DETERIORATION
OF J E R U S A L E M L I M E S T O N E
FROM A I R P O L L U T A N T S ; FIELD OBSERVATIONS AND LABORATORY SIMULATION M O R D E C H A I P E L E G , E S T E R B U R L A , IRIS C O H E N , and M E N A C H E M L U R I A
Environmental Sciences Division, School o f Applied Science and Technology, The Hebrew University o f Jerusalem, Jerusalem 91904, lsrael
(Received April 1988) Abstract. Samples of Jerusalem limestones were exposed to high levels of common air pollutants (SO2, NO) in the presence and absence of hydrocarbon, water vapor and ultra-violet light. After exposure, the outer layers were shaved off and analyzed for sulfate and nitrate. The results revealed that even after one day of exposure significant concentrations of CaSO4 and Ca(NO3) 2 could be detected in the external 40/zm layers. Sulfate formation was found to relate very strongly to relative humidity with nearly undetectable production at humidities below 10o7o.Nitrate formation was found to relate to UV light and to a limited extent to the presence of hydrocarbon but was unchanged at different humidities. Surface samples were taken from different sites of the old city wail and were also analyzed for the same substances. The data showed that the concentrations of especially CaSO4, and to a limited extend Ca(NO3)2, measured at various points along the city walls were higher than the expected values extrapolated from laboratory simulations. The elevated pollution content may be explained in part due to the deposition of transported ions, especially sulfate, onto the stone surface. It was concluded that although air pollution may not cause any structural damage in the foreseeable future it definitely deteriorates the fine details of the ancient monument.
A. Introduction The damage caused by air pollutants to historical stone buildings and stone monuments is well documented in the literature [1-4]. The list of famous historical sites damaged by air pollutants includes the Acropolis in Athens, Greece; the Roman forum in Rome as well as the St. Mark's Basilica in Venice. The state of Israel is rich in important stone archeological monuments many of which are located in Jerusalem. Of special importance are the temple mount which includes the Western (Wailing) Wall, the El Aksa and Omar shrines, the old city walls and the numerous newly excavated sites, some of which date back to the first Israeli commonwealth period (about 3000 years). In addition, the Jerusalem by-laws enforce the use of locally available limestone as the only material to be employed in the facade of all public and private buildings. Visible air pollution damages have been noted on many relatively new downtown buildings as evidenced by the blackening of the stone. The relation to air pollution is demonstrated by the high ambient pollution levels in this district [5], where 1.5 ppm of nitrogen oxides have been observed near the walls. The increased usage of fossil fuels has caused concern regarding the possible effects of air pollution on these stone monuments. Air quality data, available since 1979 indicated that the air quality in the city (population 500 000) is affected by both local Environmental Monitoring and Assessment 12: 191-201, 1989. 9 1989 Kluwer Academic Publishers. Printed in the Netherlands.
192
MORDECHAI PELEG ET AL.
sources, mainly vehicular transportation, as well as transportation of air pollutants from the heavily industrialized coastal region situated 50-100 km up wind of the city [6-71. In order to assess the damage to local limestone from sulfur dioxide and nitrogen oxides and the corresponding acids H2SO4 and HNO3, laboratory simulation were performed in which hundreds of exposure years were compressed into several days. Utilizing the laboratory developed methods, an attempts was made to determine the extent of air pollution influence on the ancient city wall.
2. Experimental The simulations were performed in an environmentally controlled exposure chamber previously used for photochemical aerosol generation [8]. The 18 liter cylindrical pyrex chamber was symmetrically surrounded by a bank of 4 GE dark light 40 w fluorescent lamps. The light intensity of this set up was estimated to photolyse NO2 at a rate of 5 min-1 (for comparison, NO2 photolysis rate peaks at 0.6 min-1 at noon, during the summer at mid-latitudes). In order to shorten the experimental duration it was necessary to use extremely high concentrations of gaseous pollutants so that a measurable effect could be obtained within a reasonable time scale. Pure (Matheson Inc.) sulfur dioxide and nitrogen monoxide were diluted to approximately 300 ppm each. The humidity for each experiment was controlled by introducing into the chamber saturated inorganic salt solutions of known water vapor pressure or water in open flasks. A similar method was used to obtain a fixed vapor pressure of hydrocarbon (hexene). The reaction mixture was brought up to atmospheric pressure by adding clean air. Excess heat was removed by air blowers that maintained the temperature at around 30 ~ The duration of a typical experiment was between 1 and 7 days. The average annual mean concentration of SO2 and NOx in Jerusalem (at roof top - 10 m above street level) is 0.005 and 0.025 ppm respectivly [7]. A comparison of air quality at roof top and street levels, has shown that the ambient pollutant concentrations are approximately four times larger at street levels than at roof tops [9]. Stone edifices at street levels will therfore be exposed to much higher pollution, and each laboratory simulation day will represent 10 to 40 years of natural exposure. This conclusion applies, however, only to the primary pollutants. The concentration of secondary pollutants (H2SO4, and HNO3), photochemically produced during the simulation experiment, were not monitored. It is expected, however, that the enhancement of the secondary pollutants concentration is not linearly proportional to the increased primary gases due to the increased loss of the free radicals under the experimental conditions employed. The limestone samples used for the exposure were taken from five different quarries in the immediate vicinity of the city and typical of the stone used through out the years in the area.. A 5 cm cylindrical core was drilled from the freshly mined stone block and then cut into 1 cm thick slices that were used in the
DETERIORATION OF JERUSALEM LIMESTONE FROM AIR POLLUTANTS
193
experiments. Before each experiment the stones were cleaned to remove any deposits, and then dried. After exposure the samples were prepared for ion chromatographic analysis (IC) for sulfate and nitrate. This included the removal of about a 40/tin deep layer by carefully filing away 200 mg of the surface. The resulting powder was added to 50 ml of deionized water which was treated in an ultrasonic bath for one hour. The samples were allowed to stand for 24 hours and then retreated in the bath for 15 additional minutes. The relativly high volume of water was needed to ensure that all the CaSO 4 produced would be extracted into the aqueous solution. The minimum detection limit was better than 0.1 mg per sample analysed for both substances. The study was extended to the field by implementing the same methodology. Stone samples were taken from the northwest corner of the old city walls between the New Gate and Jaffa Gate. Samples were taken from areas that are subject to direct rain washout and others in areas that are protected from direct rainfall, inside the arches of the Jaffa Gate. Details of the sample area and the traffic density in the immediate vicinity are presented in Figure 1. The sampling of the walls was performed by grinding off a designated area to a depth of 200/tm and a second layer that extended approximately 1 mm inside the stone. Due to the heterogenic nature of the stone surface the method of sectioning was of necessity crude and thus introduced divergency into the final results.
Q
I: 2 5 0 0
(vehicles per dey) Fig. 1. Map of old city wall and adjacent roads, illustrating sampling locations and traffic densities.
194
MORDECHAI PELEG ET AL.
3. Results
The parameters investigated in the present study were initial concentration of S O 2 and NO, relative humidity, radiation intensity, hydrocarbon presence and exposure time. In order to identify the rate limiting steps, the first set of experiments was performed in the presence of all the aforementioned parameters. List of initial conditions and characteristics of stone examined appears in Table I. In the succeeding experiments a given parameter was excluded in order to determine its effect on the overalI rate of sulfate and nitrate formation. 3.1. Ca(SO4) FORMATION The results of Ca(SO4) formation under the various experimental conditions are summerized in Table II. A similar rate of sulfate formation was observed in presence TABLE I Initial experimental conditions for limestone exposures Parameter
Value
SO 2 NO Relative humidity H C (hexane) Duration
280-300 p p m 290-300 p p m 100~ 50 tort 1-10 days
Limestone Characteristics Density Porosity [CasSO4]o [CaNO3]0
2.58-2.64 g m cm -3 3-4% 0.55-0.65% < 0.02~
T A B L E II Percent sulfate formation after exposure o f Jerusalem limestone to 300 p p m o f SO 2 Exp. time (days)
0
1
3
4
5
All parameters stone 1
0.6
0.6
0.8
1.1
1.25
All parameters stone 2
0.65
Light excluded stone 3
0.65
0.85
1.1
1.15
0.55
0.6
H C excluded stone 4
1.1
1.7
0.45
0.7
0.95
1.4
H20 excluded a stone 3
0.6
0.5
0.5
0.7
a Relative humidity lower than 10~
2
0.5
6
7
2.3
0.5
DETERIORATION OF JERUSALEMLIMESTONE FROM AIR POLLUTANTS
195
CoS0 4 FORMATION IN LIMESTONE vs Relative
2.4 t-
.9 2.0 13
E ~-
0 LL
1.6
9 - 0%
RH
0-25%
RH
Humidity
o - 5 0 % RH A - 7 5 % RH 9
-90%
0-100%
RH RH
1.2
0
O0
o 0.8
rj
0.4 ~
0 I 0
I 2
t.) Fig. 2.
I 4
I 6
I 8
I I0
E xposure period (doys) The production of CaSO4 resulting from exposure of limestone to 300 ppm SO2 for various time periods at different humidities,
and absence of ultra violet (UV) radiation and hydrocarbon (HC). In these experiments, a near linear relationship was obtained between exposure time and S O 4 2 f o r m a t i o n . A significant reduction in sulfate formation was observed in the experiments where the relative humidity was kept below 10%, with almost no increase in the CaSO4 content observed. An additional set of experiments was performed at different relative humidities and the results are presented in Figure 2. The data, shown in this figure, clearly demonstrates the effect of relative humidity. There is a sharp increase in the reaction rate once the relative humidity exceeds 50~ 3.2. NITRATE FORMATION
The formation of CaNO3 does not follow the same pattern as for Ca(SO4) formation, as shown in Table III. Exclusion of UV light reduced the nitrate formation by about 50~ It also appears that in the absence of HC, nitrate formation was slightly reduced. On the other hand, the presence of water did not seem to affect the reaction rate as shown in Figure 3 which illustrates the amount of nitrate produced under various relative humidities. In all experiments Ca(NO3) 2 formation showed a near linear relationship with exposure time.
196
MORDECHAI PELEG ET AL.
Ca(NO 3)2
i.0[
vs
o-
FORMATION IN LIMESTONE Relative Humidity
0% RH
o
/ o- ~OO,o . ,
,,o ~
~-
~O,o
.H
/
/
I
/~
I "-'~176176176
z 0.4 0.2
~o o
0
~,
2
E
Fig. 3.
4
6
8
I0
Exposure period (days) The production of Ca(NO3)2 resulting from exposure of limestone to 300 p p m NOx for various time periods at different humidities.
T A B L E III Percent nitrate formation after exposure of Jerusalem limestone to 300 p p m NO Exp. time (days)
0
1
3
4
5
6
All parameters stone 1
n.d.
0.09
0.13
0.22
0.28
0.34
All parameters stone 2
n.d.
0.10
0.15
0.22
0.26
0.37
Light excluded stone 3
n.d.
0.06
0.07
0.14
H C excluded stone 4
n.d.
0.04
0.10
0.18
H 2 0 excluded a stone 3
n.d.
0.08
0.19
0.17
a Relative humidity lower than 10~/0. n.d. = not detectable.
2
0.32
7
0.14
0.28
DETERIORATION OF JERUSALEM LIMESTONE FROM AIR POLLUTANTS
197
3.3. PENETRATION DEPTH A series o f experiments was performed in order to evaluate the degree o f penetration o f sulfate and nitrate into the stone. The results are presented in Table IV. In the penetration experiments a second 40 micron deep layer was shaved o f f the stone. The data show that in the case o f sulfate there was no added detectable sulfate in the second layer even after continued exposure to 300 ppm SO2 for 2 days. In the nitrate study it was possible to observe penatration of nitrate into the second layer which amounted to 25-30% of the nitrate level in the outer layer. It should be noted, however, that since the unexposed stone contained no detectable nitrate it was possible to observe an even small increase of Ca(NO3)2 which was not possible in the sulfate case. TABLE IV Penetration degree of sulfate and nitrate in Jerusalem limestone a Exp. Time
[CaSO4] 1st layer (per cents)
[CaSO4] 2nd layer (per cents)
[Ca(NO3h] 1st layer (per cents)
[Ca(NO3)2] 2nd layer (per cents)
0
0.65
0.45
n.d.
n.d.
1
0.65
0.60
0.10
n.d.
3
0.85
0.70
0.15
0.06
4
1.10
0.60
0.22
0.06
5
1.15
0.65
0.26
0.10
0.37
0.13
6
a Initial exposure concentrations, [SOz] = 300 ppm and [NO] = 300 ppm.
3.4. REACTION RATES From the above data it is possible to estimate the lower limit of the rate of formation for CaSO4 and Ca(NO3)2 in the outer surface 40 ~tm layer. This estimate assumes linearity with concentration and does not take into consideration that the rate of secondary pollutant production is not linear. At 100% relative humidity, 0.26% of sulfate is produced upon exposure to 1 ppm of SO 2 for a period o f one year and only 0.09070 at humidities below 50%. The nitrate formation rate is 0.09% y r - 1 upon exposure to 1 ppm NO under daylight conditions, and is reduced by about 50% at night-time, (the reaction rate takes also into consideration some penertration into the second layer). Thus an average value o f 0.07% ppm NO y r - l is applicable as a lower limit. 3.5. OLD CITY WALL SAMPLES Old wall samples were taken twice. Once in the early summuer (May 1985), after the end o f the rainy season and again, at the same location in December 1985. The results of the CaSO4 and Ca(NO3)2 levels present in the wails are summarized in Table V,
198
MORDECHAI PELEG ET AL.
(the CaSO 4 levels are corrected for background content of sulfate). The results can be divided into two separate groups; exposed sites - I, II, III, and V, and protected site - IV (see Figure 1 for locations). The levels in the protected site as compared to the exposed sites are 3-4 higher for sulfate and about 50~ larger for nitrate. The results for the two layers are comparable, indicating that the pollution penetrated at least several mm into the stone. 4. Discussion
The type of stone typically used in the Jerusalem area is a finely crystalline limestone, white to biege in color, whose major component is calcite (CaCO3) combined with small quantities of acid insoluble residues (5-7%) and iron oxides (> 1%). It is a relatively hard stone, although softer than marble, making it suitable for many building purposes. Since its major component is a salt of a very weak acid, it is readily attacked even by weak acids to form the corresponding salt causing destruction of the crystalline structure and hence damage to the mechanical strenght of the stone. Additionally, the salts formed are much more soluble in water and can be washed out from the stone surface. It is therefore obvious that the result of air pollution will be the formation of the slightly soluble CaSO 4 and the extremely soluble Ca(NO3) 2, both of which are easily removed from the stone surface by powdering or washout. Three mechanisms describing the formation of CaSO 4 have been proposed. At low relative humidities (< 50%), Gauri et al. (1973) [10] have suggested a direct attack of SO2 on the stone to form CaSO 3 which may further oxidize only in the presence of catalyst and water to produce CaSO4. Above 50% humidity heterogeneous SO2 oxidation occurs on the stone surface to form H2SO 4 due to the catalytic influence of the trace compounds in the stone. The newly formed sulfuric acid attacks the CaCO 3 immediately. The third mechanism involves the direct attack of the sulfuric acid, produced in the atmospheric oxidation of SO 2. In this case the relative humidity is not expected to affect the rate of attack. The results of the present study confirm the effect of relative humidity on the rate of attack. As shown in Figure 2 the reaction rate is decreased by a factor of six when the relative humidity is lowered below 50%. Since the presence or absence of UV light does not effect the reaction rates a high and low relative humdities it seems that heterogeneous oxidation on the surface is the most significant contributor to stone decay resulting from sulfur dioxide attack. To the best of our knowledge, there is no information on the rate and mechanism of nitrogen oxides attack on limestone. It is obvious that the primary emission (NO) or even the first oxidation product (NO 2) can not attack directly the limestone due to their low dry deposition velocities, limited solubility and low acidity. The nitrogen oxides must be first converted in the atmosphere to nitrous and nitric acid before attacking the stone. The atmospheric conversion occurs via the rapid free radical termination reactions:
199
D E T E R I O R A T I O N OF J E R U S A L E M L I M E S T O N E FROM AIR P O L L U T A N T S
NITRATE FORMATION IN LIMESTONE 0.4 -O,o-AII parameters :~
o -Light excluded
~
E O.3-z~-HC excluded IJ0_ - 9 -H 20 excluded ._N O.2z
/.-I
/..,,~'~/
j r
_
o
0.1-
( -5 E
=
o'-
0
I
2
Exposure
5
4
5
6
7
period (days)
Fig. 4. The rate of calcium nitrate formation under various environmentalconditions under exposure of 300 ppm NOx. NO + H O - * H O N O NO 2 + HO--* H O N O 2
(Nitrous acid) (Nitric acid).
(1) (2)
The results o f the present study are consistent with the theory, Figure 3 shows that nitrate formation is unaffected by changes in relative humidty, unlike the pronounced humidity effect in the case o f sulfate formation (compare with Figure 2 and Table II). Secondly, the production rate decreases by a factor o f almost three in the absence of UV light (see Figure 4). Although it was to be expected that in the absence o f UV radiation the nitric and nitrous acid formation would be zero, the experimental conditions were such that certain side reactions occurred. These reactions are possible due to the high exposure concentrations employed and are not to be expected under normal polluted atmospheric conditions. The field observations indicate that significant amounts of both sulfates and nitrates are present in the surface layers (up to 1 mm in depth) with sulfate levels higher than nitrate by a factor of about five. This result is somewhat surprising since nitrogen oxide is the major pollutant in the area. Previous air monitoring data obtained near the old city wall indicated that the mean annual NO x concentration (due to the dense traffic near the wall) is about 100-120 ppb while the mean annual
200
MORDECHAI PELEG ET AL.
level is only 20 ppb [9]. The laboratory data indicate that sulfate is produced about 3 times faster than nitrate at high humidities (> 75~ and about equivalent amounts for low humidity (< 50%). Thus one would expect at least comparable levels of nitrate and sulfate in the city walls. The concentration of CaSO 4 in the city wails is much higher than can be extrapolated from laboratory simulation especially when possible wash-out effects by rain are considered. Using the laboratory results and the above mentioned environmental pollution levels, it can be estimated that it would take thousand of years of exposure at the current pollution levels, to reach the measured values of CaSO4 for the exposed sites (I, II, III and V). For the protected site (IV) it would take even more time to reach the measured sulfate content. In the case of nitrate even though a much closer comparison is observed between the extrapolated simulation studies and actual measured field values, under present ambient nitrogen levels it would still require hundreds of years for the sites to reach their present values. The differences between the laboratory observations and the field data may indicate that the damage to the stones can not be linearly expedited by increasing the concentration of the primary ambient pollutants to unrealistic simulation levels. Since the effects on the stone are caused by the secondary pollutants, a true simulation should include exposures to the actual high concentrations of H 2 S O 4 and HNO3 in the vapor (or aerosol) phase. However, it is rather difficult to simulate the actual situation due to the necessity to utilize high pollution levels, in order to establish a measuable affect in a reasonable time period. Additionally, short exposures do not allow for the slow diffusion into the stone. The fact that the sulfate is much higher than the nitrate values in the walls might be related to the higher solubility of nitrate in rain water wash out and the fact that SO2 dry deposition SO 2
TABLE V Levels (per cent) of CaSO4* and Ca(NO3) 2 in Jerusalem city wall samples Site*
May 1985 1st layer
December 1985 2nd layer
1st layer
2nd layer
NO 3-
SO42-
NO 3--
SO42-
NO 3-
SO42-
NO 3--
5042
I
0.08
0.50
0.14
0.65
0.12
0.25
0.14
0.25
II
0.05
0.24
0.13
0.26
0.12
0.30
0.20
0.25
III
0.12
0.18
0.13
0.24
0.11
0.24
0.14
0.22
IV
0.25
1.45
0.18
0.95
0.20
1.80
0.10
1.20
V
0.06
0.40
0.12
0.44
0.12
0.45
0.18
0.12
--
* Background corrected for CaSO 4. * Site locations are as follows: (I) Outside New Gate, exposed to rain; (II) Same as (I), on other side of the gate; (III) Near Jaffa Gate, exposed to rain; (IV) Inside Jaffa Gate, protected; (V) T o p of the wall near Jaffa Gate, exposed to rain.
DETERIORATION OF JERUSALEM LIMESTONE FROM AIR POLLUTANTS
201
velocity is much higher than for NO and NO2. Therefore, the sulfur sticks to the walls and is oxidized on the surface at a much higher efficiency. Comparison of the results for the samples taken at the old city walls both before and after the rainy periods (see Table V), shows no significant difference between the two sets of data indicating the extended nature of the process and the limited washout effects of the rain.
5. Conclusions This study provided further evidence that sulfur dioxide and nitrogen oxides attack limestone by different mechanisms. Laboratory simulations indicated that the presence of high (> 50~ humidity is important for the sulfur dioxide attack, while for nitrogen oxide reaction the presence of UV light and hydrocarbons are important. Field data showed the presence of much higher sulfate than nitrate in the long term exposed stone wall, inspite of the much higher nitrogen oxides ambient levels. Lower nitrate levels were explained by the slower reaction rate and greater solubility of the Ca(NO3)2 in rain water.
Acknowledgment The authors wish to thank Mr. M. Turner, the former manager, and Mr. R. Leshem, the present manager of the environmental protection division of the city for intitiating the project and providing partial financial support. We also like to thank Prof. A. Katz from the department of Geology for assisting in the development of the analytical methods and for helpfull discussion.
References [1] Cheng, R. J. and Castilo, R.: 1984, 'A Study of Marble Deterioration at City Hall Schenectady, New York', oi. Air Pollut. Cont. Assoc. 34, 16. [2] Alessandrini, G., Sala, G., Biscontin, G., and Lazzarini, L.: 1982, 'The Arch of Peace in Milan 1. Researches on Stone Deterioration', Studies in Conserv. 27, 8-18. [3] Fassina, V." 1978, 'A Study of Air Pollution and Deterioration of Stonework in Venice', Atmos. Environ. 12, 2205-2211. [4] Gauri, K. L. and Holden, G. L.: 1981, 'Pollutant Effect on Stone Monuments', Environ. Sci. Technol. 15, 386-390. [5] Luria, M., Wiesinger, A., and Peleg, M.: 1988, 'CO and NO x Levels at the Center of City Roads', J. Air Pollut. Cont. Assoc. (submitted). [6] Luria, M., Almog, H., and Peleg, M.: 1984, 'Transport and Transformation of Air Pollutants from Israel Coastal Area', Atmos. Environ. 18, 2215-2221. [7] Luria, M., David, T., and Peleg, M.: 1984, 'Five Year Air Quality Trends in Jerusalem, Israel', Atmos. Environ. 19, 715-726. [8] Luria, M. and Sharf, G.: 1985, 'The Influence of light Intensity, SO2, NOx and C3H 6 on Sulfate Aerosol Formation', J. Atmos. Chem. 2, 321-329. [9] Luria, M., Amit, U., and Peleg, M.: 1985, 'Comparison of Air Quality Data Obtained from Roof Top, Sidewalk and Suburban Areas', Environ. Monit. Assess. 5, 249-254. [10] Gauri, K. L., Doderer, G. C., Thornton Lipscomband, N., and Sarma, A. C.: 1973, 'Reactivity of Treated and Untreated Marble Specimens in an SO 2 Atmosphere', Studies in Concervation 18, 25-35.
OF J E R U S A L E M L I M E S T O N E
FROM A I R P O L L U T A N T S ; FIELD OBSERVATIONS AND LABORATORY SIMULATION M O R D E C H A I P E L E G , E S T E R B U R L A , IRIS C O H E N , and M E N A C H E M L U R I A
Environmental Sciences Division, School o f Applied Science and Technology, The Hebrew University o f Jerusalem, Jerusalem 91904, lsrael
(Received April 1988) Abstract. Samples of Jerusalem limestones were exposed to high levels of common air pollutants (SO2, NO) in the presence and absence of hydrocarbon, water vapor and ultra-violet light. After exposure, the outer layers were shaved off and analyzed for sulfate and nitrate. The results revealed that even after one day of exposure significant concentrations of CaSO4 and Ca(NO3) 2 could be detected in the external 40/zm layers. Sulfate formation was found to relate very strongly to relative humidity with nearly undetectable production at humidities below 10o7o.Nitrate formation was found to relate to UV light and to a limited extent to the presence of hydrocarbon but was unchanged at different humidities. Surface samples were taken from different sites of the old city wail and were also analyzed for the same substances. The data showed that the concentrations of especially CaSO4, and to a limited extend Ca(NO3)2, measured at various points along the city walls were higher than the expected values extrapolated from laboratory simulations. The elevated pollution content may be explained in part due to the deposition of transported ions, especially sulfate, onto the stone surface. It was concluded that although air pollution may not cause any structural damage in the foreseeable future it definitely deteriorates the fine details of the ancient monument.
A. Introduction The damage caused by air pollutants to historical stone buildings and stone monuments is well documented in the literature [1-4]. The list of famous historical sites damaged by air pollutants includes the Acropolis in Athens, Greece; the Roman forum in Rome as well as the St. Mark's Basilica in Venice. The state of Israel is rich in important stone archeological monuments many of which are located in Jerusalem. Of special importance are the temple mount which includes the Western (Wailing) Wall, the El Aksa and Omar shrines, the old city walls and the numerous newly excavated sites, some of which date back to the first Israeli commonwealth period (about 3000 years). In addition, the Jerusalem by-laws enforce the use of locally available limestone as the only material to be employed in the facade of all public and private buildings. Visible air pollution damages have been noted on many relatively new downtown buildings as evidenced by the blackening of the stone. The relation to air pollution is demonstrated by the high ambient pollution levels in this district [5], where 1.5 ppm of nitrogen oxides have been observed near the walls. The increased usage of fossil fuels has caused concern regarding the possible effects of air pollution on these stone monuments. Air quality data, available since 1979 indicated that the air quality in the city (population 500 000) is affected by both local Environmental Monitoring and Assessment 12: 191-201, 1989. 9 1989 Kluwer Academic Publishers. Printed in the Netherlands.
192
MORDECHAI PELEG ET AL.
sources, mainly vehicular transportation, as well as transportation of air pollutants from the heavily industrialized coastal region situated 50-100 km up wind of the city [6-71. In order to assess the damage to local limestone from sulfur dioxide and nitrogen oxides and the corresponding acids H2SO4 and HNO3, laboratory simulation were performed in which hundreds of exposure years were compressed into several days. Utilizing the laboratory developed methods, an attempts was made to determine the extent of air pollution influence on the ancient city wall.
2. Experimental The simulations were performed in an environmentally controlled exposure chamber previously used for photochemical aerosol generation [8]. The 18 liter cylindrical pyrex chamber was symmetrically surrounded by a bank of 4 GE dark light 40 w fluorescent lamps. The light intensity of this set up was estimated to photolyse NO2 at a rate of 5 min-1 (for comparison, NO2 photolysis rate peaks at 0.6 min-1 at noon, during the summer at mid-latitudes). In order to shorten the experimental duration it was necessary to use extremely high concentrations of gaseous pollutants so that a measurable effect could be obtained within a reasonable time scale. Pure (Matheson Inc.) sulfur dioxide and nitrogen monoxide were diluted to approximately 300 ppm each. The humidity for each experiment was controlled by introducing into the chamber saturated inorganic salt solutions of known water vapor pressure or water in open flasks. A similar method was used to obtain a fixed vapor pressure of hydrocarbon (hexene). The reaction mixture was brought up to atmospheric pressure by adding clean air. Excess heat was removed by air blowers that maintained the temperature at around 30 ~ The duration of a typical experiment was between 1 and 7 days. The average annual mean concentration of SO2 and NOx in Jerusalem (at roof top - 10 m above street level) is 0.005 and 0.025 ppm respectivly [7]. A comparison of air quality at roof top and street levels, has shown that the ambient pollutant concentrations are approximately four times larger at street levels than at roof tops [9]. Stone edifices at street levels will therfore be exposed to much higher pollution, and each laboratory simulation day will represent 10 to 40 years of natural exposure. This conclusion applies, however, only to the primary pollutants. The concentration of secondary pollutants (H2SO4, and HNO3), photochemically produced during the simulation experiment, were not monitored. It is expected, however, that the enhancement of the secondary pollutants concentration is not linearly proportional to the increased primary gases due to the increased loss of the free radicals under the experimental conditions employed. The limestone samples used for the exposure were taken from five different quarries in the immediate vicinity of the city and typical of the stone used through out the years in the area.. A 5 cm cylindrical core was drilled from the freshly mined stone block and then cut into 1 cm thick slices that were used in the
DETERIORATION OF JERUSALEM LIMESTONE FROM AIR POLLUTANTS
193
experiments. Before each experiment the stones were cleaned to remove any deposits, and then dried. After exposure the samples were prepared for ion chromatographic analysis (IC) for sulfate and nitrate. This included the removal of about a 40/tin deep layer by carefully filing away 200 mg of the surface. The resulting powder was added to 50 ml of deionized water which was treated in an ultrasonic bath for one hour. The samples were allowed to stand for 24 hours and then retreated in the bath for 15 additional minutes. The relativly high volume of water was needed to ensure that all the CaSO 4 produced would be extracted into the aqueous solution. The minimum detection limit was better than 0.1 mg per sample analysed for both substances. The study was extended to the field by implementing the same methodology. Stone samples were taken from the northwest corner of the old city walls between the New Gate and Jaffa Gate. Samples were taken from areas that are subject to direct rain washout and others in areas that are protected from direct rainfall, inside the arches of the Jaffa Gate. Details of the sample area and the traffic density in the immediate vicinity are presented in Figure 1. The sampling of the walls was performed by grinding off a designated area to a depth of 200/tm and a second layer that extended approximately 1 mm inside the stone. Due to the heterogenic nature of the stone surface the method of sectioning was of necessity crude and thus introduced divergency into the final results.
Q
I: 2 5 0 0
(vehicles per dey) Fig. 1. Map of old city wall and adjacent roads, illustrating sampling locations and traffic densities.
194
MORDECHAI PELEG ET AL.
3. Results
The parameters investigated in the present study were initial concentration of S O 2 and NO, relative humidity, radiation intensity, hydrocarbon presence and exposure time. In order to identify the rate limiting steps, the first set of experiments was performed in the presence of all the aforementioned parameters. List of initial conditions and characteristics of stone examined appears in Table I. In the succeeding experiments a given parameter was excluded in order to determine its effect on the overalI rate of sulfate and nitrate formation. 3.1. Ca(SO4) FORMATION The results of Ca(SO4) formation under the various experimental conditions are summerized in Table II. A similar rate of sulfate formation was observed in presence TABLE I Initial experimental conditions for limestone exposures Parameter
Value
SO 2 NO Relative humidity H C (hexane) Duration
280-300 p p m 290-300 p p m 100~ 50 tort 1-10 days
Limestone Characteristics Density Porosity [CasSO4]o [CaNO3]0
2.58-2.64 g m cm -3 3-4% 0.55-0.65% < 0.02~
T A B L E II Percent sulfate formation after exposure o f Jerusalem limestone to 300 p p m o f SO 2 Exp. time (days)
0
1
3
4
5
All parameters stone 1
0.6
0.6
0.8
1.1
1.25
All parameters stone 2
0.65
Light excluded stone 3
0.65
0.85
1.1
1.15
0.55
0.6
H C excluded stone 4
1.1
1.7
0.45
0.7
0.95
1.4
H20 excluded a stone 3
0.6
0.5
0.5
0.7
a Relative humidity lower than 10~
2
0.5
6
7
2.3
0.5
DETERIORATION OF JERUSALEMLIMESTONE FROM AIR POLLUTANTS
195
CoS0 4 FORMATION IN LIMESTONE vs Relative
2.4 t-
.9 2.0 13
E ~-
0 LL
1.6
9 - 0%
RH
0-25%
RH
Humidity
o - 5 0 % RH A - 7 5 % RH 9
-90%
0-100%
RH RH
1.2
0
O0
o 0.8
rj
0.4 ~
0 I 0
I 2
t.) Fig. 2.
I 4
I 6
I 8
I I0
E xposure period (doys) The production of CaSO4 resulting from exposure of limestone to 300 ppm SO2 for various time periods at different humidities,
and absence of ultra violet (UV) radiation and hydrocarbon (HC). In these experiments, a near linear relationship was obtained between exposure time and S O 4 2 f o r m a t i o n . A significant reduction in sulfate formation was observed in the experiments where the relative humidity was kept below 10%, with almost no increase in the CaSO4 content observed. An additional set of experiments was performed at different relative humidities and the results are presented in Figure 2. The data, shown in this figure, clearly demonstrates the effect of relative humidity. There is a sharp increase in the reaction rate once the relative humidity exceeds 50~ 3.2. NITRATE FORMATION
The formation of CaNO3 does not follow the same pattern as for Ca(SO4) formation, as shown in Table III. Exclusion of UV light reduced the nitrate formation by about 50~ It also appears that in the absence of HC, nitrate formation was slightly reduced. On the other hand, the presence of water did not seem to affect the reaction rate as shown in Figure 3 which illustrates the amount of nitrate produced under various relative humidities. In all experiments Ca(NO3) 2 formation showed a near linear relationship with exposure time.
196
MORDECHAI PELEG ET AL.
Ca(NO 3)2
i.0[
vs
o-
FORMATION IN LIMESTONE Relative Humidity
0% RH
o
/ o- ~OO,o . ,
,,o ~
~-
~O,o
.H
/
/
I
/~
I "-'~176176176
z 0.4 0.2
~o o
0
~,
2
E
Fig. 3.
4
6
8
I0
Exposure period (days) The production of Ca(NO3)2 resulting from exposure of limestone to 300 p p m NOx for various time periods at different humidities.
T A B L E III Percent nitrate formation after exposure of Jerusalem limestone to 300 p p m NO Exp. time (days)
0
1
3
4
5
6
All parameters stone 1
n.d.
0.09
0.13
0.22
0.28
0.34
All parameters stone 2
n.d.
0.10
0.15
0.22
0.26
0.37
Light excluded stone 3
n.d.
0.06
0.07
0.14
H C excluded stone 4
n.d.
0.04
0.10
0.18
H 2 0 excluded a stone 3
n.d.
0.08
0.19
0.17
a Relative humidity lower than 10~/0. n.d. = not detectable.
2
0.32
7
0.14
0.28
DETERIORATION OF JERUSALEM LIMESTONE FROM AIR POLLUTANTS
197
3.3. PENETRATION DEPTH A series o f experiments was performed in order to evaluate the degree o f penetration o f sulfate and nitrate into the stone. The results are presented in Table IV. In the penetration experiments a second 40 micron deep layer was shaved o f f the stone. The data show that in the case o f sulfate there was no added detectable sulfate in the second layer even after continued exposure to 300 ppm SO2 for 2 days. In the nitrate study it was possible to observe penatration of nitrate into the second layer which amounted to 25-30% of the nitrate level in the outer layer. It should be noted, however, that since the unexposed stone contained no detectable nitrate it was possible to observe an even small increase of Ca(NO3)2 which was not possible in the sulfate case. TABLE IV Penetration degree of sulfate and nitrate in Jerusalem limestone a Exp. Time
[CaSO4] 1st layer (per cents)
[CaSO4] 2nd layer (per cents)
[Ca(NO3h] 1st layer (per cents)
[Ca(NO3)2] 2nd layer (per cents)
0
0.65
0.45
n.d.
n.d.
1
0.65
0.60
0.10
n.d.
3
0.85
0.70
0.15
0.06
4
1.10
0.60
0.22
0.06
5
1.15
0.65
0.26
0.10
0.37
0.13
6
a Initial exposure concentrations, [SOz] = 300 ppm and [NO] = 300 ppm.
3.4. REACTION RATES From the above data it is possible to estimate the lower limit of the rate of formation for CaSO4 and Ca(NO3)2 in the outer surface 40 ~tm layer. This estimate assumes linearity with concentration and does not take into consideration that the rate of secondary pollutant production is not linear. At 100% relative humidity, 0.26% of sulfate is produced upon exposure to 1 ppm of SO 2 for a period o f one year and only 0.09070 at humidities below 50%. The nitrate formation rate is 0.09% y r - 1 upon exposure to 1 ppm NO under daylight conditions, and is reduced by about 50% at night-time, (the reaction rate takes also into consideration some penertration into the second layer). Thus an average value o f 0.07% ppm NO y r - l is applicable as a lower limit. 3.5. OLD CITY WALL SAMPLES Old wall samples were taken twice. Once in the early summuer (May 1985), after the end o f the rainy season and again, at the same location in December 1985. The results of the CaSO4 and Ca(NO3)2 levels present in the wails are summarized in Table V,
198
MORDECHAI PELEG ET AL.
(the CaSO 4 levels are corrected for background content of sulfate). The results can be divided into two separate groups; exposed sites - I, II, III, and V, and protected site - IV (see Figure 1 for locations). The levels in the protected site as compared to the exposed sites are 3-4 higher for sulfate and about 50~ larger for nitrate. The results for the two layers are comparable, indicating that the pollution penetrated at least several mm into the stone. 4. Discussion
The type of stone typically used in the Jerusalem area is a finely crystalline limestone, white to biege in color, whose major component is calcite (CaCO3) combined with small quantities of acid insoluble residues (5-7%) and iron oxides (> 1%). It is a relatively hard stone, although softer than marble, making it suitable for many building purposes. Since its major component is a salt of a very weak acid, it is readily attacked even by weak acids to form the corresponding salt causing destruction of the crystalline structure and hence damage to the mechanical strenght of the stone. Additionally, the salts formed are much more soluble in water and can be washed out from the stone surface. It is therefore obvious that the result of air pollution will be the formation of the slightly soluble CaSO 4 and the extremely soluble Ca(NO3) 2, both of which are easily removed from the stone surface by powdering or washout. Three mechanisms describing the formation of CaSO 4 have been proposed. At low relative humidities (< 50%), Gauri et al. (1973) [10] have suggested a direct attack of SO2 on the stone to form CaSO 3 which may further oxidize only in the presence of catalyst and water to produce CaSO4. Above 50% humidity heterogeneous SO2 oxidation occurs on the stone surface to form H2SO 4 due to the catalytic influence of the trace compounds in the stone. The newly formed sulfuric acid attacks the CaCO 3 immediately. The third mechanism involves the direct attack of the sulfuric acid, produced in the atmospheric oxidation of SO 2. In this case the relative humidity is not expected to affect the rate of attack. The results of the present study confirm the effect of relative humidity on the rate of attack. As shown in Figure 2 the reaction rate is decreased by a factor of six when the relative humidity is lowered below 50%. Since the presence or absence of UV light does not effect the reaction rates a high and low relative humdities it seems that heterogeneous oxidation on the surface is the most significant contributor to stone decay resulting from sulfur dioxide attack. To the best of our knowledge, there is no information on the rate and mechanism of nitrogen oxides attack on limestone. It is obvious that the primary emission (NO) or even the first oxidation product (NO 2) can not attack directly the limestone due to their low dry deposition velocities, limited solubility and low acidity. The nitrogen oxides must be first converted in the atmosphere to nitrous and nitric acid before attacking the stone. The atmospheric conversion occurs via the rapid free radical termination reactions:
199
D E T E R I O R A T I O N OF J E R U S A L E M L I M E S T O N E FROM AIR P O L L U T A N T S
NITRATE FORMATION IN LIMESTONE 0.4 -O,o-AII parameters :~
o -Light excluded
~
E O.3-z~-HC excluded IJ0_ - 9 -H 20 excluded ._N O.2z
/.-I
/..,,~'~/
j r
_
o
0.1-
( -5 E
=
o'-
0
I
2
Exposure
5
4
5
6
7
period (days)
Fig. 4. The rate of calcium nitrate formation under various environmentalconditions under exposure of 300 ppm NOx. NO + H O - * H O N O NO 2 + HO--* H O N O 2
(Nitrous acid) (Nitric acid).
(1) (2)
The results o f the present study are consistent with the theory, Figure 3 shows that nitrate formation is unaffected by changes in relative humidty, unlike the pronounced humidity effect in the case o f sulfate formation (compare with Figure 2 and Table II). Secondly, the production rate decreases by a factor o f almost three in the absence of UV light (see Figure 4). Although it was to be expected that in the absence o f UV radiation the nitric and nitrous acid formation would be zero, the experimental conditions were such that certain side reactions occurred. These reactions are possible due to the high exposure concentrations employed and are not to be expected under normal polluted atmospheric conditions. The field observations indicate that significant amounts of both sulfates and nitrates are present in the surface layers (up to 1 mm in depth) with sulfate levels higher than nitrate by a factor of about five. This result is somewhat surprising since nitrogen oxide is the major pollutant in the area. Previous air monitoring data obtained near the old city wall indicated that the mean annual NO x concentration (due to the dense traffic near the wall) is about 100-120 ppb while the mean annual
200
MORDECHAI PELEG ET AL.
level is only 20 ppb [9]. The laboratory data indicate that sulfate is produced about 3 times faster than nitrate at high humidities (> 75~ and about equivalent amounts for low humidity (< 50%). Thus one would expect at least comparable levels of nitrate and sulfate in the city walls. The concentration of CaSO 4 in the city wails is much higher than can be extrapolated from laboratory simulation especially when possible wash-out effects by rain are considered. Using the laboratory results and the above mentioned environmental pollution levels, it can be estimated that it would take thousand of years of exposure at the current pollution levels, to reach the measured values of CaSO4 for the exposed sites (I, II, III and V). For the protected site (IV) it would take even more time to reach the measured sulfate content. In the case of nitrate even though a much closer comparison is observed between the extrapolated simulation studies and actual measured field values, under present ambient nitrogen levels it would still require hundreds of years for the sites to reach their present values. The differences between the laboratory observations and the field data may indicate that the damage to the stones can not be linearly expedited by increasing the concentration of the primary ambient pollutants to unrealistic simulation levels. Since the effects on the stone are caused by the secondary pollutants, a true simulation should include exposures to the actual high concentrations of H 2 S O 4 and HNO3 in the vapor (or aerosol) phase. However, it is rather difficult to simulate the actual situation due to the necessity to utilize high pollution levels, in order to establish a measuable affect in a reasonable time period. Additionally, short exposures do not allow for the slow diffusion into the stone. The fact that the sulfate is much higher than the nitrate values in the walls might be related to the higher solubility of nitrate in rain water wash out and the fact that SO2 dry deposition SO 2
TABLE V Levels (per cent) of CaSO4* and Ca(NO3) 2 in Jerusalem city wall samples Site*
May 1985 1st layer
December 1985 2nd layer
1st layer
2nd layer
NO 3-
SO42-
NO 3--
SO42-
NO 3-
SO42-
NO 3--
5042
I
0.08
0.50
0.14
0.65
0.12
0.25
0.14
0.25
II
0.05
0.24
0.13
0.26
0.12
0.30
0.20
0.25
III
0.12
0.18
0.13
0.24
0.11
0.24
0.14
0.22
IV
0.25
1.45
0.18
0.95
0.20
1.80
0.10
1.20
V
0.06
0.40
0.12
0.44
0.12
0.45
0.18
0.12
--
* Background corrected for CaSO 4. * Site locations are as follows: (I) Outside New Gate, exposed to rain; (II) Same as (I), on other side of the gate; (III) Near Jaffa Gate, exposed to rain; (IV) Inside Jaffa Gate, protected; (V) T o p of the wall near Jaffa Gate, exposed to rain.
DETERIORATION OF JERUSALEM LIMESTONE FROM AIR POLLUTANTS
201
velocity is much higher than for NO and NO2. Therefore, the sulfur sticks to the walls and is oxidized on the surface at a much higher efficiency. Comparison of the results for the samples taken at the old city walls both before and after the rainy periods (see Table V), shows no significant difference between the two sets of data indicating the extended nature of the process and the limited washout effects of the rain.
5. Conclusions This study provided further evidence that sulfur dioxide and nitrogen oxides attack limestone by different mechanisms. Laboratory simulations indicated that the presence of high (> 50~ humidity is important for the sulfur dioxide attack, while for nitrogen oxide reaction the presence of UV light and hydrocarbons are important. Field data showed the presence of much higher sulfate than nitrate in the long term exposed stone wall, inspite of the much higher nitrogen oxides ambient levels. Lower nitrate levels were explained by the slower reaction rate and greater solubility of the Ca(NO3)2 in rain water.
Acknowledgment The authors wish to thank Mr. M. Turner, the former manager, and Mr. R. Leshem, the present manager of the environmental protection division of the city for intitiating the project and providing partial financial support. We also like to thank Prof. A. Katz from the department of Geology for assisting in the development of the analytical methods and for helpfull discussion.
References [1] Cheng, R. J. and Castilo, R.: 1984, 'A Study of Marble Deterioration at City Hall Schenectady, New York', oi. Air Pollut. Cont. Assoc. 34, 16. [2] Alessandrini, G., Sala, G., Biscontin, G., and Lazzarini, L.: 1982, 'The Arch of Peace in Milan 1. Researches on Stone Deterioration', Studies in Conserv. 27, 8-18. [3] Fassina, V." 1978, 'A Study of Air Pollution and Deterioration of Stonework in Venice', Atmos. Environ. 12, 2205-2211. [4] Gauri, K. L. and Holden, G. L.: 1981, 'Pollutant Effect on Stone Monuments', Environ. Sci. Technol. 15, 386-390. [5] Luria, M., Wiesinger, A., and Peleg, M.: 1988, 'CO and NO x Levels at the Center of City Roads', J. Air Pollut. Cont. Assoc. (submitted). [6] Luria, M., Almog, H., and Peleg, M.: 1984, 'Transport and Transformation of Air Pollutants from Israel Coastal Area', Atmos. Environ. 18, 2215-2221. [7] Luria, M., David, T., and Peleg, M.: 1984, 'Five Year Air Quality Trends in Jerusalem, Israel', Atmos. Environ. 19, 715-726. [8] Luria, M. and Sharf, G.: 1985, 'The Influence of light Intensity, SO2, NOx and C3H 6 on Sulfate Aerosol Formation', J. Atmos. Chem. 2, 321-329. [9] Luria, M., Amit, U., and Peleg, M.: 1985, 'Comparison of Air Quality Data Obtained from Roof Top, Sidewalk and Suburban Areas', Environ. Monit. Assess. 5, 249-254. [10] Gauri, K. L., Doderer, G. C., Thornton Lipscomband, N., and Sarma, A. C.: 1973, 'Reactivity of Treated and Untreated Marble Specimens in an SO 2 Atmosphere', Studies in Concervation 18, 25-35.
