ARTICLE IN PRESS

1352-2310/$ - se

doi:10.1016/j.at

�Correspondfax: +4686741

E-mail addr

Atmospheric Environment 41 (2007) 1374–1382

www.elsevier.com/locate/atmosenv

Atmospheric corrosion effects of HNO3—Influence oftemperature and relative humidity on laboratory-exposed copper

Farid Samiea,�, Johan Tidblada, Vladimir Kuceraa, Christofer Leygrafb

aCorrosion and Metals Research Institute, Drottning Kristinas vag 48, SE-114 28 Stockholm, SwedenbRoyal Institute of Technology, Division of Corrosion Science, Drottning Kristinas vag 51, SE-100 44 Stockholm, Sweden

Received 26 April 2006; received in revised form 4 October 2006; accepted 10 October 2006

Abstract

The effect of HNO3 on the atmospheric corrosion of copper has been investigated at varied temperature (15–35 1C) and

relative humidity (0–85% RH). Fourier transform infrared (FT-IR) spectroscopy and X-ray diffraction (XRD) confirmed

the existence of cuprite and gerhardtite as the two main corrosion products on the exposed copper surface. For

determination of the corrosion rate and for estimation of the deposition velocity (Vd) of HNO3 on copper, gravimetry and

ion chromatography has been employed. Temperature had a low effect on the corrosion of copper. A minor decrease in the

mass gain was observed as the temperature was increased to 35 1C, possibly as an effect of lower amount of cuprite due to a

thinner adlayer on the metal surface at 35 1C. The Vd of HNO3 on copper, however, was unaffected by temperature. The

corrosion rate and Vd of HNO3 on copper was the lowest at 0% RH, i. e. dry condition, and increased considerably when

changing to 40% RH. A maximum was reached at 65% RH and the mass gain remained constant when the RH was

increased to 85% RH. The Vd of HNO3 on copper at X65% RH, 25 1C and 0.03 cm s�1 air velocity was as high as

0.1570.03 cm s�1 to be compared with the value obtained for an ideal absorbent, 0.1970.02 cm s�1. At sub-ppm levels of

HNO3, the corrosion rate of copper decreased after 14 d and the growth of the oxide levelled off after 7 d of exposure.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Nitric acid; Material degradation; Deposition velocity; Gerhardtite; Cuprite

1. Introduction

Air pollutants together with moisture are keycontributors to atmospheric corrosion, and theircorrosivity depends largely on their surface reactiv-ity and the deposition velocity on the material. Assoon as a pure metal surface is exposed to humidity,it will be immediately covered with a thin layer of

e front matter r 2006 Elsevier Ltd. All rights reserved

mosenv.2006.10.018

ing author. Tel.: +46 704076788;

800.

ess: Farid.Samie@kimab.com (F. Samie).

water, which provides a medium for dissolving thepollutant. The dissolved corrosive gases serve as anelectrolyte for corrosion reactions on the metalsurface and corrosion is initiated by an electro-chemical mechanism. The thickness of the waterlayer depends on factors such as the surfaceroughness, relative humidity (RH) and temperature(Leygraf and Graedel, 2000).

In ambient environment, the effect of tempera-ture on atmospheric corrosion is hard to predict.Although an increased temperature accelerates thechemical reaction activity, it does not necessarily

.

ARTICLE IN PRESS

Table 1

Exposure conditions

Exposure time (h) (48–336)70.1

Temperature (1C) (15–35)70.5

[HNO3] (mgm�3) (80–400)720

RH (%) (0–85)74

Flow rate (cm3min�1) 60075

Air velocity (cm s�1) 0.03

F. Samie et al. / Atmospheric Environment 41 (2007) 1374–1382 1375

increase the corrosion rate. For example, as asurface is exposed to sunlight the temperatureincreases and the surface dries, reducing thecorrosion. However, a higher temperature can alsoincrease the condensation of an aqueous film andenhance the corrosion (Jones, 1996).

In the field of atmospheric corrosion the effects ofpollutants such as sulphur dioxide (SO2), nitrogendioxide (NO2) or ozone (O3) have been comprehen-sively investigated (Graedel et al., 1987; Graedel,1987a, b, 1989; Tidblad et al., 1991; Eriksson, 1992;Svensson, 1995; Tidblad and Kucera, 1996;Strandberg, 1997; Weissenrieder, 2003; Chen 2005;Tidblad et al., 2005). The role of NO2 on corrosion ofcopper was investigated extensively by Tidblad andLeygraf (1995) and Oesch and Faller (1997). However,so far very little effort has been dedicated to researchon the effects of the secondary pollutant nitric acid(HNO3). The high sticking coefficient of this gas, lackof commercial analytical instruments for continuousconcentration measurements and its relatively lowambient concentrations (Ferm et al., 2005) may bereasons why this gas has been so little studied.

HNO3 is formed mainly by reaction of NO2

with the hydroxyl radical (OHd) (Grosjean andBytnerowicz, 1993), and its high daytime and lownight time concentration is consistent with the involvedphotochemical processes (Bytnerowicz et al., 2005).

There are a few reports on HNO3-induced atmo-spheric corrosion effects on the deterioration ofcalcareous stones (Lipfert, 1989; Haneef et al., 1992;Kirkitsos and Sikiotis, 1995, 1996; Fenter et al.,1995). The only laboratory investigations so farreported on corrosion effects of HNO3 on metalsare, to our knowledge, our previous work presentedin a series of articles. In the first paper, in this series,a new method was developed and successfully testedfor the atmospheric corrosion effects of HNO3 onlaboratory-exposed copper (Samie et al., 2005). Inthe preceding paper, results were presented of theinfluence of HNO3 concentration and air velocityon the corrosion of copper (Samie et al., 2006a, b).The aim of the present study was to investigate therole of temperature and humidity in the presence ofsub-ppm levels of HNO3 on the atmosphericcorrosion of copper, using the same experimentalset-up and techniques as in the previous papers.

This work is a part of an EU project entitledMULTI-ASSESS with the main aim to estimatedose–response functions suitable for the combinedeffects of pollutants and climate parameters oncorrosion and soiling of significant materials in

cultural heritage objects, e.g. copper, zinc, paintedsteel, glass and limestone. For further details on thisproject see Kucera et al. (2006). In a future article,results on the corrosion effects of HNO3 on copper,zinc, carbon steel and Portland limestone will becompared with each other and also with reportedresults on their corrosion rate in presence of otherpollutants, e.g. SO2, NO2 and O3.

2. Experimental

The experimental technique was described indetail previously (Samie et al., 2005); however, forcompleteness a shorter description is given herebelow.

2.1. Sample preparation and exposure

Copper sheet (1mm), with a minimum coppercontent of 99.99%, were cut to shape of10mm� 50mm and a hole was drilled for suspend-ing the sample vertically on a glass rod in theexposure chamber. The copper samples were po-lished with SiC-paper in ethanol to 2400 mesh,degreased twice in acetone using ultrasonic agita-tion and dried with a hairdryer. The samples wereafter the preparation stored in a desiccator at least1 d and at most 4 d before being introduced in theexposure chamber. Immediately before and afterexposure, the coupons were weighed with a micro-balance with a precision of 72.0 mg (70.2 mg cm�2).

The experimental conditions are summarized inTable 1. The relatively high concentrations wereused to obtain measurable corrosion rates and alsoto enable comparison with studies of other corrosivegases performed.

The HNO3 source was a temperature- andpressure-controlled permeation tube (Kin-Tek,La Marque, TX). Perfluoroalkoxy (PFA) Teflonfitting and tubing was used to connect the HNO3

source to the mixing chamber and exposurechamber to minimize HNO3 adsorption (Neuman

ARTICLE IN PRESSF. Samie et al. / Atmospheric Environment 41 (2007) 1374–13821376

et al., 1999). As the HNO3 is readily dissolved inwater the exit air from the chamber was directedinto a gas wash bottle containing ultra pure water.Thus, the hazardous air could be cleaned fromHNO3, and the concentration of HNO3 could beestimated by quantifying the nitrate in the gas washbottle with ion chromatography.

It should be mentioned that, in a previouspublication by the same authors (Samie et al.,2005), the gravimetrical concentration measure-ments of the permeation tube were compared withion chromatographic results on gas concentration.It was found that there was a good agreementbetween these two methods. Thus, the formation ofother nitrogen-containing compounds at exposureconditions here is minimal, if existing.

2.2. Ideal absorber

The deposition of HNO3 on perfect/ideal absor-bent was measured using a sodium hydroxide(NaOH)-impregnated Munktell filter paper of thesame dimensions as the copper sample. Eachexperiment included two copper samples and onefilter paper exposed in the same chamber. Afterexposure, the deposited HNO3 on the filter wasleached in ultra pure water, which was thereafteranalysed by ion chromatography for quantificationof nitrates.

2.3. Mass change

After exposure the metal samples were stored in adesiccator and weighed within 24 h using themicrobalance. The quantification of mass loss ofexposed copper was performed according to thefollowing pickling procedure. Selected exposedsamples were introduced in a solution of 5wt%amidosulphonic acid (H3NO3S) for 30 s usingultrasonic agitation followed by water and ethanolrinses. The samples were thereafter dried with ahairdryer and weighed with the microbalance. Theremoval of bulk metal during the pickling procedurewas determined by pickling of un-exposed samples.

2.4. Deposition velocity

For estimation of the deposition velocity (Vd) ofHNO3 on copper the following parameters wereused: the flow velocity, the exposure duration, theHNO3 concentration of the air exiting the chamberbefore and during exposure of the samples, i.e. one

filter and two copper coupons, and the amount ofHNO3 deposited on the filter after the exposure. Adetailed description of the Vd estimation on copperand filter is given in Samie et al. (2006a, b).

2.5. Ion chromatography

Analysis of nitrates were performed by an ionchromatography model MetroOhm with 733 ICSeparation Center, 709 IC pump and 752 PumpUnit. The column chosen for analysing the samplesis of model 6.1006.5� 0 Metrosep A Supp 5. Theprepared eluent is 3.2mmol L�1 Na2CO3 and1.0mmol L�1 NaHCO3 in 5% acetone, with flow0.70mlmin�1. Detection was done with anelectric conductivity detector model Metrohm 732.Reproducibility of IC nitrate analysis was testedwith prepared standards in the range of0.1–6.0mgmL�1. The differences in repeated sam-ple analyses were lower than 5%.

2.6. Surface analysis

Surfaces of the exposed samples were analysedusing Fourier transform infrared (FT-IR) micro-scopy model Bio-Rad UMA 500, equipped with abroad band mercury cadmium telluride (MCT)detector (cut-off number at 450 cm�1). Also, X-raypowder diffraction (XRD) was used to characterizethe solid phases formed on the copper surface.

3. Results and discussion

It may be argued that the HNO3 concentrations(80–400 mgm�3) used are too high, compared toactual field-exposure conditions (0.5–4.3 mgm�3,Ferm et al., 2005). This issue is dealt with in moredetail in a forthcoming paper where HNO3-inducedcorrosion effects from laboratory exposures arecompared with those obtained from field exposures.Without providing further evidence here, it turnsout that the corrosion effects at the field stationsdominated by HNO3 influence can be fairly wellpredicted from laboratory exposures. To conclude,the actual HNO3 concentrations used in thelaboratory bear relevance, at least for those fieldsites having the highest HNO3 concentrations.

3.1. Characterization of the corrosion products

Analysis of the corrosion products using FT-IR,XRD, IC and mass analysis (mass gain and mass

ARTICLE IN PRESSF. Samie et al. / Atmospheric Environment 41 (2007) 1374–1382 1377

loss), all resulted in gerhardtite, Cu2(NO3)(OH)3,and cuprite, Cu2O, as the major corrosion productsformed on copper exposed to HNO3. Eriksson et al.(1993) also identified low amounts of gerhardtiteand Cu2O on copper, however, exposed to humidair containing NO2.

3.2. Effects of temperature

Laboratory experiments on the effect of tempera-ture on the atmospheric corrosion of metals arerelatively rare. When accelerated corrosion tests aredesigned, temperature often is increased consider-ably comparing to the mean ambient values. Thisincrease seems to be based on the assumption thatthe chemical activity, and thus, the corrosion rate isenhanced. In the present study, we varied thetemperature between +15 and +35 1C.

3.2.1. Mass gain

The effect of temperature on the mass gain isshown in Fig. 1. In a previous study on thecorrosion effects of HNO3 on copper exposed atvaried concentration and air velocity (Samie et al.,2006a, b), it has been shown that there is a linearincrease of copper mass gain with increasedconcentration of HNO3. Thus, in the figures below(Figs. 1, 4 and 7), the mass gain is presented at therate of mass increase per 100 mgm�3 HNO3 in orderto use all data obtained at varied concentrations. Anincrease of temperature from 15 to 25 1C at 65%RH did not have significant effect on the mass gain.

0

1

2

3

4

10 15 20 25 30 35

Temperature / °C

Ma

ss g

ain

/�g

cm

-2 d

-1 p

er

10

0 �

g m

-3 H

NO

3

Fig. 1. Mass gain of copper vs. temperature at 65% RH,

0.03 cm s�1 air velocity and varied HNO3 concentrations

(100–400mgm�3). Error bars show the standard deviation of at

least three samples.

This is in agreement with other reported results oncorrosion of copper in the same temperature range(Golubyev et al., 1968; Barton, 1973).

At 35 1C the average mass gain was somewhatlower than that obtained at 15 and 25 1C. There aresome studies showing the thickness of water adlayeron the metal surface to decrease with increase oftemperature within the range of 15–35 1C (Strekalovet al., 1972; Phipps and Rice 1979). A thinner aqueouslayer may lead to reduced corrosion rate, however, aswill be presented further below. Based on mass gainand mass loss analysis, no significant variation in theamount of different corrosion products could beobserved with increased temperature.

Based on the earlier obtained results, the majorconstituents of the corrosion product on copperexposed to HNO3 and humid air were identified asthe basic copper nitrate, gerhardtite and cuprite,Cu2O. Combining this information with the resultson mass gain and mass loss, the ratio of copper inthe corrosion layer, and thus, the ratio of Cu2O andgerhardtite, could be estimated (Fig. 2). Also, thetheoretical lines for Cu2O and gerhardtite areincluded in the same figure.

As can be seen in the figure, a minor decrease ofthe average ratio, i.e. decreased ratio of Cu2O in thecorrosion layer, can be observed with increasedtemperature.

3.2.2. Deposition velocity

The deposition velocity, Vd, of HNO3 on copperand filter was tested at 65% RH, 0.03 cm s�1 air

0.5

0.6

0.7

0.8

0.9

10 15 20 25 30 35

Ratio c

opper

in t

he c

orr

osio

n layer Cu2O 0.888

Cu2(NO3)(OH)3 0.529

Temperature / °C

Fig. 2. Ratio of copper in the corrosion layer vs. temperature at

65% RH, 0.03 cm s�1 air velocity and varied HNO3 concentra-

tions (100–400mgm�3). The dashed lines represent the theoretical

ratio of copper for 100% gerhardtite (lower dashed line) or 100%

Cu2O (the upper dashed line) in the corrosion layer. Error bars

show the standard deviation of at least three samples.

ARTICLE IN PRESSF. Samie et al. / Atmospheric Environment 41 (2007) 1374–13821378

velocity and varied temperatures. The results arepresented in Fig. 3. The average Vd of HNO3 oncopper was in all measurements no less than 75% ofthat of the ideal absorbent, i.e. impregnated filter,which confirms the high sticking coefficient ofHNO3. The Vd of HNO3 on copper, as for theideal absorbent, did not change significantly withincreased temperature.

In a previous investigation on the influence of airvelocity on the corrosion of copper (Samie et al.,2006a, b), we presented the Vd of HNO3 on the filterand on the copper to be nearly the same at0.03 cm s�1 air velocity. However, in the presentstudy, based on more data, it can bee seen that theaverage Vd of HNO3 on the copper is slightly lowerthan that on the filter at 15–35 1C.

The results on the mass gain of copper and theratio of corrosion products with increased tempera-ture (Figs. 1 and 2) showed constant amount ofgerhardtite and a minor decrease in the amount ofCu2O in the corrosion layer with increased tem-perature. Also, as the Vd of HNO3 on copper didnot change significantly (Fig. 3), it is evident thatthe corrosion rate of copper in the present study waslittle affected by a raise of the temperature from 15to 35 1C at 65% RH.

3.3. Effects of relative humidity

The quantity of adsorbed water on a surfaceincreases with the RH. This aqueous layer willprovide a bulk for the acidic pollutant to dissolveand the corrosion rate is increased. As will be

0.0

0.1

0.2

0.3

0.4

10 15 20 25 30 35

Vd H

NO

3/c

m s

-1

Temperature /°C

Fig. 3. HNO3 deposition velocity on filter (circles) and copper

(triangles) vs. temperature at 65% RH, 0.03 cm s�1 air velocity

and varied HNO3 concentrations (100–400mgm�3). Error bars

show the standard deviation of at least three samples.

presented in the results below, due to the highsticking coefficient and corrosivity of HNO3,significant corrosion of copper was observed atRH as low as 20%.

3.3.1. Mass changes

According to Leygraf (2002), the number ofmonolayers on a metal surface is below 2 mono-layers up to 40% RH (see Table 2).

The effect of RH on the mass gain of copperexposed at 25 1C and sub-ppm levels of HNO3 ispresented in Fig. 4. At dry condition (0% RH), themass increase was at minimum, and as the RH wasincreased to 20% RH, significant mass gain couldbe observed. This mass gain increased relativelylittle with increased humidity up to 40% RH(o1 mg cm�2 d�1 per 100 mgm�3 HNO3) and at65% RH it increased to nearly 3 mg cm�2 d�1 per100 mgm�3 HNO3. For the aqueous layer toapproach bulk properties, the number of watermonolayers on the metal surface must be thickerthan approximately three (Phipps and Rice, 1979).In such condition, the gas can dissolve in theaqueous layer forming an electrolyte. The chemicalreaction on the metal surface is enhanced and themetal corrodes more rapidly and homogenously.Thus, the high increase of the mass gain from 40%to 65% RH (Fig. 4) is suggested to be an effect ofthe adlayer to take on bulk properties.

As the RH was increased from 65% to 85% themass gain did not change significantly. However, aswill be presented further below, the formation ofCu2O was enhanced at 85% RH. The mass gain ismeasured and estimated as the amount of otherthan Cu elements in the corrosion layer, i.e. O witha molecular mass of 16.0 gmol�1 for Cu2O andNO3(OH)3 with a molecular mass of 113.0 gmol�1

for gerhardtite. Hence, increased amount of Cu2Oat levels presented here have a relatively low impacton the mass gain.

Table 2

Approximate number of water monolayers on different metal and

alloy surfaces, including gold, cobalt, nickel, iron and nickel-iron

alloys at 25 1C vs. RH

RH (%) Number of monolayers

20 1

40 1.5–2

60 2–5

80 5–10

Data from Leygraf (2002).

ARTICLE IN PRESS

0

1

2

3

4

0 20 40 60 80 100

RH /%

Ma

ss g

ain

/�g

cm

-2 d

-1 p

er

10

0 �

g m

-3 H

NO

3

Fig. 4. Mass gain of copper vs. relative humidity at 25 1C,

0.03 cm s�1 air velocity and varied HNO3 concentrations

(100–400mgm�3). Error bars show the standard deviation of at

least three samples. The line is a guide to the eye.

0.5

0.6

0.7

0.8

0.9

Ra

tio

co

pp

er

in t

he

co

rro

sio

n la

ye

r Cu2O

Cu2(NO

3)(OH)

3

0 20 40 60 80 100

RH /%

Fig. 5. Ratio of copper in the corrosion layer vs. relative

humidity at 25 1C, 0.03 cm s�1 air velocity and varied HNO3

concentrations (80–400mgm�3). The dashed lines represent the

theoretical ratio of copper for 100% gerhardtite (lower dashed

line) or 100% Cu2O (the upper dashed line) in the corrosion

layer. Error bars show the standard deviation of at least three

samples.

0 20 40 60 80 100

RH /%

0.0

0.1

0.2

0.3

0.4

Vd H

NO

3/c

m s

-1

Fig. 6. HNO3 deposition velocity on filter (circles) and copper

(triangles) vs. relative humidity at 25 1C, 0.03 cm s�1 air velocity

and varied HNO3 concentrations (80–400mgm�3). Error bars

show the standard deviation of at least three samples. The full

line is a guide to the eye.

F. Samie et al. / Atmospheric Environment 41 (2007) 1374–1382 1379

In Fig. 5 the ratios of copper in the corrosionlayer at varied RH are estimated based on theresults on mass gain and mass loss of exposedsamples. Noteworthy, the results of the estimatedratio of copper in the corrosion layer at low RH(p20% RH) shown in Fig. 5, have a relativelyhigh statistical error due to somewhat low massgain and mass loss and are based on only twoexperiments.

In the previous section on the effects of tempera-ture, we suggested a slightly lower formation ofCu2O as the temperature was increased from 25 to35 1C. Also, from the literature, we found that thethickness of the water layer on the metal surfacedecreases with increased temperature. As can beseen in Fig. 5, the ratio of copper increased slightlyas the RH was increased from 65% to 85%,indicating a higher amount of Cu2O in the corrosionlayer.

3.3.2. Deposition velocity

Fig. 6 illustrates the Vd of HNO3 on copper andon filter at varied RH. The deposition of HNO3 on aperfect absorbent is dependent on the number ofmolecules that strike its surface since the diffusioncoefficient of HNO3 in air (T ¼ 25 1C, P ¼ 1 atm) isindependent of RH (Durham and Stockburger,1986). This is confirmed by the results presented inFig. 6.

However, from the results plotted in Fig. 6 at dryconditions, i.e. 0% RH, it can be seen that Vd ofHNO3 on the copper surface was far below that of afilter. The trend for Vd of HNO3 on copper looked

very similar to that obtained for the mass gain atvaried RH (Fig. 4). That is, a significant Vd alreadyat 20% RH and slow increase up to 40% RH,indicating the high sticking coefficient of HNO3,and the maxima of Vd reached at 65% RH(approximately 0.15 cm s�1). At 85% RH the Vd

remained unchanged from that at 65% RH.Independent of the exposure time, an increase of

RH results in increased amount of physisorbedwater (Aastrup, 1999). As has been described in theearlier section, for water to take on bulk property

ARTICLE IN PRESSF. Samie et al. / Atmospheric Environment 41 (2007) 1374–13821380

on a surface, the number of monolayers have to be 3or more, excluding interfacial water adjacent to thegas phase or the solid phase. This may be achievedat X60% RH (Table 2) and at this condition thewater layer is likely to act as a perfect sink forHNO3. The maximum Vd of HNO3 on copper wasreached at 65% RH (Fig. 6) and remained un-changed as the RH was increased to 85% RH. Incomparison to Vd for other corrosive gases, e. g.SO2, NO2 or O3 on copper, Vd for HNO3 is farabove any of the mentioned gases (Samie et al.,2006a, b).

3.4. Corrosion kinetics

It is well known that the corrosion rate of a metalinitially is high, and as the oxide layer grows thecorrosion rate decreases and levels off. In earliersections, we presented the influence of temperatureand RH on the Vd of HNO3 on copper and on theformation of Cu2O and gerhardtite. We now presentthe influence of time on the growth of the coppercorrosion products which was examined after 2, 7and 14 d.

The corrosion rate of the samples exposed to sub-ppm levels of HNO3 is presented in Fig. 7, as thetotal mass gain (triangles) and the correspondingmass gain due to Cu2O (squares) as a function oftime. The linear increase of the total mass gain upto 7 d somewhat decreased at 14 d of exposure andthe contributing mass gain from Cu2O levelled offafter 7 d suggesting no further formation of Cu2O.It is important to bear in mind that the mass

0

10

20

30

40

0 2 4 6 8 10 12 14

Time /days

Mass g

ain

/�g c

m-2

per

100 �

g m

-3 H

NO

3

Fig. 7. Total mass gain of copper (triangles) and corresponding

mass gain due to Cu2O (squares) vs. the exposure time at 65%

RH, 25 1C, 0.03 cm s�1 air velocity and varied HNO3 concentra-

tions (80–400mgm�3).

gain due to Cu2O, relative to gerhardtite(Cu2(NO3)(OH)3), is very low. Combining theobtained mass gain and mass loss results it wasevident that after 2 d of exposure, nearly the sameamount of copper was consumed for formation ofgerhardtite as for formation of Cu2O. After 14 d thefraction of Cu2O in the corrosion layer decreasedremarkably and the thickness of the oxide wasestimated to approximately 14 nm, when exposed to80 mgm�3 HNO3. The corresponding oxide thick-ness was approximately 9 nm after only 2 d ofexposure.

With the growth of the corrosion layer, thetransport of the corrosive pollutant to the basemetal will be constrained, and consequently, thecorrosion rate may decrease.

4. Conclusions

The effect of temperature on the corrosion ofcopper in the presence of sub-ppm levels of HNO3

has been shown in the present study to be very low.An increase of temperature from 15 to 35 1Cresulted in a minor decrease of copper mass gain.Such decrease in the mass gain is suggested to bedue to a reduced formation of Cu2O, which can bean effect of a thinner adlayer on the metal surface,as the temperature is increased. However, the Vd ofHNO3 on copper and filter at varied temperatureremained constant at 65% RH and above.

The HNO3-induced atmospheric corrosion ofcopper was found to be significant already at 20%RH and the highest corrosion rate was reached at65% RH. A similar trend was also found for theHNO3 deposition velocity (Vd), where Vd increasedslowly up to 40% RH and reached a plateau at 65%RH, corresponding to the formation of an aqueousadlayer that takes on bulk properties. An increase inRH from 65% to 85% RH did not show anysignificant change in neither the mass gain nor theVd of HNO3 on copper. The relative amount ofCu2O in the corrosion layer, however, increased athigh RH.

At 65% RH, 25 1C and 0.03 cm s�1 air velocity,the corrosion rate of copper showed a trend oflevelling off after 14 d and no further growth of theoxide was observed after 7 d of exposure. Withincreased thickness of the oxide layer, the transportof HNO3 to the base metal is believed to beconstrained and the corrosion rate decreases as aconsequence, which seems to be the case in thepresent study.

ARTICLE IN PRESSF. Samie et al. / Atmospheric Environment 41 (2007) 1374–1382 1381

Our results confirm the high corrosivity of HNO3,its high sticking coefficient and high Vd on copper.

5. Further and ongoing work

In a coming paper, we will compare the effects ofHNO3 on laboratory-exposed copper, zinc, carbonsteel and lime stone.

Acknowledgements

This study was supported financially by theEuropean Community, within the MULTI-ASSESSproject (Contract no.: EVK4-CT-2001-00044), TheSwedish Council for Environment, AgriculturalScience and Spatial Planning (FORMAS) and TheSwedish Agency for Innovation Systems (VINNO-VA). The authors gratefully acknowledge helpfuldiscussions with Dr. Martin Ferm at the SwedishEnvironmental Research Institute.

References

Aastrup, T., 1999. In situ investigations of the metal/atmosphere

interface. Influence of humidity, sulfur dioxide, ozone and

nitrogen dioxide on copper. Ph.D. Thesis (ISBN 91-7170-481-7).

Barton, K., 1973. Schutz gegen atmospharische Korrosion,

Theorie und Technik. Verlag Chemie GMBH Weinheim,

Germany (in German).

Bytnerowicz, A., Sanz, M.J., Arbaugh, M.J., Padgett, P.E.,

Jones, D.P., Davila, A., 2005. Passive sampler for monitoring

ambient nitric acid (HNO3) and nitrous acid (HNO2)

concentrations. Atmospheric Environment 39 (14),

2655–2660.

Chen, Z.Y., 2005. The Role of Particles on Initial Atmospheric

Corrosion of Copper and Zinc. Ph.D. Thesis, ISBN:91-7178-

155-2.

Durham, J.L., Stockburger, L., 1986. Nitric acid-air diffusion

coefficient: experimental determination. Atmospheric Envir-

onment 20 (3), 559–563.

Eriksson, P., 1992. Effects of SO2 and NO2 on metal surfaces:

atmospheric corrosion of copper and surface reactions on

gold. Ph.D. Thesis, ISBN 91-7032-661-4.

Eriksson, P., Johansson, L.G., Strandberg, H., 1993. Initial

stages of copper corrosion in humid air containing SO2 and

NO2. Journal of the Electrochemical Society 40 (1), 53–59.

Fenter, F.F., Caloz, F., Rossi, M.J., 1995. Experimental evidence

for the efficient ‘‘dry deposition’’ of nitric acid on calcite.

Atmospheric Environment 29 (22), 3365–3372.

Ferm, M., De Santis, F., Varotsos, C., 2005. Nitric acid

measurements in connection with corrosion studies. Atmo-

spheric Environment 39, 6664–6672.

Golubyev, A.I., Kadyrov, M.Kh., 1968. Metal corrosion forecast

under atmospheric conditions. Corrosion Week, Budapest,

pp. 1039–1044.

Graedel, T.E., 1987a. Copper patinas formed in the atmo-

sphere—II. A qualitative assessment of mechanisms. Corro-

sion Science 27 (7), 721–741.

Graedel, T.E., 1987b. Copper patinas formed in the atmo-

sphere—III. A semi-quantitative active assessment of rates

and constraints in the greater New York metropolitan area.

Corrosion Science 27 (7), 741–769.

Graedel, T.E., 1989. Corrosion mechanisms for zinc exposed to

the atmosphere. Journal of the Electrochemical Society 136

(4), 193C–203C.

Graedel, T.E., Nassau, K., Franey, J.P., 1987. Copper patinas

formed in the atmosphere—I. Introduction. Corrosion

Science 27 (7), 639–657.

Grosjean, D., Bytnerowicz, A., 1993. Nitrogenous air pollutants

at a southern California mountain forest smog receptor site.

Atmospheric Environment 27A (4), 483–492.

Haneef, S.J., Johnson, J.B., Dickinson, C., Thompson, G.E.,

Wood, G.C., 1992. Effect of dry deposition of NOx and

SO2 gaseous pollutants on the degradation of calcareous

building stones. Atmospheric Environment 26A (16),

2963–2974.

Jones, D.A., 1996. Principles and Prevention of Corrosion,

second ed. Prentice-Hall, Inc. Simon and Shuster/A Viacom

Company, USA.

Kirkitsos, P., Sikiotis, D., 1995. Deterioration of Pentelic marble,

Portland limestone and Baumberger sandstone in laboratory

exposures to gaseous nitric acid. Atmospheric Environment

29 (1), 77–88.

Kirkitsos, P., Sikiotis, D., 1996. Deterioration of pentelic marble,

Portland limestone and Baumberger sandstone in laboratory

exposures to NO2: a comparison with exposures to gaseous

HNO3. Atmospheric Environment 30 (6), 941–950.

Kucera, V., Tidblad, J., Kreislova, K., Knotkova, D., Faller, M.,

Reiss, D., Snethlage, R., Yates, T., Henriksen, J., Schreiner,

M., Melcher, M., Ferm, M., Lefevre, R.-A., Kobus, J., 2006.

The UN/ECE ICP materials multipollutant exposure on

effects on materials including historic and cultural monu-

ments. Water, Air and Soil Pollution, accepted.

Leygraf, C., 2002. Atmospheric corrosion. P. Marcus (Ed.),

Corrosion Mechanism in Theory and Practice, second ed.,

pp. 529–562.

Leygraf, C., Graedel, T., 2000. Atmospheric Corrosion. Wiley,

New York.

Lipfert, F.W., 1989. Atmospheric damage to calcareous stones:

comparison and reconciliation of recent experimental find-

ings. Atmospheric Environment 23 (2), 415–429.

Neuman, J.A., Huey, L.G., Ryerson, T.B., Fahey, D.W., 1999.

Study of inlet materials for sampling atmospheric nitric acid.

Environmental Science and Technology 33, 1133–1136.

Phipps, P.B.P., Rice, W., 1979. The role of water in atmospheric

corrosion. American Chemical Society 8, 236–261.

Oesch, S., Faller, M., 1997. Environmental effects on materials:

the effect of the air pollutants SO2, NO2, NO and O3 on the

corrosion of copper, zinc and aluminium. A short literature

survey and results of laboratory exposures. Corrosion Science

39 (9), 1505–1530.

Samie, F., Tidblad, J., Kucera, V., Leygraf, C., 2005. Atmo-

spheric corrosion effects of HNO3—method development and

results on laboratory exposed copper. Atmospheric Environ-

ment 39, 7362–7373.

Samie, F., Tidblad, J., Kucera, V., Leygraf, C., 2006a. Atmo-

spheric corrosion effects of HNO3—influence of concentration

ARTICLE IN PRESSF. Samie et al. / Atmospheric Environment 41 (2007) 1374–13821382

and air velocity on laboratory exposed copper. Atmospheric

Environment 40, 3631–3639.

Samie, F., Tidblad, J., Kucera, V., Leygraf, C., 2006b. Atmo-

spheric corrosion effects of HNO3—comparison of labora-

tory exposed copper, zinc and carbon steel. Atmospheric

Environment, submitted for publication.

Strandberg, H., 1997. Perspectives of bronze sculpture conserva-

tion; modelling copper and bronze corrosion. Ph.D. Thesis,

ISBN:91-7197-539-X.

Strekalov, P.V., Agafonov, V.V., Mikhailovski, Yu.N., 1972.

Effect of temperature on the adsorption of moisture and the

corrosion rate of zinc under atmospheric condition. Protec-

tion of Metals 8, 521–523.

Svensson, J.-E., 1995. The influence of different air pollutants on

the atmospheric corrosion of zinc. Ph.D. Thesis, ISBN:91-

7197-086-X.

Tidblad, J., Aastrup, T., Leygraf, C., 2005. GILDES model

studies of aqueous chemistry VI. Initial SO2/O3- and SO2/

NO2-induced atmospheric corrosion. Journal of the Electro-

chemical Society 152 (5), B178–B185.

Tidblad, J., Kucera, V., 1996. The role of NOx and O3 in the

corrosion and degradation of materials. Technical Report No.

1996:4E, Swedish Corrosion Institute.

Tidblad, J., Leygraf, C., 1995. Atmospheric corrosion effects of

SO2 and NO2: a comparison of laboratory and field exposed

copper. Journal of the Electrochemical Society 142 (3),

749–756.

Tidblad, J., Leygraf, C., Kucera, V., 1991. Acid deposition effects

on materials: evaluation of nickel and copper. Journal of the

Electrochemical Society 138, 3592–3598.

Weissenrieder, J., 2003. Molecular level studies of the metal/

atmosphere interface. Ph.D. Thesis, ISBN:91-7283-399-8.