTAL

The analysis of the dried petals of Hibiscus subdariffa showed it to consist of: 7-1° 3"5 proteins, 10 liquids, 63-5 glucides, 11 cellulose and 12 per cent ash (Fe, Mn, AI, Ca, Mg, Na, K, SO4, PO4 and CI). The glucides are: 22 organic acids, 16 reducing sugars and 25 per cent other non-nitrogenous substances. The water-soluble acids consist of: 77 per cent citric acid, 22 per cent malic acid and traces of tartaric acid. The colouring matter contained in the flowers of H. subdariffa is a mixture of hibiscitrin and g o s s y t r i n , v-9.n-xs The two compounds are the glycosides of hibiscetin (3 : 5 : 7 : 8 : 3' : 4' : 5'-heptahydroxyflavone), (I) and of gossypetin (3 : 5 : 7 : 8 : 3' : 4'-hexahydroxyflavone) (II), respectively. The glycosidic bond is at the 3-position. The dried drug contains also about 40-50 mg ascorbic acid (vitamin C)/100 g.lS-lD A stock solution of the inhibitor was prepared by extracting 10.0 g of the dried petals 5 times with 250 ml hot distilled water. The collected extract was faltered and portions thereof were titrated pH-metricaUy with standard N a O H . The free acids neutralized along a single step. The concentration of the stock solution was expressed in terms of the normality of the free acids. The effect of both the colouring material and the free acids of the karkade extracts on the dissolution rates of the two metals was tested separately. The colouring material was isolated as follows: A known volume of the stock solution was neutralized

Corrosion inhibition by naturally occurring substances--I

899

with N a O H till p H ~ 6. The formed citrate and malate salts were precipitated with BaCI2 and filtered off. The volume of the filtrate was readjusted to the original value by evaporation on the water bath. On the other hand, decolorization of ttie solution was effected through adsorption on active charcoal. Analysis of the decolorized solution revealed that the treatment removed simultaneously between 20-30 per cent of the acid content, depending on the amount of charcoal used and the time of treatment. The new formalities were used in the calculations. The karkade stock solution was kept in the ice chest. Over a period of 3 m, neither the composition (as gleaned from the titration of the acid) nor the inhibition efficiency of the solution was affected. Other solutions, prepared during the same period, showed also the same efficiency, when their normalities were comparable. This last finding indicates that the ratios of the various components of the extract are always constant, as long as the same plant variety is used. In the thermometric corrosion test, °,20,21 pieces of the two metals measuring 1 X 10 cm were used. These were degreased and cleaned as previously described, zl Experiments were carried out with 15 ml of 2N HCI or I-5N N a O H , and the starting temperature was 25 °. The weight-loss experiments were performed with metal test pieces of the same dimensions as in the thermometric technique. These were suspended in 50 ml of the corroding solutions to which increasing quantities of the inhibitor were added. The loss in weight after 60 rain of immersion was determined in duplicates and the results were averaged. In both the thermometric and the weight-loss experiments, the test metal piece was used only once, and new portions of the fresh solutions were always employed. The dissolution rate of Zn in 1.5N N a O H was too slow to be accurately determined by the thermometric or the weighr-loss techniques. For this reason the inhibition efficiency of karkade extracts was established by the galvanostatic polarization technique. A B.D.H. Zn rod of a total surface area of 3-7 cm z was anodically polarized in 1-0N N a O H solution free from and containing varying amounts of the additive. A constant polarizing current of 80 mA/electrode was used in this series of experiments. Before anodization, the working electrode was first subjected to a 15 min cathodic pre-treatment process with the same polarizing current to be later employed, in order to reduce any oxide present on the surface of the electrode. Experiments were carried out at 25 °, and the potential of the working electrode was measured relative to a saturated calomel half cell. The polarization cell and the circuit used in tracing galvanostatic oxidation curves have already been described. 2z RESULTS Curves O of Fig. 1 are the characteristic thermometric corrosion curves of AI in 2N HCI (Fig. la) and in 1-5N N a O H (Fig. lb). The maximum temperature measured in the acid solution is 80 °, and is attained after 15 rain. This corresponds to a reaction number (R.N.) of 3-67°/rain. The dissolution of the metal in the basic solution is somewhat slower. T,, amounts to 85 °, but is reached only after 32-5 min. The R.N. is 1-85°/min. When to both corrosive media definite quantities of the aqueous extracts of karkade were added, the other curves of Fig. 1 were obtained. As is clear, the additive interferes with the dissolution reaction and causes the R.N. to decrease below that determined in additive-free media. The percentage decrease in R.N., i.e.

900

A . A . EL Hos~dtY, R.. M. SALEH and A. M. SHAMS EL Dir~

80

0.3

7%1.0

o

o

(a)

~

~ OlOl~

...,..

(b)

L o°6%~1~l'.<~ p~°'5

~3. 6O

o

I--

40

20O

20

x

4C1

I

I

80

Time ,

1

i

0

I

I

40

I

x

O0

rain

FIG. 1. Effect of karkade extract on the temperature-time curves of AJ in 2N HCI (a) and in 1.5N NaOH (b). Numbers ort the curves are concentrations of the additive.

(a)

-

40

/

(b)

\

20

~

..2

l

3

.c 1

.c

~ s0

80

I -5'0

I -4'0

I -3.0

I -2.0

1

I

-l-O

-5-0

I ~4-0

I -3-0

I -2"0

Log C,N FIG. 2. Inhibition efficiency of karkade extracts on the dissolution of AI in 2N HCI (a) and 1.5N N a O H (b). ( a ) ~ ( l ) karkade extract (R..N.), (2) karkade extract (weightloss), (3) decolourized extract (R.N.), (4) colouring material (l~N.). (b)--(l) karkade extract (R.N.), (2) colouring material (K.N.).

I

120

C o r r o s i o n inhibition by naturally occurring s u b s t a n c e s - - I

901

0 l'-

~O~o

0-1

",,,%

(a)

(b) 3

~

Z

4O "0 6O 20

r

0

I

t

I0

N

I

I

20

Time ,

-

I

4 "0

I

- 3 "0

rnin

-I'0

-Z'O

Log C,N

FIG, 3. (a)--Effect o f k a r k a d e extract on the t e m p e r a t u r e - t i m e curves of Zn in 3N HCI. N u m b e r s on curves are concentrations of the additive_ (b)--Percentage reduction in R.N. (1) and weight-loss (2) of Zn in 3 N HCI with the concentration of karkade extract.

2.0

-

\ L..,..

(a)

1"6

12 _

o

0-8

A

0-4

o

in 0 -

0

_

> LIJ

?:,

6

~n O

~

0

O

-0-4

o c

-0"8

u

(b)

40

60 -I,2

-I,6

"~~! I 0

v

80

I

i

15

Time,

I 30

rain

t

I 45

t

I 60

I

I

-3.0 Log

-2"0

-I-O

C,N

FIG. 4. (a)---Effect of karkade extract additions on the galvanostatic polarization curves of Zn in I N H a O H at 21.62 rnA/cm I. N u m b e r s on the curves are concentrations of the additive. (b)--Percentage variation of Q with the concentration of karkade extract (see text).

902

A.A. EL HOSARY,R. M. SALEI--Iand A. M. SHAMSEL DIN

100 × (R.N.)rr,~ -- (R.N.)inhibitor/(R.N.)rrce, is plotted in Fig. 2 as a function of log Cinhibitor-

Curves having the same shape as those of Fig. 1 were obtained when the decolorized - - o r the acid-free coloured extracts were used instead of the original aqueous infusion. The variation of the percentage reduction in R.N. with the logarithm of the concentration of the active components of the additive is similarly drawn in Fig. 2. In Fig. 3(a) curves are given representing the effect of different concentrations of the karkade extract on the dissolution of Zn in 3N HCI as determined by the thermometric technique. The extent of corrosion inhibition, in per cent, is given by curve 1, Fig. 3(b) as a function of log C i n h i b i t o rThe loss in the weight of AI strips in 2N HCI and of Zn strips in 3N HCI in the absence and presence of increasing amounts of karkade extracts was determined over a period of 60 min at room temperature. The inhibition efficiency of the additive was computed according to 100 × Wfr~c - - WiahJbitor/Wfrec, where W is the loss in weight/ test piece (20 cmZ). The variation of the inhibition efficiency with the logarithm of the normality of the additive is drawn in Figs. 2(a) and 3(b). The galvanostatic polarization of the Zn electrode in 1N N a O H solutions containing various additions of the karkade extracts is represented by the curves of Fig, 4(a). A constant current of 80 mA/electrode was employed. The curves exhibit one oxidation arrest, the length of which decreases with the increase in the quantity of the additive in solution. The percentage decrease in time, 100 × tfree -- tinhibitor/tfree, is plottcd as a function of log C i n h i b i t o r in Fig. 4(b). DISCUSSION Karkade extract acts as an efficient restrainer to the dissolution of A1 and Zn in acid and alkaline solutions. This is well demonstrated by the results obtained from a variety of different experimental techniques. Thus, for example, the additive reduces the R.N. of the two metals in the thermometric corrosion tests (Figs. 1 and 3). The effect depends on both the metal used and the aggressive medium. In the case of Al in the acid solution, the maximum temperature, Tin, is little affected, while the time, t, needed to reach it progressively increases as the concentration of the additive is raised' Fig. I(a). These are common features of weak adsorption. 2° The prolongz.tion of t is mainly confined to the initial segment at the foot of the curves. This time was defined as the incubation period, 21.zs during which the oxide film originally present on the surface of the metal protects it from the acid attack. Accordingly, one can safely conclude that the additive acts in the acid solution as a cathodic inhibitor, zl Supporting this conclusion is the fact that the slope of the rising parts of the temperaturetime curves (corresponding to the actual dissolution of the metal) is practically the same and independent of the inhibitor concentration. In N a O H , on the other hand, the incubation period is largely absent. It is replaced by art induction period a3 representing the breakdown of the film and the start of the attack. The rate of the temperature-rise during this stage is higher than that of the incubation period. When present in enough concentration, the karkade extract intereferes with the dissolution of the metal and lowers the R.N. This is produced mainly through a lowering in T= and a corresponding increase in t. The first effect indicates strong adsorption of the additive, z° Further, in contrast to its behaviour in

Corrosion inhibition by naturally occurring substances--I

903

HCI solution, the restrainer appears to act predominantly as an anodic inhibitor. This is readily gleaned from the fact that the slope of the rising parts of the temperaturetime curves progressively decreases with the increase in the bulk concentration of the additive. The thermometric curves representing the dissolution of Zn in HCI, Fig. 3, differ from those for A1 in that the incubation and the induction periods are absent. The temperature of the system rises--almost linearly--from the moment the metal is allowed to react with the aggressive electrolyte. The air formed oxide on the Zn metal ---in contrast to that on Al--does not offer any resistance to the acid attack. From the curves obtained in the presence of increasing concentrations of the inhibitor it is concluded that the additive is strongly adsorbed on the anodic sites of the metal. In Figs. 2 and 3 curves are drawn representing the variation of the percentage decrease in R.N. as a function of log ~ i b i t o r for the two metals in HCI and NaOH. These are invariably sigmoid in nature, substantiating the idea that the karkade extracts reduce the corrosion by way of adsorption. The turn in the curves is rather sharp, suggesting that it is associated with the formation of a monolayer of the inhibitor on the surface of the corroding metal. In the case of A1 in the acid solution, the inflexion in the curve occurs around 10-aN. In N a O H , on the other hand, the same takes place at ~ 5 × 10-4N. The smaller value in the latter medium is in accordance with the above given conclusion, namely, that the additive in the basic solutions adsorbs on the anodic areas. Since these on A1 are small compared to the "cathodic sites, the quantity of the restrainer necessary to stifle the dissolution to the same extent will be smaller in N a O H than in HCI. It is of interest to note also that the extent of dissolution retardation by karkade before the inflexion in N a O H exceeds that determined in HCI under the same conditions. This also substantiates the conclusion that the additive is more strongly adsorbed in the former medium than in the latter. When present in enough amounts the additive reduces the extent of dissolution of AI in both solutions to less than 20 per cent of the value measured in additive-free media. The effect of karkade extracts--as corrosion restrainer--on the dissolution of AI and Zn in acid solutions was also established by the loss in weight method. The results are expressed in terms of percentage reduction in the extent of dissolution, and are given by curve 2 of Figs. 2(a) and 3(b). The agreement between these curves and those determined by the thermometric method is more than accidental, and strongly recommends the latter as a rapid and simple technique for the evaluation of inhibition properties of additives. In view of the facts that the karkade extract contains a small amount of free acids (mainly citric artd malic), and that these are expected to exert a retarding effect on the dissolution of AI and Zn, 24 it appeared of interest to establish whether the high inhibition efficiency of the crude extract is due to the free acids or to the red colouring material (hibiscitrin and gossytrin). The two components were therefore separated as described in the Experimental and each was tested for inhibition. The thermometric procedure was used, mainly with A1 in HCI. The percentage decrease in R.N. by both additive components is represented by curves 3 and 4 of Fig. 2(a). The two parts reduce the dissolution rate, but the effect produced by the colouring material greatly overweighs that of the acids, and approaches the one produced by the original extract. In

904

A . A . EL HOSARY, R. M. SALEH and A. M. SHAMSEL DIN

NaOH, on the other hand, the effect of the colouring material similates to a great extent that of the main infusion. This is to be expected since the extent of inhibition by organic acid anions in basic media is limited. 24 The dissolution of Zn in 1 and 1.SN N a O H was too slow to be followed by the thermometric method. For this reason the effect of the karkade extract on the corrosion of the metal was examined by the galvanostatic polarization technique. The curve O of Fig. 4 was obtained when the Zn electrode was anodically polarized with a constant current density of 21.62 m A / c m ~ in IN NaOH. This exhibits a clear oxidation step at -- 1.4 V (vs SCE) due to the formation of ZnO or Zn(OH)2 on the electrode surface. 2s-29 At the end of this stage, the potential of the working electrode changed abruptly towards the value characteristic for the evolution of O2 on the passive electrode. In the presence of increasing quantities of the additive in solution, the oxideformation step progressively decreased in length. This is most probably the result of the blocking of part of the electrode surface by the inhibitor. In agreement with this conclusion is the fact that the plot of the percentage decrease in the quantity of electricity, Q, consumed along the oxidation step, Fig. 4(b), has the same sigmoid form as those determined by the weight-loss or the thermometric corrosion tests. As is the case with these last techniques, the extent of anodic corrosion can be lowered to about 20 per cent of its value in the presence of enough of the additive in solution. REFERENCES 1. J. H~.vRovsrY and J. KtrrA, Principles ofPolarography, p, 299. Publishing House of the Czechoslovak Academy of Sciences, Prague (1965). 2. S. G. M.AIRANOVS~I, Catalytic and Kinetic Waves in Polarography (Eng.'Trans.), p. 93. Plenum Press, New York (1968). 3. C. N. REILLEV and W. STUMM, Progress in Polarography, Vol. I, p. 81 (Ed. P. ZtrMkrq and I. M. KoLraOrr). Interscieace, New York (1962). 4. B. KASTE~qINOand L. HOLLECK, J. electroanal. Chem. 27, 355 (1970). 5. U. R. EVANS, The Corrosion and Oxidation o f Metals, p. 178. Edward Arnold, London (1960). 6. F. MVLn~S, Z. Metallic. 14, 233 (1922). 7. T. A. G~S~tAN, The Chemistry of Flavonoid Compounds, pp. 13,345, 428, 431. Pergamon Press, Oxford (1962). 8. G. REAUaOtmG and R. H. MO~qSEAUX,J. Pharm. Chim. 1, 392 (1940); C.A. 35, 8209 (1941). 9. S. COPERI"INI, ,4gricottura Colon. 30, 182 (1936); C..4.30, 6505 (1936). 10. G. BUOGO and D. RICCHINENNA,Ann. Chim. applicata 27, 577 (1937); C.A. 32, 5951 (1938). 11. P. S. RAO and T. R. S~HADRL Proc. Indian Acad. Sci. 15A, 148 (1942); 16A, 323 (1942). 12. P. R. RAO, P. S. RAo and T. R. SESHADRLProc. Indian Acad. Sci. 19A, 88 (1944); 27A, 104 (1948). 13. A. G. PERKI~q,J. chem. Soc. 95, 1855 (1909); 103, 650 (1913). 14. W. BAg.ER, R. NODZU and R. RomrqSON, J. chem. Soc. 74, 74 (1929). 15. R. YAMAMOTOand T. OSIMA, Sci. Pap. Inst. phys. Chem. Res. Tokyo 19, 134 (1932); C.A. 27, 521 (1933). 16. G. LORENZI~, Archo. Ist. biochem, ital 9, 123 (1937); C.A. 31, 8601 (1937). 17. N. M ~ . r r o , Archo. 1st. biochem, ital9, 383 (1937); C.A. 32, 6279 (1938). 18. G. LA BRUTO and E. BRurqo, Annali Chim. 50, 1357 (1960); C.A. 55, 6611 (1961). 19. C. S. TtmG, Chung Kuo Nung Yeh Hun Hsneh Hui Chih 1-2 (1964); C.A. 61, 16437 (1964). 20. K. Azlz and A. M. SHAMS EL DIN, Corros. Sci. 5, 489 (1965). 21. J. M. AnD EL KADER and A. M. SHAMSEL DIN, Corros. Sci 10, 551 (1970). 22. A. M. SHAMSEL DIN and F. M. AnD EL WAHAB, Electrochim. Acta 9, 113 0964). 23. M. A. STRrdCHER, Trans. electrochem. Soc. 93, 285 (1948). 24. R. M. SALEH and A. M. SHAMSEL DIN, Corros. Sci. 12, 689 (1972). 25. K. HUlaER, Heir. Chim. Acta 26, 1037 (1943). 26. T. P. DntzsE, J. electrochem. Soc. 102, 497 0955). 27. S. E. S. EL WAKKAD, A. M. SHAMSEL DIN and H. KOTa, J. electrochem. Soc. 105, 47 (1948). 28. M. A. V. DEVENATHANand S. LAKSHMANAN,Electrochim. Acre 13, 667 (1968).