Environmental mud adhesion on optical glass surface: Effect of mud drying temperature on surface properties

Solar Energy 150 (2017) 73–82

Contents lists available at ScienceDirect

Solar Energy journal homepage: www.elsevier.com/locate/solener

Environmental mud adhesion on optical glass surface: Effect of mud drying temperature on surface properties B.S. Yilbas a,b,⇑, H. Ali b, A. Al-Sharafi a, N. Al-Aqeeli a a b

ME Department and Center of Excellence in Renewable Energy, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia ME Department, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia

a r t i c l e

i n f o

Article history: Received 19 September 2016 Received in revised form 16 April 2017 Accepted 17 April 2017

Keywords: Environmental dust Mud formation Humid air Glass surface Optical transmittance

a b s t r a c t Mud formation in humid air ambient from dust particles on the glass surfaces is investigated. Mud drying temperature and the characteristics of mud liquid created at the interface between wet mud and the glass surface are examined. Effects of dried mud liquid layer on the glass surface are assessed using analytical tools including scanning electron and atomic force microscopes, energy dispersive spectroscopy, X-ray diffraction, and Fourier-transform infrared spectroscopy. Dry mud adhesion on the glass surface is evaluated incorporating the micro-tribometer and UV visible transmittance tests are carried out to determine optical properties of the glass substrate after dry mud removal. It is found that mud drying temperature has significant effect on the glass surface chemistry and topology; in which case, increasing mud drying temperature increases OH
and KOH attack on the glass surface while forming cavities-like structures and randomly distributed few micro-cracks. The tangential force required for dry mud removal from the glass surface increases with increasing mud drying temperature. Transmittance of UV visible spectrum is suppressed by the mud residues on the glass surface, which is more pronounced at high temperatures. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Climate change caused frequent occurrence of severe weather conditions in recent years, particularly in the Middle East, in terms of regular dust storms, increased humidity, and high air temperatures, which remain well above the yearly averages. Regular storms in desert environments carry the dust particles to urban areas while causing damages and performance degradation in energy harvesting systems such as PV panels, concentrated solar receivers, etc. (Semaoui et al., 2015). Creating self-cleaning surfaces, via mimicking the nature, provides the promising solution to the problem of dust accumulation in dry air conditions (Lin et al., 2016). The dust particles settle at the surface while forming a dust layer with varying sizes and compositions (Yilbas et al., 2016a). Some of the dust compounds are chemically active and cause local damages on the surfaces as well as alter the optical properties of the settled surface. Some of these surface properties are critical for transmittance, absorption, and reflection of the incident optical radiation. In humid environments, such as regions close to the seashores - as it is the case in the Arabian Gulf, water ⇑ Corresponding author at: ME Department, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia. E-mail address: [email protected] (B.S. Yilbas). http://dx.doi.org/10.1016/j.solener.2017.04.041 0038-092X/Ó 2017 Elsevier Ltd. All rights reserved.

vapor condensate on the dust surfaces and gives rise to dissolution of some dust compounds, such as alkaline (Na and K) and alkaline earth metals (Ca) compounds. The solution composing of dissolved compounds and water remains chemically active and possesses the state of high bases that is high pH (Yilbas et al., 2016b). Since the dust particles consist of different sizes and shapes, when they settle at the surface, they form like pores structures with high permeability (Yilbas et al., 2016b). This provides access to the liquid solution, composing of dissolved compounds and water, passing among the dust particles and accumulating at the surface under the gravity. As the rate of condensation increases in humid air environments, water content increases on the dust particles while increasing the amount of dissolution of dust compounds. This further increases the bases state of the liquid solution and gives rise to a liquid film formation at the interface between the wet dust particles and the solid surface. As the air humidity reduces, the liquid film dries out and forms intermediate layer at the interface. The dried liquid layer not only alters the optical characteristics of the surface, but also increases the adhesion between the accumulated dust particles and the solid surface (Yilbas et al., 2016b). The effect of the dry liquid solution on the properties of the solid surface and the adhesion of the accumulated dust particles can change with air temperature due to changing of the drying rate at the surface. Consequently, investigation of the effect of dried solution, which is

74

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82

formed at different air temperatures, on the properties of the glass surface and adhesion characteristics of the accumulated dust particles becomes essential. Considerable research studies were carried out to examine the dust accumulation on the surfaces and its effects on the performance of the energy harvesting devices. Paudyal and Shakya (2016) highlighted the importance of dust accumulation on PV panel modulus. The dust accumulation at the bottom of the PV modules resulted in a high risk of hot spots, which could eventually lead to permanent module damage. Rifai et al. (2016) presented the removal mechanism of the dust particles from the rotating disk incorporating the centrifugal forces. The force balance revealed that the centrifugal forces remained higher than the adhesion, friction, drag, lift, and gravitational forces in the region away from the rotational center. In addition, the dust particle size and rotational speed significantly influenced the rate of dust removal from the disk surface. Zaihidee et al. (2016) studied the degradation of PV panel surfaces and the findings revealed that the degradation depended mainly on the dust deposition density, which was governed by the various factors. The dust accumulation of 20 g/m2 on a PV panel reduced a short circuit current, open circuit voltage and efficiency by 15–21%, 2–6% and 15–35%, respectively. AlShehri et al. (2016) investigated the cleaning of photovoltaic panel from dust particles. It was demonstrated that cleaning efficiency of the nylon brushes was not as high as the cleaning by using water and delicate wipers. Moreover, some small damages were observed on the glass surfaces after brushing; however, this was shown not to have a permanent effect on the optical characteristics of the glass. Sarver et al. (2013) performed a comprehensive overview of soiling problems, primarily those associated with the dust particles (sand) and combined dust–moisture conditions. They reviewed the key indicators of dust effects on the device performance. Klimm et al. (2016) examined the soiling behavior of the surfaces due to dust accumulation. The soiling decreased the transmittance while limiting the overall performance of the solar devices. Qian et al. (2012) developed a scaling analysis to identify a single key dimensionless parameter influencing photovoltaic panel efficiency and determined the optimal values for turn-on and turn-off voltage ratio in terms of the parameters selected. The findings revealed that the efficiency decreased rapidly as the turn-off voltage ratio was raised above the optimal value. Yilbas et al. (2016c) studied dust and mud accumulation of laser textured and sol-gel coated alumina surfaces. The laser treated and sol–gel coated alumina surfaces provided superior surface characteristics in the harsh environments because of the weak adhesion between the mud formed from the dust particles and the coating surface. This was associated with the small texture height of the sol–gel coating, which lowered the area of the interfacial contact between the mud and the coated surface, and relatively lower surface energy of the sol–gel coating as compared to that of the laser treated surface. Dastoori et al. (2016) examined the impact of the static electric charge of the accumulated dust particles on the photovoltaic module performance. The charge level of the accumulated dust particles had significant impact on photovoltaic module output and the dust particles accumulation was not strongly associated with panel tilt angle. Sueto et al. (2013) studied the effects of anti-soiling photocatalytic coating on surfaces characteristics. The findings revealed that the presence of electrostatic charges on the surfaces was the main factor for the adhesion of sand, and it could be suppressed by the anti-soiling photocatalytic layer. Elminir et al. (2006) demonstrated the reduction of transmittance due to surface soiling. In this case, the dust deposition density, in conjunction with the plate tilt angle, and the orientation of the surfaces, with respect to the dominant wind directions, influenced significantly the surface soiling. Tanesab et al. (2015) showed the type of dust particles deposited on PV module surface. The dusts were

dominated by the fine particles of quartz (SiO2), followed by calcium oxide (CaO) and some minors of feldspars minerals (KAlSi3O8). In this case, quartz particles were the main contributors for lowering the transmittance and reducing the PV module performance. Klugmann-Radziemska (2015) investigated degrading performance of a crystalline photovoltaic module due to the dust deposition on the surface. The findings revealed that performance loss was closely related to the tilt angle of the module, the exposure period, site climate conditions, wind movement, and the dust properties. In addition, the energy yield losses took place due to the dust deposition on the photovoltaic modules. Smallwood et al. (2016) introduced a finite element modeling to compute the adhesion factor of the particles for dielectrophoretic adhesion. In the analysis, the density of particles and the dielectric constant were incorporated. Sayyah et al. (2014) reported the degradation performance of the solar thermal system. The analysis was introduced for the advantages of cleaning processes that included natural, manual, automatic, and passive methods. The type of solar collectors, geographical location, local climate, and exposure period of the collectors influenced the efficiency of the solar thermal system. Yilbas et al. (2015) investigated laser texturing of alumina surfaces and dry mud effect on the textured surfaces. The laser texturing increased the microhardness and enhanced surface hydrophobicity due to the formation of nitride species. The mud residues did not influence the fracture toughness and microhardness of the laser textured surface; however, they reduced the surface hydrophobicity significantly. Hegazy (2001) showed, in dusty environment, the surface transmittance strongly related to the dust deposition in conjunction with plate tilt angle, as well as on the exposure period and site climate conditions. In this case, the empirical correlation was developed for the reduction of transmittance of a glass wafer with a fixed tilt angle. Beattie et al. (2012) described the reduction of the active surface area due to the formation of clusters of particles on the surface. Such clusters could support particles in upper layers which reduced the available area for photon capture by a much smaller amount than the particles resting directly on the glass surface. Javed et al. (2017) characterized the dust particles accumulated on photovoltaic panel surfaces over a period of ten months in a solar test facility located in Doha, Qatar. The findings revealed that the dust collected after dust-storm events had higher proportions of halite and quartz contents than non-dust-storm days, depending on the direction of the wind. Also, dust particles accumulated on photovoltaic panels appeared to agglomerate as the exposure time increased. Abderrezek and Fathi (2017) demonstrated the importance of dust type on the photovoltaic panel performance; in which case, variation of the physical parameters including level of optical transmittance and the glazing temperature resulted in photovoltaic panel performance change. Kazem and Chaichan (2016) examined the dust depositions on photovoltaic modules. It was demonstrated that the weight and shape of dust particles had a significant effect on their deposition behavior while influencing the photovoltaic device performance. Although dust accumulation and mud formation on glass surfaces in humid air environments was studied earlier (Yilbas et al., 2015, 2016a, 2016b), the main focus was the mud formation in standard air temperature on the glass surfaces (Yilbas et al., 2015) or mud formation on the polycarbonate surface (Yilbas et al., 2016a, 2016b). However, the effect of air temperature on the dry mud adhesion and its after effects on the glass surfaces were left for future study. In the present study, the dust accumulation and the mud formation on the glass surfaces at different air temperatures are investigated. The properties of mud the solution and its influence on the characteristics of the glass surface are assessed using the analytical tools including scanning and atomic force microscopes, X-ray diffraction, energy dispersive spec-

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82

troscopy, UV visible transmittance, and Fourier-transform infrared spectroscopy. The mud adhesion on the glass surface is evaluated incorporating the micro-tribometer.

2. Experimental set up The glass samples with dimensions of 30 mm  40 mm  2 mm (width  length  thickness) were used as workpieces. The chemical composition of the glass was 76.5% SiO2, 9.9% CaO, 1.2% MgO and 12.4% Na2O. The dust was collected in the area of Dhahran in Saudi Arabia after a dust storm in 2015. The dust particles were characterized using SEM (A JEOL 6460), EDS, and XRD (Bruker D8 Advanced diffractometer). The roughness measurements and surface profile characterization were performed using a 5100 AFM/ SPM Microscope by Agilent in contact mode. The probe tip was made of silicon nitride (r = 20–60 nm) with a manufacturer specified force constant, k, of 0.12 N/m. A linear microscratch tester (MCTX-S/N: 01-04300) was used to record the tangential force required to remove the dry mud from the glass surfaces. The contact load was set at 0.03 N, and the end load was set at 2.5 N. The scanning speed was 5 mm/min, and the loading rate was 0.01 N/s. The total length for the scratch tests was 0.5 mm. The optical transmittance was measured using a UV spectrometer (Jenway – 67 Series spectrophotometer), and Fourier transform infrared spectroscopy (Bruker – VERTEX70) was performed to collect the infrared absorption spectrum of the glass. To investigate the effects of dust and mud on the surface characteristics of the glass, actual dust accumulation and mud formation were simulated in a laboratory environment. In actual environments, the mud was formed from accumulated dust particles due to the condensation of water vapor onto the particles. The accumulated dust thickness was measured over the period of two weeks during a dust storm in Saudi Arabia in 2015. This accumulation was found to be on the order of 300 lm. To simulate the dust accumulation in the laboratory, 300-lm layer of the dust particles collected from the local environment were deposited onto the cleaned glass surfaces. Desalinated water, which was equal to the amount of water vapor that condensed on the same volume of the dust in the open environment, was dispensed gradually onto the dust layer. The initial condensation tests were performed in ambient humid air to estimate the amount of condensate that accumulated over time. The dispensed water and the dust layer were left without mechanical mixing to resemble water condensation in the humid air. Therefore, the simulated formation of the wet mud on the glass surfaces was similar to the deposition that occurred naturally in the open environment. Next, the glasses were kept in ambient air at different temperatures including 30 °C, 40 °C, 50 °C, 60 °C, and 70 °C. for three days to dry. It should be noted that the maximum module temperature in Saudi humid days under the sun was in the order of 70 °C. The scratch tests were performed to measure the tangential force required to remove the mud from the glass surfaces. The tangential force provided information regarding the adhesion, cohesion and frictional work during the dry mud removal. The tests were repeated 15 times to secure the confidence levels for the experimental uncertainty assessments. Based on the distribution of the experimental data, the confidence level of 95% was resulted; in which case, the mean (l) of the data distribution was within ±1.75 of the standard deviation of the distribution of a single measurement from that distribution. The experimental uncertainty analysis revealed that the uncertainty less than 3% was resulted for the tangential force measurements. To assess the after-effects of the mud on the glass surfaces, the dry mud was removed from the glass surfaces using a desalinated

75

water jet that was 2 mm in diameter with a velocity of 2 m/s. The cleaning process was applied for 15 min to each glass surface. Finally, the morphology, optical transmittance, and molecular characteristics of the mud removed glass surfaces were analyzed using the analytical tools. The microhardness, friction and adhesion tests were repeated 12 times to secure the confidence levels for the experimental uncertainty assessments. Based on the distribution of the experimental data, the confidence level of 95% was resulted; in which case, the mean (l) of the data distribution was within ±1.75 of the standard deviation of the distribution of a single measurement from that distribution. The experimental uncertainty analysis revealed that the uncertainty less than 2% was resulted for the microhardness measurements while the uncertainty of about 3% was obtained for the friction and adhesion tests. The surface energy of the dry muds formed at different drying temperatures is measured incorporating the contact angle experiments. Kyowa (model - DM 501) goniometer was used to determine surface free energy of the glass surface after mud removal. In this case, water, Glycerol, and Diiodomethane were used (van Oss et al., 1990). The surface energy data for Lifshitz-van der Walls components and electron-donor parameters were given in Table 1. In addition, the surface energy and Lifshitz-van der Walls components and electron-donor parameters determined for dry mud removed from the glass surfaces for different mud drying temperatures are given in Table 2. However, the details of the mathematical analysis could be found in the previous study (van Oss et al., 1990).

3. Results and discussion The dust characteristics and the mud formation on the glass surfaces at different air temperatures are examined. The dust and the dry mud characteristics are investigated using analytical tools. The influence of the dry mud solution on the adhesion between the dry mud and the glass surface was assessed via tangential force measurements. Fig. 1 shows SEM micrographs of dust particles located at the glass surface prior to the mud formation. The dust particles vary in sizes and shapes (Fig. 1(a)) and the averaged size of the dust particles is in the order of 1.2 lm. In general, particle shapes can be grouped into quadrangular, round, and flake-like features (Fig. 1 (b) and (c)). Small size dust particles (<1 lm) generally attach to the large size particles (Fig. 1(d)). This is associated with the electrostatic charge on small particles, which appear as bright color in SEM micrograph (Fig. 1(b)) due to electron charging. The possible explanation of the electrostatic charges is associated with the prolonged residing duration of the small particles in the atmosphere. They interact with solar radiation for longer durations than the large particles. Consequently, prolonged exposure to the atmosphere in regions close to the sea, such as Arabian Gulf, causes the attachment of ionic compounds to small dust particles. The EDS data for the dust particles are given in Table 3, which demonstrate that oxygen, iron, sulfur, chlorine, calcium, silicon, sodium, magnesium, and potassium are present in the dust particles. The weight percentage of the elemental concentration changes with different shapes of the dust particles. The quadrangular particle, which could be result of the cubic particle deformation, is rich in sodium and chlorine. The agglomerated small particles are rich in calcium and oxygen. The flake-like particles are rich in calcium and silicon. In addition, the chlorine concentration varies in the dust particles; in which case, the molar ratio of NaCl is not satisfied. Therefore, the dust particles do not possess the salt crystals, but NaCl in the compound form. The presence of sulfur may be related to formation of a monomer layer during the aging process

76

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82

Table 1 Lifshitz-van der Walls components and electron-donor parameters used in the simulation (van Oss et al., 1990). hL

cLL (mJ/m2)

cþL (mJ/m2)

c
L (mJ/m2)

Mud drying temperature = 40 °C Water Glycerol Ethylene Glycol

36.2 35.9 36.8

72.8 64.1 48.1

25.5 3.92 0.41

21.8 34.1 19.1

Mud drying temperature = 70 °C Water Glycerol Ethylene Glycol

44.6 36.2 30.3

72.8 64.3 39.8

25.5 3.92 0.41

21.8 33.8 18.9

Table 2 Surface free energy of polycarbonate cover after mud removal (mJ/m2). Mud drying temperature = 40 °C Mud drying temperature = 70 °C

58.1 64.6

in the atmosphere. In addition, it can be associated with the calcium, such as the anhydrite or gypsum component (CaSO4) in the dust. The iron is most likely related to clay-aggregated hematite (Fe2O3). Fig. 2(a) shows the X-ray diffractogram of the dust particles and dry mud formed on the glass surface while Fig. 2(b) shows the X-ray diffractogram of the dry mud solution. The presence of potassium, sodium, calcium, sulfur, chlorine, and iron peaks are evident. The iron, the aluminum, and silicon peaks are coincident with each other. The sodium and potassium peaks in the diffractogram are related to the sea salt due to the close region of Arabian Gulf where the dust particles were gathered. The metals are found to be in oxide compounds in dust particles, such as Fe2O3, FeO, and MgO. Fig. 3 shows SEM micrograph of the dry mud surface (Fig. 3 (a) and (b)) and the dry mud cross-section (Fig. 3(c) and (d)), which are formed at two temperatures, 40 °C and 70 °C, on the glass surface. The dry mud surfaces possesses large size dust particles sur-

rounded by dry mud, which compose of fine size dissolved dust particles (Fig. 3(a)). In addition, small holes/voids at the surface are formed, which lead the formation of the porous like structures inside the dry mud (Fig. 3(d)). The formation of small holes/voids is attributed to the space created by the mud liquid after flowing towards the mud-solid surface interface under the gravity. It should be noted that the liquid solution flows through the wet mud and forms a thin layer at the interface between the wet mud and the glass surface due to gravity. Once the mud dries, some small holes/voids become apparent where the liquid solution left behind in the mud. This situation can also be observed in micrographs of mud-cross section (Fig. 3(d)). In this case, the liquid solution forms small channels among the undissolved mud particles while it is flowing towards the interface of the wet mud and the glass surface. After the mud dries, these small size channels become apparent within the dry mud cross-section as small cavities. Since orientation of the dust particles in the wet-mud does not form a regular pattern, the small size channels appear as irregular and partially joined cavities across the dry mud cross-section. A layer formed by the liquid solution appears as the bright region at the interface upon drying. In addition, some of the liquid solution residues also dry out in the small channels and they appear as locally scattered bright regions at cross-section of the dry mud (Fig. 3(d)). The dry liquid solution at the interface does not have

Fig. 1. SEM micrographs of dust particles: (a) Different shapes and sizes of dust particles, (b) rectangular and circular shapes of dust particles, (c) flake like dust particles, and (d) small size dust particles attach large particle surface.

77

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82 Table 3 Elemental composition of dust (wt.%) determined by energy dispersive spectroscopy (EDS). Si

Ca

Na

S

Mg

K

Fe

Cl

O

11.4

7.9

3.4

1.6

2.8

1.1

1.2

0.8

Balance

Fig. 2. X-ray diffractograms of (a) dust particles, dry mud surface, and (b) dry mud solution for two temperatures.

uniform thickness because of the local contact area of the dust particles on the glass surface (Fig. 3(c)). Nevertheless, it covers almost all the surface of the glass at the interface. The effect of air temperature on the morphology of the dry mud is evident; in which case, the thickness of the dried liquid solution becomes slightly thicker because of the fast drying rate at high temperature. It should be noted that thermal conductivity of the glass is in the order of 0.1 W/mK (Chung, 2001), which is much smaller than that of metals, and the thick layer of liquid solution at the interface gives rise to a slow cooling rate. In order to assess the dried liquid characteristics on the glass surface, the liquid solution is extracted from the dust particles and dispensed onto the glass surface. Later, it is left for drying over 48 h under the constant air temperature condition according to temperature range at which the wet mud is dried. The pH of the mud solution is measured and it is observed that the pH

increased to 7.5 within six hours. This increase is attributed to the dissolution of alkaline and alkaline earth metals in water. In addition, OH
ions are responsible for high pH level of the liquid solution due to the dissolution of the alkaline and alkaline earth metallic compounds in water. The potentiodynamic tests using the mud liquid are not conducted to assess the chemical activity of the mud solution; however, it was reported previously (Zhang et al., 2009) that the mud solution has chemically active characteristics (Zhang et al., 2009). Fig. 4 shows SEM micrographs of the top surface of dried mud solution at two temperatures. Solid crystals with various sizes are formed at the glass surface for both temperatures (Fig. 4(a) and (b)). In general, the crystal size reduces with increasing temperature, which may be attributed to the drying rate; in which case, the size of the crystals remains smaller at high temperature (Fig. 4(a)) than that of at low temperature (Fig. 4(b)).

78

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82

Fig. 3. SEM micrographs of dry mud surface and dry mud cross-section: (a) dry mud surface, (b) holes on dry mud surface (marked in red circle), (c) film of dried liquid solution at dry mud and glass interface, and (d) cavity filled by the dried liquid solution (marked in red circle). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4. SEM micrographs of dried liquid solution forming crystals: (a) crystals formed at the glass surface at 70 °C and (b) crystals formed at the glass surface at 30 °C.

The EDS data for the dry mud solution are given in Table 4. The presence of Na, K, Ca, and Cl in the data demonstrates that alkaline and alkaline earth metals as well as chlorine dissolve in water while increasing the pH level of the mud liquid. Fig. 5 shows SEM micrographs of the glass surfaces after the dry mud removal by using a pressurized desalinated water jet. The mud residuals are observed on the glass surface after the dry mud removal, which is true for all temperatures at which the dry muds are formed on the glass surface (Fig. 5(a)–(d)). The average coverage area of the dry mud residues is in the order of 6% and the coverage area of the mud residues increases with increasing temperature. In this case, the coverage area becomes 8% at 70 °C while it is only 5% at 30 °C. The increase in the coverage area is

associated with the adhesion of the mud residues at the glass surface, which increases with increasing temperature. This behavior is associated with the rate of evaporation of water from wet mud during the drying process. Once water evaporates from the wet mud during the drying, the volume shrinkage takes place, which increases with increasing evaporation rate at high temperatures. Consequently, some small crystals are formed in the dried mud liquid at the interface between the dry mud and the glass surface. The volume shrinkage results in increased stress levels in the dry mud while enhancing the adhesion among the dust residues and the glass surface along the interface (Style et al., 2011). However, some small cavity-like textures are observed at the glass surface after the dry mud removed. This situation is also observed from Fig. 5(e).

Table 4 EDS data (wt.%) for the elemental composition of the dried mud solution on the glass surface. Ca

Na

S

Mg

K

Fe

Cl

O

6.3

0.8

1.2

1.7

0.5

0.6

0.1

Balance

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82

79

Fig. 5. SEM micrographs of glass surface after dry mud removal by a desalinated water jet: (a) dry mud residues at 70 °C, (b) close view of dry mud residue at 70 °C and no crystals are evident, (c) dry mud residues at 30 °C, (d) close view of dry mud residues and crystals due to dried liquid solution is evident at 30 °C, (e) close view of mud residue and small voids around it at 70 °C, and (f) micro-crack at the surface at 70 °C.

The size of the damage site on the glass surface increases slightly with temperature. The presence of the damaged sites on the glass surface can be explained in terms of OH
ions in the mud liquid (pH = 7.2) at the interface prior to drying. In this case, prior to complete drying of the mud liquid at the interface, glass surface is possibly etched by KOH ions, which is present in the mud liquid. The local etching causes small cavity-like structures at the surface and enhances the surface texture of glass. This can be seen from (Fig. 5 (e) and (f)) of glass surface after the dry mud removed. The ion exchange mechanism can take place between the glass-alkaline ions at the interface as (García-Heras et al., 2005) SiAO
. . . K+(OH)
water ? „SiAOH + K+ + OH
. The breakdown of the glass network in the surface region, via alkaline attack because of OH
ions, causes the reaction SiAOASi„ + OH
„SiAOH +
OASi„. In addition, KOH attack on glass surface causes the formation of silanol groups („SiAOH, which can cause micro-cracks formation in the close region of glass surface (Fig. 5(f)). During the process of alkaline attack, OH
ions can break siloxane bonds („SiAOASi„) on the glass surface while favoring water molecules (H+ and OH
ions) penetration into the glass surface. This, in turn, results in the small cavity-like structures in the surface region of the glass. On the other hand, KOH attack also causes diffusion of potassium ions into glass surface while resulting in volume change due to the large diameter of potassium. This gives rise to chemical tough-

ening of the glass surface and changes the optical properties of the glass surface. Fig. 6(a) and (b) shows AFM image and line scan of the surface after the dry mud was removed by a water jet for two mud drying temperatures. Increasing mud drying temperature enhances the mud residuals at the surface. In addition, AFM line scan of the glass surface reveals that the average roughness of the glass surface after dry mud removal is in the order of 0.5 lm–0.6 lm depending on mud drying temperature, i.e. increasing temperature reduces the surface roughness. This is related to the crystal sizes formed at the surface, which reduces with increasing mud drying temperature. Fig. 7 shows the Fourier-transform infrared spectroscopy data for the glass surface after the dry mud removed. The data for the bare glass are also shown for the comparison reason. The bare glass surface demonstrates peaks on 550–600 cm
1 region, which correspond to SiAOASi bending vibration (Khalil et al., 2010). In addition, the stretching vibration of SiAOASi takes place at approximately 910 and 990 cm
1 (Khalil et al., 2010). In the case of dry mud removed glass surface, the peaks slightly different appearance than those corresponding to the bare glass. This difference can be explained in terms of the stress developed in the close region of the dry mud removed glass surface. In this case, stresses are formed because of the attack by the alkali and alkaline earth

80

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82

Fig. 6. AFM 3-dimensional images of glass surface after dry mud removal and line scan of the mud residues on glass surface: (a) mud drying temperature is 70 °C, and (b) mud drying temperature is 30 °C.

Fig. 7. FTIR graph for bare glass, dried liquid solution, and after mud removed glass with mud drying temperatures of 30 °C and 70 °C.

hydroxides and the diffusion of potassium in the surface region of the mud removed glass surface. The stretching vibration of nonbridging oxygens (SiAO
) also appears at 920 cm
1, which is only observed for the dry mud removed glass surface. The presence of this peak is associated with the ionic bonding between the nonbridging oxygen and the network modifiers. Consequently, the glass network dissolution and new SiO2-enriched network formation cause the stretching vibration of the non-bridging oxygen (Xiong et al., 2010). In the case of the FTIR spectrum of mud removed surface, it results peaks in the range of 3070– 3580 cm
1, which corresponds to hydroxyl groups (AOH). Fig. 8 shows the tangential force required to remove the dry mud from the glass surface. The tangential force composes of adhesion, cohesion, and frictional forces, and it corresponds to the force required to remove the dry mud from the glass surface. Moreover, several forces are involved with the adhesion of the dry mud onto the glass surface, such as van der Waals and electrostatic forces (Fraunhofer, 2012). The dried liquid solution at the interface between the dry mud and the glass surface consists of dissolved

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82

Fig. 8. Tangential force variation along the scratch length for the cases bare glass, dry liquid solution, and dry mud due to two drying temperatures, 30 °C and 70 °C.

alkaline metals, alkaline earth metals, and other ions. It should be noted that the assumption of only strong van der Waals forces contributing to the adhesion force may not be realistic due to the covalent bonds formed by the dried liquid solution at the interface. On the other hand, there is no dry mud adhesion on the bare glass surface; therefore, the tangential force measured is due to the friction effect only. The averaged tangential force required to remove the dry mud from the glass surface is higher than the frictional force obtained for the base glass. The tangential force increases with increasing mud drying temperature. This behavior is associated with the strong adhesion between the dry mud on the glass surface; in which case, the modified surface texture by small cavitylike structures and KOH attack increases surface area and causes strong bonding of the dry mud liquid at glass surface. Since the dry mud liquid forms interface between the dry mud and glass surface, it enhances the adhesion and increases the tangential force. The dried liquid solution gives rise to the highest tangential force, which indicates the strong adhesion between the dried liquid solution and the glass surface. It should be noted that the tangential force measurements are repeated five times, and the experimental error is estimated as in the order of 5%. Fig. 9 shows the transmittance data of the dry mud removed glass surface. The bare glass the transmittance is also shown for the comparison. The dry mud removed glass surface demonstrates the absorption of incident UV visible spectrum while resulting in about 15% reduction in the transmittance when mud is dried at 30 °C. Moreover, increasing mud drying temperature to 70 °C further increases the transmittance loss, which is in the order of 60%. The reduced

Fig. 9. Transmittance of glass after mud removal by a desalinated water jet for the cases mud drying temperatures 30 °C and 70 °C. Transmittance data for liquid solution at two temperatures are shown together with bare glass for comparison.

81

transmittance of the dry mud surface is associated with one or all of the followings: (i) presence of the mud residues at the glass surface after the dry mud is removed; in which case, the mud residues partially block the incident UV visible radiation, (ii) the structural and topological changes at the surface because of the ion attacks due to the mud liquid presence at the interface, and (iii) presence of dry liquid solution residues at the surface after desalinated water jet cleaning. It should be noted that optical transmittance reduces significantly by the dried liquid solution and the reduction is in the order of 50% for the liquid solution drying temperature 30 °C and over 80% for drying temperature 70 °C. Since direct relation is present between the optical transmittance and photovoltaic panel efficiency (Ghazi et al., 2013), dry mud formation on the surface reduces the device efficiency by a factor of almost 50%, which is similar to the finding of the previous study (Ghazi et al., 2013). This is because of significant reduction in the optical transmittance due to dry mud formation in humid air ambient. In the case of high temperature (70 °C) of mud drying, the panel efficiency loss increases almost 65%, which is higher than that of the dry mud formed at low ambient temperatures. However, cleaning the dry mud from the surface by a water jet improves the efficiency; however, full recovery of the efficiency by a water jet cleaning is less likely. Consequently, the dry mud has significant effect on the photovoltaic panel efficiency, which is more pronounced when the mud is dried at high ambient temperatures.

4. Conclusion The mud formed from the environmental dust particles in humid air environments is investigated and its effects on the properties of glass surface are analyzed. In humid environments, water condensates on the dust particles forming a wet mud on the glass surfaces. Since the wet mud drying at different temperatures can demonstrate different adhesion characteristics, the effect of mud drying temperature on the dry mud adhesion characteristics is examined. The mud liquid extracted from the wet mud is analyzed and the influence of the mud liquid on the chemistry and topology of the glass surface are investigated incorporating the analytical tools including scanning electron and atomic force microscopes, energy dispersive spectroscopy, X-ray diffraction, and Fouriertransform infrared spectroscopy. The tangential force required to remove the dry mud from the glass surface is measured using the micro-tribometer. Optical properties of the dry mud removed glass surfaces are assessed through UV visible transmittance tests. It is found that the dust particles poses alkaline (Na, K) and alkaline earth metal (Ca) compounds, metallic oxides (MgO, FeO, Fe2O3, etc.), and chlorine. They vary in size and shapes, and the average size of the dust particles is in the order of 1.2 lm. When water condensates on the dust particles, due to air humidity, alkaline and alkaline earth metal compounds dissolves in water and forms the mud solution, which gradually settles on the glass surface due to gravity. The mud liquid forms a layer between the wet mud and the glass surface. Since the mud liquid is chemically active, the ion attacks cause structural deformation of the glass surface. In addition, drying temperature of the wet mud changes the characteristics of the glass surface. In this case, OH
and KOH attack causes formation of small size cavity-like structures and some minor micro-cracks on the glass surface. These surface defects are more pronounced at high mud drying temperatures. The lateral force, to remove the dry mud from the glass surface, increases with increasing mud drying temperature. This behavior is attributed to strong adhesion between the dry mud and the surface because of the presence of the dried mud liquid at the interface. Mud residues are observed after removal of the dry mud from the glass surface

82

B.S. Yilbas et al. / Solar Energy 150 (2017) 73–82

by using a desalinated water jet. The coverage area of the dry mud residuals on glass surface is in the order of 4% for low mud drying temperature (30 °C) and it increases to 8% for high drying temperature (70 °C). The UV visible transmittance of the dry mud removed glass surface reduces with increasing mud drying temperature. In addition, optical transmittance reduces significantly through the dried liquid solution, which is more pronounced at high mud drying temperature (70 °C). Consequently, mud drying temperature has significant effect on the mud adhesion on the glass surface and its after effects are also observed to be significant in terms of chemistry and topology of the dry mud removed glass surface, which result in low UV visible transmittance. The study demonstrates the importance of environmental temperature on the mud formation from the environmental dust particles on the glass surfaces in humid air ambient. It also provides information on the mud after effects on the glass surfaces. Acknowledgements The authors acknowledge the financial support of King Fahd University of Petroleum and Minerals (KFUPM) through Project# RG 1406, and King Abdulaziz City for Science and Technology (KACST) through project# 11-ADV2134-04 to accomplish this work. References Abderrezek, M., Fathi, M., 2017. Experimental study of the dust effect on photovoltaic panels’ energy yield. Sol. Energy 142, 308–320. Al-Shehri, A., Parrott, B., Carrasco, P., Al-Saiari, H., Taie, I., 2016. Impact of dust deposition and brush-based dry cleaning on glass transmittance for PV modules applications. Sol. Energy 135, 317–324. Beattie, N.S., Moir, R.S., Chacko, C., Buffoni, G., Roberts, S.H., Pearsall, N.M., 2012. Understanding the effects of sand and dust accumulation on photovoltaic modules. Renew. Energy 48, 448–452. Chung, D.D.L., 2001. Materials for thermal conduction. Appl. Therm. Eng. 21, 1593– 1605. Dastoori, K., Al-Shabaan, G., Kolhe, M., Thompson, D., Makin, B., 2016. Impact of accumulated dust particles’ charge on the photovoltaic module performance. J. Electrostat. 79, 20–24. Elminir, H.K., Ghitas, A.E., Hamid, R.H., El-Hussainy, F., Beheary, M.M., AbdelMoneim, K.M., 2006. Effect of dust on the transparent cover of solar collectors. Energy Convers. Manage. 47 (18–19), 3192–3203. Fraunhofer, J.A., 2012. Adhesion and cohesion. Int. J. Dentistry 2012, 1–8. García-Heras, M., Carmona, N., Ruiz-Conde, A., Sanchez-Soto, P., Benítez, J.J., 2005. Application of atomic force microscopy to the study of glass decay. Mater. Charact. 55, 272–280. Ghazi, S., Ip, K., Sayigh, A., 2013. Preliminary study of environmental solid particles on solar flat surfaces in the UK. Energy Procedia 42, 765–774. Hegazy, A.A., 2001. Effect of dust accumulation on solar transmittance through glass covers of plate-type collectors. Renew. Energy 22, 525–540. Javed, W., Wubulikasimu, Y., Figgis, B., Guo, B., 2017. Characterization of dust accumulated on photovoltaic panels in Doha, Qatar. Sol. Energy 142, 123–135. Kazem, H.A., Chaichan, M.T., 2016. Experimental analysis of the effect of dust’s physical properties on photovoltaic modules in Northern Oman. Sol. Energy 139, 68–80.

Khalil, E.M.A., El Batal, F.H., Hamdy, Y.M., Zidan, H.M., Aziz, M.S., Abdelghany, A.M., 2010. Infrared absorption spectra of transition metals-doped soda lime silica glasses. Physica B 405, 1294–1300. Klimm, E., Ost, L., Kohl, M., Weiss, H.-A., 2016. Microscopic measurement and analysis of the soiling behavior of surfaces with standardized and real dust – a parameter study. Energy Procedia 91, 338–345. Klugmann-Radziemska, E., 2015. Degradation of electrical performance of a crystalline photovoltaic module due to dust deposition in northern Poland. Renew. Energy 78, 418–426. Lin, C.-A., Tsai, M.-L., Wei, W.-R., Lai, K.-Y., He, J.-H., 2016. Packaging glass with a hierarchically nanostructured surface: a universal method to achieve selfcleaning omnidirectional solar cells. ACS Nano 10, 549–555. Paudyal, B.R., Shakya, S.R., 2016. Dust accumulation affects on efficiency of solar PV modules for off grid purpose: a case study of Kathmandu. Sol. Energy 135, 103– 110. Qian, D., Marshall, J.S., Frolik, J., 2012. Control analysis for solar panel dust mitigation using an electric curtain. Renew. Energy 41, 134–144. Rifai, A., Abu-Dheir, N., Yilbas, B.S., Khaled, M., 2016. Mechanics of dust removal from rotating disk in relation to self-cleaning applications of PV protective cover. Sol. Energy 130, 193–206. Sarver, T., Al-Qaraghuli, A., Kazmerski, L.L., 2013. A comprehensive review of the impact of dust on the use of solar energy: history, investigations, results, literature, and mitigation approaches. Renew. Sust. Energy Rev. 22, 698–733. Sayyah, A., Horenstein, M.N., Mazumder, M.K., 2014. Energy yield loss caused by dust deposition on photovoltaic panels. Sol. Energy 107, 576–604. Semaoui, S., Arab, A.H., Boudjelthia, E.K., Bacha, S., Zeraia, H., 2015. Dust effect on optical transmittance of photovoltaic module glazing in a desert region. Energy Procedia 74, 1347–1357. Smallwood, J., Praeger, M., Chippendale, R., Lewin, P., 2016. Dielectrophoretic adhesion of 50–300 lm particles under ambient atmospheric conditions. J. Electrostat. 82, 1–6. Style, R.W., Peppin, S.S.L., Cocks, A.C.F., 2011. Mud peeling and horizontal crack formation in drying clays. J. Geophys. Res. 116, 1–8. Sueto, T., Ota, Y., Nishioka, K., 2013. Suppression of dust adhesion on a concentrator photovoltaic module using an anti-soiling photocatalytic coating. Sol. Energy 97, 414–417. Tanesab, J., Parlevliet, D., Whale, J., Urmee, T., Pryor, T., 2015. The contribution of dust to performance degradation of PV modules in a temperate climate zone. Sol. Energy 120, 147–157. van Oss, C.J., Good, R.J., Busscher, R.J., 1990. Estimation of the polar surface tension parameters of glycerol and formamide for use in contact angle measurements on polar solids. J. Disper. Sci. Technol. 11, 75–81. Xiong, J., Das, S.N., Kar, J.P., Choi, J.-H., Myoung, J.-M., 2010. A multifunctional nanoporous layer created on glass through a simple alkali corrosion process. J. Mater. Chem. 20, 10246–10252. Yilbas, B.S., Ali, H., Khaled, M., Al-Aqeeli, N., Abu-Dheir, N., Varanasi, K.K., 2015. Characteristics of laser textured silicon surface and effect of mud adhesion on hydrophobicity. Appl. Surf. Sci. 351, 880–888. Yilbas, B.S., Ali, H., Al-Aqeeli, N., Abu-Dheir, N., Khaled, M., 2016a. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol. Energy 125, 282–293. Yilbas, B.S., Ali, H., Al-Aqeeli, N., Khaled, M., Abu-Dheir, N., Varanasi, K., 2016b. Characterization of environmental dust in the dammam area and mud aftereffects on bisphenol-a polycarbonate sheets. Sci. Rep. 6, 24308. Yilbas, B.S., Ali, H., Al-Aqeeli, N., Oubaha, M.M., Khaled, M., Abu-Dheir, N., 2016c. Laser gas assisted nitriding and sol–gel coating of alumina surfaces: effect of environmental dust on surfaces. Surf. Coat. Tech. 289, 11–22. Zaihidee, F.M., Mekhilef, S., Seyedmahmoudian, M., Horan, B., 2016. Dust as an unalterable deteriorative factor affecting PV panel’s efficiency: why and how. Renew. Sust. Energy Rev. 65 (1), 1267–1278. Zhang, H., Li, X.-G., Du, C.-W., Qi, H.-B., 2009. Corrosion behavior and mechanism of the automotive hot-dip galvanized steel with alkaline mud adhesion. Int. J. Miner. Metall. Mater. 16, 414–421.