Plants can deposit inorganic materials in the form of crystals in their tissues as calcium salts (Franceschi and Horner, 1980) and silica hydrated oxides (Postek, 1981), often aggregated as druses, sphaerites, styloids, raphides inside the cytoplasm or vacuoles, but also many times spread in the cell walls and other tissues.
These minerals have very distinctive and strong bands in the TIR, therefore if they are present in the leaf surface (epidermis cell wall + cuticle), an ATR transmittance measurement will probably show it. For example, Magnolia grandiflora known for having a silica rich cuticle (Postek, 1981) shows characteristic strong Silica oxides (Si-O) molecular vibrations bands at 1045 and 780 cm-1 in its ATR transmittance spectrum. When M. grandiflora spectrum (between 1287 and 747 cm-1) was used in a search in a mineral spectral library (Minerals and rocks library) the first mineral in the result was Mordenite, second Silica gel, third Ferrierite (Figure 1), fourth Clinoptilolite, fifth Phillipsite, sixth Erionite. With the exception of Silica gel all the other minerals are Zeolites (http://en.wikipedia.org/wiki/Zeolite).
Figure 1. ATR spectra of a fresh leaf of Magnolia grandiflora, compared to silica gel and the Zeolites: Mordenite and Ferrierite, the three first matches in a Spectral library search. The two Si-O strong molecular vibrations between 780 and 1048 cm-1 are present in all spectra. Note weak wide OH band between 3700 and 2982 cm-1 and H20 band around 1630 cm-1 related to the minerals and leaf hydration.
To verify if these spectral features in leaves were from silica and also if any of the Zeolites could be part of the leaf surface, wavelength-dispersive X-ray elemental maps of Al, Ca, Mg, Na (common elements in several Zeolites), and Si were obtained using a JEOL JXA-8900 electron microprobe from leaves from 7 species: 3 species that are well known for having silica rich tissues as Curatella americana (Silica plant), Equisetum sp and Magnolia grandiflora (Magnolia) (Figure 2a); 3 species without known studies about their Silica content, but showing silica features in their ATR spectra, as Acer rubrum (Red Maple), Fagus grandifolia (Beech) and Quercus alba (White oak) (Figure 2b) and two species that seemed to lack silica features as Cornus florida (Dogwood) and Liriodendron tulipifera (Tulip Poplar) (Figure 2c).
Figure 2. a) Leaf spectra of species that are known to be silica rich, b) Leaf spectra of species that show silica features, but without studies of silica content. c) Leaf spectra not showing silica features.
The X-ray elemental maps results are semi quantitative because leaves are not as flat as it is required for a quantitative analysis.
Figure 3. Curatella americana adaxial surface shows relatively small cells and three sets of trichomes (CP). They are extremely rich in silica, but with traces only of the other elements Al, Mg and Ca most of what is just background noise (Na analysis is lacking).
Figure 4. Equisetum sp is quite rich in Ca and Si, with some traces of Mg and no Al or Na.
Figure 5. Magnolia grandiflora cuticle is formed by relatively large plates (CP) rich in silica, covering smaller cells from the epidermis underneath, which are well delineated in the Ca image, showing relatively high levels of Ca in the cells, and low levels between them. Traces of Mg, and no sign of Na or Al
Figure 6. Acer rubrum distinctive epidermal cells show a relative rich content of Si, Ca and Mg inside cells, but not in the cell walls between them; but no signs of Na or Al.
Figure 7. The irregular interstices of epidermal cells of Fagus grandifolia are quite rich in Si, with traces of Ca in the cell walls, but none of the other elements.
Figure 8. Quercus alba show some interstitial Si, and traces of intracellular Ca and Mg. No Na or Al.
Figure 9. Cornus florida CP image shows several T shaped trichomes very rich in Ca. The irregular shape epidermal cells show Ca rich interstices also, but the surface lacks the other elements in the analysis, with just background noise. Na analysis was not performed.
Figure 10. Liriodendron tulipifera shows some irregular small content of Ca, that seems to be extracellular, possibly localized in the cuticle.
From 8 species analysed with X-ray elemental maps 6 have shown the presence of Silica at the leaf surface. These same 6 species also show Silica oxide bands in their spectra. Curatella americana that has an extremely high content of silica has also strong and characteristic silica oxides (Si-O) bands in its spectrum. Equisetum sp and Magnolia grandiflora, show a relative high amount of Silica, that seems to be spread out in the cuticle. Acer rubrum shows an intermediate amount of intracellular silica, maybe in its cell walls, with no evidence of silica in the intercellular substance. Fagus grandifolia on the contrary shows relative high content of silica in its intercellular substance and practically no intracellular silica. Quercus alba shows a relative low amount of intra and intercellular silica probably in its cell walls. All these species show at least some degree of Si-O content in their spectra, that corresponds to strong to weak Si-O spectral features (Fig 2a and b). Cornus florida and Liriodendron tulipifera do not show Si-O spectral features what agrees with the lack of Si element in their mapping (Fig 2c).
Calcium was mapped in most species being analyzed except in Curatella americana, Fagus grandifolia and Quercus alba. The highest amount of Ca is seen in Cornus florida trichomes, although its spectrum does not show any spectral features of calcite or aragonite two common Calcium carbonates in organisms.
It is an appealing hypothesis the presence of organophilic/hydrophobic catalyst silica chambers, or micropores in leaf and cell surfaces, as zeolites, to assemble or extrude long chain aliphatic compounds as hydrocarbons, alcohols, esters and fatty acids found for instance in cuticle waxes. Natural Zeolites are crystalline hydrated alumino silicates with a tridimensional open (or hollow) structure, with organophlic and catalytic inner surfaces. Because of these qualities they have been used as selective sieves or filters in many industrial and agricultural applications (Mumpton, 1999), and it has been hypothesized that silicalite, a dealuminated zeolite, could have been the precursor catalysts of the first replicating biopolymers and perhaps assembling primitive cellular organisms during the Archean (Smith, 1998), due to their characteristics. Silicalites, or aluminum-free/mesopore zeolites are not very common in nature, but they have been synthesized and used by oil companies to catalyze aliphatic and aromatic hydrocarbons from smaller molecules (Derouane et al., 1978, Olsbye et al., 2012, Sartipi et al. 2013). More recently, synthesized silica nanopores are being used to transport DNA and chemicals into isolated plant cells and intact leaves ( Torney et al., 2007).
Although some silica bands in leaves are similar to zeolites as Mordenite and Ferrierite, elemental map of several leaves containing silica did not find Aluminum almost always found in natural zeolites. Na and Mg other elements common in Mordenite or Ferrierite were not found in most leaves. Maybe silica in leaves have some crystal structure similar to silicalites ( a synthetic zeolite free of Aluminum), or maybe there is some correlation with organic compounds present in leaves tricking the correlation of leaf spectra with Zeolites spectra.
To check for confusion with organic matter, leaves were submitted to 10-14 days in a plasma oven to burn organic matter in a low enough temperature not to change crystallinity of silica or other mineral compounds.
To check for crystallinity leaf ashes prepared in the low temperature oven were submitted to X Ray analysis.