Atmospheric Nanoparticles in Photocatalytic and Thermal Production of Atmospheric Pollutants

R.R. Chianelli1 and M.J. Yacaman2

1Chemistry Department, Physical Sciences Building, University of Texas at El Paso, Texas, 79968-0513. Phone: (915)-747-7555, Fax: (915)-747-5748; 2Instituto de Fisica, Universidad Nacional Autonoma de Mexico, Apdo. Postal20-364, Delegacion AlvaroObregon, 01000Mexico, D. F., Mexicoand Instituto Nacional de Investigaciones Nucleares (ININ), Carr, Mexico-Toluca. Phone: (525)518-4150, Fax: (525)616-1535

Abstract

Atmospheric aerosols which occur above heavily polluted areas such as Mexico City are characterized and found to be complex materials which have the potential to accelerate important ozone-forming reactions photocatalytically and thermocatalytically. In addition, because the particles are respirable, they represent a considerable health hazard. The aerosols consist of two intermixed components. The first component consists of amorphous carbonaceous materials of variable composition with "fullerene like" materials dispersed throughout. The second component is an inorganic material consisting of nanoparticles of oxides and sulfides "supported" on clay minerals. This inorganic component has all of the characteristics of an airborne photocatalyst. Nanoparticles of Fe203, MnO2 and FeS2 have demonstrated catalytic properties, particularly when occurring in the nanoparticle range as they do in the subject aerosol materials. These materials have band-gaps which occur in the broad solar spectrum enhancing the photocatalytic adsorption of solar radiation beyond that of the wider band-gap aluminosilicate and titanate materials which also occur in the aerosols. In addition, the materials are acidic and probably are coated with moisture when suspended in air, further enhancing the catalytic ability to crack hydrocarbons and create free radicals.

Keywords: atmospheric nanoparticles, photocatalyst, atmospheric pollutants

Introduction

Contaminated cities suffer from high levels of airborne particles which, during adverse weather conditions, form visible clouds suspended above the cities as seen over Ciudad Juarez in Figure 1. In addition to the catalytic effects suggested in this report, particles of the type described are also believed to have adverse public health effects. The United States Environmental Protection Agency (USEPA) has recently reported that particulate matter smaller than 10µm (PM-10) increased the level of respiratory illness in children when exposed to levels of 40-58 µg/m3 for extended periods (EPA, 1993) and has recently proposed to further restrict particles smaller than 2.5µm (PM-2.5) (Chemical and Engineering News, 1997). Additionally, these airborne particulates and the associated polluting molecules are held responsible for a general degradation in the quality of life in these cities, including deterioration of monuments and buildings, visibility degradation, and acid rain. In spite of the negative impact of air particulates, little is known regarding their composition and properties, especially those which fall below 2.5µm. In this report, we characterize nanoparticles present in aerosols from Mexico City and emphasize the possible role of these atmospheric nanoparticles in photocatalytically producing alkyl free radicals and accelerating other reactions discussed below.

We present evidence that particles smaller than 2.5µm can play an important but as yet undetermined role in catalytically accelerating or guiding atmospheric reactions which occur in polluted air. The importance of excess hydrocarbons (HCs) and nitrous oxides (NOxs) from vehicular emissions in producing ozone has long been recognized (National Research Council, 1991). Recently, Blake and Rowland have pointed to the role of leaking alkane hydrocarbons from liquefied petroleum gas (LPG) in increasing the amount of ozone produced over Mexico City (Blake,1995). The authors present evidence for very high concentrations of C3 and C4 alkanes and reactive olefinic and acetylenic products from incomplete combustion of LPG and conclude that these molecules play a major role in the production of ozone in the Valley of Mexico. It is likely that a similar situation exists in many other cities where LPG use is high.

The role of catalytic surfaces present in atmospheric aerosols has not been well studied and little appears in the literature regarding the role of indirect photocatalytic pathways. The general role that photocatalytic particles play on global atmospheric chemistry has been reviewed by Zamaraev, et al. (Zamaraev, 1994) The presence of a photocatalyst will permit photo-reactions to occur within the main spectral range of available sunlight. Atmospheric nanoparticles arising from disturbed soil particles (and other sources) are expected to contain photocatalytic solids which affect key reactions. The major solid component of atmospheric aerosols are nanoparticulate silicates which adsorb short wavelength light. The concentration of such particles is highest in dry areas with values in the neighborhood of 100mg/m3. Surface areas of these atmospheric nanoparticles are reported to be within the range that characterize heterogeneous catalysts (10-100m2/gm). The particles are located primarily near the earth's surface and are usually covered with an adsorbed layer of water which is important in considering the potential photocatalytic pathways occurring. Additionally, the presence of sulfur in heavily polluted areas leads to the formation of sulfuric acid which creates an acidic environment in the encapsulating water layer.

Zamaraev further points out that adsorption of a quanta of light above the band-gap causes excitation of an electron from the valence band to the conduction band with this excited electron acting as a reductant and the hole in the valence band acting as an oxidant. Compounds with band-gaps appearing in the solar spectrum region such as Fe2O3 (2.2ev) and TiO2 (3.0ev) are known photocatalysts and appear in the atmosphere. Frei has reported that visible light and O2 are all that are required to partially oxidize hydrocarbons which are trapped in a zeolite (Frei,1996). Recently, Vinodgopal et al., reported the photosensitized destruction of dyes on semiconductor surfaces such as TiO2 in the presence of oxygen (Vinodgopal,1996).

Materials and Methods

Samples reported in this study were collected using standard stacked Nucleopore filter techniques and have been previously described (Aldape,1991a, Aldape, 1991b, Aldape,1993). Six samples were collected at four sites in Mexico City: Nezahualcoyotl (A), Portales (B), Taxqueña (C), Nezahualcoyotl II (D), Hangares (E), and Secretaria de Hacienda (F). The samples were analyzed with XFA (x-ray fluorescence analysis), transmission electron microscopy (TEM) and IR (infrared spectroscopy). The powders directly from the filters were suspended in alcohol using an ultrasonic bath. A drop of solution was placed on a 200-mesh copper grid (covered with a thin carbon layer) and air dried. Samples were examined using a JOEL 4000-EX high resolution microscope. Some of the images were computer processed to increase the signal to noise ratio. Electron diffraction patterns were measured using a CCD camera and analyzed using a standard commercial software package. Samples were also examined with an SEM JOEL 6400 and a JOEL 2010-EX microscope fitted with x-ray fluorescence analysis. Samples were also examined by infrared spectroscopy (IR) in KBr pellets using a Perkin Elmer Model 238B spectrometer.

Results and Discussion

Both Mexico City, Mexico (>5000ft.) and El Paso,Texas/Juárez, Mexico (>4000ft.) are at high altitudes, surrounded by mountains confining air movement, resulting in atmospheric inversion especially in cold seasons. Both cities sustain large populations, concentrated vehicular activity, and concentrated industrial activity. Particulate input to the environment due to the above activities is high, creating aerosols of nanoparticles. The composition of aerosols (respirable particles < 2.5µm) over Mexico City has been studied in a series of recent articles (8-10). Elemental analysis was performed using the Proton Induced X-ray Emission (PIXE) technique and the most abundant elements found were typical of soil components (Si, Al, Ca, and Fe). Anthropogenic elements were also found (soot, S, and Pb). The amount of anthropogenic versus soil components of the aerosol varied widely but was generally in the neighborhood of 50%. Anthropogenic material (soot) is approximately represented by the carbon analysis and was found to be in the range of 3-7 µg/m3. Inorganic material (soil) is approximately represented by silicon analysis and was found to be in the range of 1.5-5 µg/m3. Seasonal variations were observed with maximum values occurring in fall and winter. Daily values peaked between 6 and 12 A.M. Levels were compared with levels found in Los Angeles (1976) prior to the introduction of automotive pollution controls and were generally higher. Total particulate concentrations were in the range of 120 µg/m3 which compares with global measurements in the range of 65-200 µg/m3 for wind-driven dust, which is thought to affect global climate with 50% of this number is believed to be a result of human activity (Li, 1996 and Tegen,1996).

A typical SEM image of particles collected over Mexico City is shown in Figure 2 with the corresponding x-ray analysis shown in Figure 3. In the figures it is possible to see elongated platelet particles and some filaments. X-ray fluorescence analysis was used to identify the chemical composition of the particulate and fibrous materials. It can be seen from Figure 3 how the elements are associated with the various particles imaged by the SEM. Table 1 shows the carbon and silicon analyses for the six samples with four analyses of different portions of the sample. It can be seen from Table 1 that the carbon content lessens as the silicon content increases and we use these measurements as an indication of the amount of soot (carbon analysis) and the amount of soil (silicon) in the sample. Table 2 indicates the full analysis of the samples associated with the highest and lowest value of the carbon analysis.

All other samples follow similar trends and it can be seen from this data that the soil content increases as the soot content decreases. X-ray diffraction was used to identify the main phases which are present on the dust. In addition, selected area electron diffraction and nanodiffraction in the TEM were used to identify nanoparticles supported in soil or carbonaceous material. Table 3 indicates the crystalline phases with the samples which were identified by x-ray and electron diffraction. Particle samples from the atmospheric aerosols were also examined by infrared spectroscopy. The resulting spectra were extremely broad indicating very disordered, small particulate material. The major peaks are indicated in Table 4 and are consistent with the data for crystalline phases in Table 3. The composition of the particles is also consistent with reports of the composition of these materials in the atmosphere (Vinodgopal, 1996). The pH of an aqueous phase in which particles were suspended was approximately 4.5 indicating that acidic oxides were also present as indicated in the IR.

A TEM image of a typical nanocrystalline particle is shown in Figure 4a. The highly defined surfaces can be clearly seen. The size distribution of particles is shown in Figure 4b. Note that the particles have a very sharp distribution in the nanometer size range with an average particle size between 10 and 12 nm. Most of the nanoparticles observed in Figure 4a correspond to Fe203 with some of the particles being other oxides such as Mn02 and sulfides such as FeS2. These nanoparticles are dispersed on clay particles in the micrometer range size much as the active phase of a commercial catalyst is dispersed on a support material. Figure 5a shows a high resolution image of nanoparticles in a sample from Hangares. The insert shows some high resolution images of Fe203 as determined from the lattice spacings. The corresponding size distribution is shown in Figure 5a and again the sharp distribution in the lower nanosize interval is clearly seen in Figure 5b. Additionally, the material contains large amounts of amorphous carbon. In several areas we found carbon tubes and well-defined onion structures of the fullerene family similar to those reported recently in the literature (Haggin, 1994. ). An example of those structures is shown in Figure 6.

The above data give us a picture of the particulate matter in the Mexico City region as nanoparticulate agglomerates containing high surface area soils (silicates/oxides), semiconducting particles (hematite/pyrite), and other acidic and carbonate solids, probably encapsulated in water. Accompanying these particles is an approximately homogeneous distribution of carbonaceous materials.

Atmospheric aerosols which occur above heavily polluted areas are complex materials which have the potential to accelerate important ozone-forming reactions photocatalytically and possibly thermocatalytically (Zamaraev,1994). In addition, there is abundant evidence for deleterious health effects from inhaled particles in this size range (Chemical and Engineering News, 1997). The particlulate component of the aerosols consist of two intermixed components, inorganic and organic as described above. The organic component consists of amorphous carbonaceous materials of variable composition with "fullerene-like" materials dispersed throughout. The origin of this component is generally attributed to soot production from burning hydrocarbon based fuels. The carbon-based component of these aerosols will be the subject of a future article.

The inorganic component consists of nanoparticles of oxides and sulfides "supported" on clay minerals. Additionally, there is a considerable amount of sulfur present rendering the materials acidic in the presence of water. The origin of this component is generally thought to be from windblown dusts created naturally or by human activity such as vehicular traffic or quarrying operations. This inorganic component has all of the characteristics of an airborne photocatalyst. Nanoparticles of Fe203, MnO2, and FeS2 have demonstrated catalytic properties, particularly when occurring in the nanoparticulate range as they do in the subject of aerosol materials. These materials have band-gaps which occur in the broad solar spectrum enhancing the photocatalytic adsorption of solar radiation beyond that of the wider band-gap aluminosilicate and titanate materials which also occur in the aerosols. In addition to this, the materials are acidic and probably are coated with moisture when suspended, further enhancing the catalytic ability to crack hydrocarbons and create free radicals.

The formation of ozone is not direct but is the result of photochemical interactions of volatile hydrocarbons and oxides of nitrogen (NOX). Volatile hydrocarbons present in the atmosphere are of many types and follow many different reaction pathways making detailed analysis difficult (Carter, 1994). Blake and Rowland presented a simplified reaction scheme which starts with the production of alkyl free radicals from hydroxyl free radicals (HO):

HO + RH H2O + R (R-=alkyl radical) (1)

The alkyl radicals then proceed to produce alkyl peroxides which in turn react to produce aldehydes, ketones, and nitrogen dioxide (NO2).

R + O2 ROO (ROO =peroxide radical) (2)

ROO + NO RO + NO2 (RO = aldehyde or ketone) (3)

The NO2 photocatalytically produces excited atomic oxygen which produces ozone with free oxygen.

NO2 + h NO + O* (O* = atomic oxygen [0{1D}]) (4)

O* + O2 + M O3 + M (M = mediator molecule) (5)

The detailed reaction pathway and estimate of rate constants are given in Atkinson (Atkinson, 1990). The formation of the radicals in equation 1 is slow due to the high activation energy of hydrogen atom transfer with room temperature rate constants on the order of 10-14 to 10-15 cm3s-1. Under atmospheric conditions, the free radicals are rapidly converted to the peroxyradicals. The peroxyradicals are stable with respect to most atmospheric components and react predominately with NO to form NO2 (Francisco, 1995). Rate constants for alkyl radical formation increase as the carbon chain gets longer supporting the contention that presence of C3 and C4 hydrocarbons accelerate the overall rates. The overall reactivity is thought to be further accelerated by the olefinic and acetylenic content of the atmospheric HCs which appear as additional reactants. The alkyl radical formation should also be enhanced by the presence of catalytic surface.

Atomic oxygen (O{1D}) is produced photochemically from NO2 and this reaction has a DHo(298K) of 300 k J mol-1 and an threshold of 398 nm. Hydroxyl radicals are also formed photochemically with a DHo(298K) of 491 k J mol-1 and a threshold of 243 nm:

H2O + h OH + H

TiO2 is a well-recognized photocatalyst which completely oxidizes hydrocarbons, photocatalytically producing only CO2 and H2O (Hoffmann, 1995). However, the phase and structure are crucial for this to occur effectively. If the TiO2 is in the rutile phase or other titanium-containing oxide phases are used, a considerable amount of "coke" is produced reducing activity. Thus, ineffective catalysts produce coke and this means the production of carbon containing free radicals and unsaturated products. The photocatalysts described in this report are relatively poor photocatalysts and are

coated with carbon indicating the production of free radicals and partially oxidized

products (Cheng, 1995).

Hydroxyl radicals (reaction 6) and peroxide radicals can also be provided by photocatalytic dissociation of water:

H + O2 HOO (7)

Semiconducting oxides and sulfides are known to possess water-splitting activity in aqueous solution (Fox, 1993). Gas phase photoreactions of these materials are virtually unstudied. However, since the nanoparticles are probably covered with a thin layer of water, the reactions maybe similar. Nevertheless, these reactions need to be studied experimentally to confirm their importance in the overall scheme.

Conclusion

The catalytic role that these materials may play in tropospheric chemistry remains to be elucidated and is the subject of a continuing study. But there is no doubt that the chemistry and physics of these nanoparticulate materials, which occur in increasing amounts over polluted cities, must be understood in greater details to determine the role they play both in accelerating atmospheric reactions and in creating human health hazards. It should also be added that it is also necessary to understand the role that human versus natural activities play in creating these materials. In this respect it is worth mentioning the very uniform distribution of nanoparticulate materials. In future studies we will investigate by what mechanism such uniformity is created.

Acknowledgements

The authors would like to acknowledge useful conversations with Professors Jorge Gardea-Torresdey, William Herndon, N.J. Parks, and Nicholas Pingitore of the University of Texas at El Paso. We would also like to acknowledge support from the CERM (Center for Environmental Resource Management) and PACES (Pan American Center for Environmental Studies) at the University of Texas at El Paso. The authors are also indebted to Dr. F. Aldalpe of ININ for providing the samples used on this work, and to J. Arenas (ININ) for his TEM and SEM work.

References

Aldape, F., Flores M. J., R.V. Diaz, and D. Crumpton,1993. Nuclear Instruments and Methods in Physics Research, vol. B75, p. 304.

Aldape, F., M. J. Flores, R. V. Diaz, J. Miranda, J.R. Morales, T.A. Cahill, and L. Saravia, 1991a. International Journal Of PIXE, vol. 1(4), p. 355.

Aldape, F., M. J. Flores, R. V. Diaz, J. Miranda, T.A. Cahill, and J.R. Morales,1991b. International Journal Of PIXE, vol. 1(4), p. 373.

Atkinson, R., 1990. Gas Phase Tropospheric Chemistry of Organic Compunds: A Review, Atmos. Environ., vol. 24A, p.1.

Blake, D.R., and F. S., Rowland, Blake. Science, vol. 269, p. 953.

Carter, W.P.L.,1994. Air Waste, vol. 44, p. 881.

Chemical and Engineering News, February 3, 1997. page 28.

Cheng, S., S. J. Tsai, and Y. F. Lee,1995. Catalysis Today, vol. 26, p. 87.

EPA, April 1993. "Mobile Vehicle Relate Air Toxics Studies," Office of Mobile Sources.

Fox, M.A., and M. T. Dulay,1993. Chem Rev. vol. 93, p. 341.

Francisco, J.S., and M. M. Maricq, 1995. Advances in Photochemistry, vol. 20, p. 79.

Frei, H., April 1,1996. Chemical and Engineering News, p. 5.

Haggin, J., 1994. Chemical & Engineering News, vol. 72(7), p. 22.

Hoffmann, M.R., S.T. Martin, W. Choi, and D.W. Bahnemann,1995. Chem. Rev., vol. 95, p. 69.

Li, X., H. Maring, D. Savole, K. Voss, and J.M. Prospero, 1996. Nature, vol. 380, p. 416.

National Research Council, 1991. Rethinking the Ozone Problem in Urban and Regional air Pollution, National Academy Press, Washington, DC.

Nyquist, R.A., and R.O. Kagel, 1971. Infrared Spectra of Inorganic Compounds, Academic Press, New York.

Tegen, l., A.A. Lacis, and I. Fung,1996. Nature, vol. 380, p. 419.

Vinodgopal, K., D.E. Wynkoop, and P.V. Kamat,1996. Environ. Sci. Technol., vol. 30, p. 1 660

Zamaraev, K.l., M.l. Khramov, and V.N. Parmon, 1994. Catal. Rev.-Sci.Eng., vol. 36(4), p. 617.



Table 1. Atomic weight % carbon and silicon
sample at. wt. % 1 2 3 4
A carbon 82.71 85.84 85.03 42.79
silicon 3.37 2.22 2.82 15.53
B carbon 57.94 63.38 56.20 72.17
silicon 11.17 11.52 11.29 4.60
C carbon 56.35 55.11 61.06 66.39
silicon 11.28 19.61 7.68 7.39
D carbon 64.88 65.28 53.32 63.96
silicon 9.63 11.88 10.64 7.20
E carbon 59.25 54.31 - -
silicon 9.77 10.50 - -
F carbon 67.99 77.77 79.12 78.45
silicon 7.08 4.77 2.46 4.30










Table 2. Atomic weight % of detectable elements
C O Na Mg Al Si P S Cl
82.71 9.40 0.70 0.12 0.84 3.37 0.07 0.38 0.04
42.79 28.16 2.15 0.45 4.99 15.53 0.21 0.24 0.05


K Ca Ti Cr Mn Fe Cu Zn
0.15 0.61 0.07 0.00 0.00 1.11 0.27 0.17
0.78 2.43 0.28 0.00 0.00 1.21 0.42 0.32






Table 3. Chrystaline Phases in Aerosol Samples
Phase Mineral
NaAlSi3O8 Albite
CaCO3 Calcite
NaCl Halite
Na2SO4-1O H2O Mirabilite
Na2CO3-1O H2O Natron
Na2CO3-NaHCO3-2H2O --
NaHCO3 --
CaSO4-2 H2O Gypsum
Fe2O3 Hematite
FeS2 Pyrite
MnO2 --






Table 4. Infrared Spectrum of Particles from Aerosols
Peak (cm-1) Strength Attribution

(Nyquist, 1971)

3600-2400 broad, strong water, hydration
2000-1400 very broad, medium carbonate, water
1250-650 broad, strong sulfate, silicate, carbonate
600-250 broad, very strong hematite, pyrite






Figure 1. Dust cloud above Ciudad Juarez, Mexico/El Paso, Texas. Photograph was taken during March 1996 from west El Paso looking south into Ciudad Juarez (courtesy of Dr. L. Murr, Chairman and Professor of Materials Science, University of Texas at El Paso)



Figure 2. SEM image of typical dust particle.



Figure 3. SEM image of the particulate material collected from the atmosphere with associated x-ray fluorescence analysis: SEM, Carbon, Calcium, Aluminum, Copper, Iron, Potassium, Magnesium, Sodium Oxygen, Silicon, and overlay of all measured elements.

Figure 4a. High resolution image (at low magnification) of a dust particle.



Figure 4b. Particle size distribution of Fe203 nanoparticles in dust collected from the Mexico City Atmosphere



Figure 5a. High resolution image of a dust containing a clay with act as a support for nano-particles. The inset shows some high magnification of the Fe203 particles.



Figure 5b. Corresponding size distributions.

Figure 6. Fullerene cages formed in some soots observed in the sample of Nezahualcoyotl.