Chapter 4 Composition of the coatings.

Results

Most of the glacial sand on Long Island is composed of quartz that is covered with iron-stained coatings. The color of these coatings varies from bright orange-yellow to pale beige. In most cases, the proportion of coatings in the sediment determines the sorption capacity of the sediments, even though the coatings constitute only 0.8 to

13 % of the sediment (Table 1), (Coston et. al, 1995). As a result, it was necessary to evaluate the mineralogical and the chemical composition of the coatings. Coarse and medium-coarse sand from Port Jefferson (samples 1 and 5) and South Setauket (samples 6 and 7) and fine sand from beach cliff at David Weld Preserve were chosen because these sediments contain a significant proportion of coatings and represent different depths and environments (see Methods).

The morphology of the grains surfaces and coatings of the whole untreated sand was examined by SEM. Transmission electron microscopy (TEM) was used to examine coating separates. Chemical analysis and selected-area electron diffraction (SAED) patterns of individual fine particles were collected to aid in mineral identification. The clay assemblage of the coatings was identified by X-ray diffraction (XRD). Also total organic carbon was measured (TOC) on both the whole sand and the coatings.

X-ray diffraction (XRD)

X-ray diffraction was used to distinguish the clay minerals in the < 2mm fraction of the fine coatings from fine sand at the beach cliff at David Weld Preserve (Figure 1 and Table 1).

Table 1. Basal Reflection d-spacings for common clay minerals.

Clay

14

10

7

5

3.5-3.6

Chlorite

X

 

X

 

X

Kaolinite

   

X

 

X

Illite

 

X

 

X

X

Vermiculite

X

 

X

 

X

Illite is unambiguously recognized by the 10 and 5 reflection peaks. Chlorite has 14 , 7 , and 3.5 basal reflections, which coincide with basal reflections for kaolinite and vermiculite. An additional test was necessary to distinguish between the second chlorite reflection at 7 and the first kaolinite reflection at 7 using the procedure of Moore (1989). An aliquot of the sample was boiled for 2 hours in 6M HCl. Under these conditions most chlorite is dissolved and any residual 7 peak presumed to indicate kaolinite. In this sample the residual 7 peak was observed which indicates the presence of kaolinite (Figure 2 A). To distinguish between the 14 chlorite reflection and the 14 vermiculite reflection, the sample was heated at 3000C for an hour (Moore,1989). This treatment collapses vermiculite to 10 , but chlorite retains its 14 peak. In this sample 14 peak is observed after heat treatment, which indicates the presence of chlorite (Figure 2 B). After heat treatment the 14 peak is smaller than the 10 peak suggesting that vermiculite or a mixed chlorite-vermiculite may be present. Illite, kaolinite and chlorite are unambiguously present in this sample.

Scanning Electron Microscopy (SEM)

Scanning Electron Microscopy (SEM) was used to examine the surface of untreated medium-coarse sand from Port Jefferson (samples 1 and 5), and fine coated sand from the beach cliff at the David Weld Preserve (sample 10). SEM examination was done to evaluate whether the morphology of the particles in the coatings could clarify whether the particles are authigenic or detrital.

The surface of the quartz grains are extensively covered with fine particles of the different sizes and shapes (Figures 3, 4, 5, 6 and 9). The images of the medium-coarse sand from sample 1 are shown in the Figures 3, 4, 5, 6, 7, and 8. The SEM images of the fine coated sand from sample 5 are shown in Figures 9 and 10. The SEM images of sample 10 are shown on the Figures 11 and 12.

Most particles have irregular shapes. Almost all quartz grain surfaces are covered with coatings (Figures 5 and 9). The uncoated areas of quartz show signs of extensive weathering (etch pits) (Tazaki, Fyfe, 1986) (Figure 8). Within the coatings the smaller particles are observed to be on the surfaces of larger particles (Figures 5 and 6). Grains with laminar structure are observed in Figure 11. In sample 5 (coarse-medium sand from Port Jefferson) platy grains, possibly clay particles, were noticed (Figures 9 and 10). Pseudo-hexagonal crystals were identified in sample 5 (Figure 10). This may be the evidence of authigenic clay (Tazaki, Fyfe, Van der Gaast, 1989; Small 1992). Except for this sample most of the grains have irregular shapes.

The fine sand from the beach cliff at the David Weld Preserve contains a significant fraction of mica shaped grains and grains with platelet morphology

(Figure 11).

Transmission Electron Microscopy (TEM)

Transmission electron microscopy was undertaken to determine the mineralogy and composition of the coating particles. Ultrasonically separated coatings from sample 1 (coarse-medium sand from Port Jefferson) were chosen because this sample is heavily covered with a bright brown-orange colored coating. The TEM images are shown on the Figures 13A, 13B, 13C, and 13D.

The coatings contain particles of different sizes. Figure 13A shows a large grain 3 to 4 m m in size. The dark grain in the upper right corner is almost 1 m m. The smaller particles on the surface of the large grain vary in size from 1.2 m m (the largest one) to the less than 1/10 m m. A lot of very small particles are around the large grain, most of them form aggregates less than 1 m m. The largest grain in Figure 13C is 2.5 m m long and 1 m m wide. A lot of very small particles (sizes much less than 1 m m) are in the right low corner. Some small particles form aggregates, which are hard to distinguish from the individual grains. Most of the particles are platelets, consistent with layered silicates, but show irregular shape. Small particles are observed on the surface of the larger grains and around them (Figures 13A and 13C).

Selected-area electron diffraction (SAED) patterns of the individual grains were obtained to identify the mineralogy of the particles. The dispersed grain technique used for sample preparation results in a unique orientation of platy grains. Because of the strong preferred orientation of the platelet, all SAED patterns revealed dominantly a single zone-axis orientation of reciprocal space, i.e. the a-b plane. The large isolated grains give single-crystal spot patterns, whereas the aggregates of very small grains give ring patterns (Figures 13B and 13D). Essentially all particles showed diffraction contrast and indicating that all grains are crystalline.

The diffraction patterns were indexed using an independently calibrated camera constant l L and the relationship

R*d = l L                                                                  (eq. 1)

Where R is the magnitude of the diffraction vector measured at the film, and d is the interplater spacing. All particles have a 4.2-4.4 spacing for the a-b plane. The ring patterns of the aggregated small grains are superimposed on the spots. This indicates the similarity of crystal structure of the small particles and the large isolated grains. This spacing is characteristic of the clay minerals, kaolinite, smectite and illite. To distinguish among these clay minerals we need to rely on chemical analysis of the individual grains and the aggregates.

Chemical analyses of the individual grains and the particles aggregates were obtained by TEM. 1000 beam size was used for chemical analyses, this is 10 times smaller than the beam typically used in an electron microprobe. We were only able to analyze chemical composition of the large grains; for the small particles this beam size was too large as a result it was necessary to analyze aggregates of the finer particles.

The integrated intensities of the characteristic peaks for the different elements were measured. The peak intensities were converted to concentration using procedure introduced by Cliff and Lorimer (1975). The chemical compositions of the individual grains in the sample 1 are plotted on a ternary diagram (Figure 14) along with the compositional range of illite, chlorite, kaolinite, montmorillonite and celadonite. Except for kaolinite these clay minerals show broad compositional variability. The compositional range data of these minerals were taken from Newman (1987). Ternary diagram was constructed using molar ratios of K, Al and Fe+Mg to Si.


Figure 14. Ternary diagram shows compositional range for the fields of illite, chlorite, montmorillonite, kaolinite and celadonite and  chemical composition  of the coatings from sample 1 (red dots). The ternary plot is constructed based on the molar K, Al and Mg+Fe to Si molar ratios. 


The coatings are composed predominantly of Si, Al, with some Fe, K and Mg and minor amounts of Mn and Ca. Most of analyses plot close to illite, some are close to chlorite (Figure 14). Some spread in data was observed it might be due to the analytical uncertainty (5%) or variations in the chemical composition of the particles. The low number of counts is the main source of uncertainty. The measurement of the k-factor contributes to the uncertainty, but it was determined using the same procedure as the sample.

Results of the TEM examinations show that coatings have a large range of grain sizes, and in most cases irregular shape. Small and large particles have identical SAED patterns (4.2-4.4 ).The TEM data as well as the XRD data suggest that the clay minerals are. kaolinite, illite and chlorite.

Total Organic Carbon (TOC)

It is important to determine the content of organic matter in sediment because organic matter has high cation exchange capacity (200-400 meq/100g), (Appolo,1992) and is the major sorbent for hydrophobic organic compounds (HOC). The weight fraction of TOC of whole sand varies between 0.0002-0.0003 (Table 2). Coatings have an order of magnitude higher fractions of TOC varying from 0.002 (coarse, heavily coated sand from Port Jefferson, sample 5) to 0.008 (coarse-medium sand from Port Jefferson, sample 1). The natural organic matter is typically made up of about half carbon so the fraction of organic carbon can be converted to fraction of organic matter by multiplying by 2 (Schwarzenbach et. al, 1993).

fom=foc*2                                                                      (eq. 2)

Where fom is fraction of organic matter and foc is fraction of organic carbon.

According to Schwazenbach et al (1993) Long Island sand contains relatively small fraction of organic carbon (foc= 0.0002-0.003).


Table 2. Fraction of the coatings to the whole sand mass and fraction of total organic carbon (foc) measured on the whole sand and coatings. Sample 1 is coarse heavily coated sand from Port Jefferson (depth 100-120 cm below the surface). Sample 5 is medium-coarse sand from Port Jefferson (depth 160-180 cm below the surface). Sample 6 is medium-coarse sand from South Setauket (140-150 cm below the surface). Sample 7 is coarse sand (200-220 cm below the surface) from South Setauket. 

Samples

Fraction of the coatings

Fraction of total organic carbon (foc)

Whole sand

Fraction of total organic carbon (foc)

Coatings

1

0.027

0.0003

0.008

5

0.065

0.0003

0.002

6

0.008

0.0002

0.005

7

0.009

0.0002

0.003

Discussion

TEM, SEM and XRD analyses show that the coatings consist dominantly of illite, with minor kaolinite and chlorite. The particles in the coating material have a large range in grain sizes and shapes. The SAED analyses show that the large and fine particles all have identical diffraction patterns consistent with illite or kaolinite. Only very minor fractions of organic matter and iron compounds were identified in the coatings.

The Long Island stratigraphic sequence consists of highly weathered basement gneisses, schists and granites overlain by Cretaceous sands and clays of the Raritan and Magothy formations which are in turn overlain by glacial sediments (Sirkin, 1995). Significant fractions of the underlying Cretaceous sediments were picked up by the Wisconsinan glaciers and were incorporated in the upper glacial sediments (Sirkin, 1995).

Liebling and Scherp (1975) found kaolinite and illite present in all cores in the Cretaceous Raritan Formation on Long Island. They found kaolinite to be generally more abundant with chlorite making up one tenth to one third of the clay. Wakeland (1979) reported that Long Island clays are dominated by illite with lesser amounts of smectite, chlorite and kaolinite. Liebling (1973) found that kaolinite and illite were the dominant clays in the weathered bedrock samples from the northern part of Nassau County with minor amounts of intrastratified vermiculite-montmorilonite.

There are several studies of the mineralogy of the particles in coatings of glacial sediments in environments similar to those on Long Island. The following two studies were conducted on Cape Cod, Massachusetts. Wood et al, (1990) found a clay assemblage of illite (25%), kaolinite (25%) and smectite (3%) in coatings on glacial sediments on the southern border of Otis Air Base, Cape Cod, Massachusetts.

Coston et al, (1995) found that the glacial sand in Falmouth, Cape Cod, Massachusetts consists primarily of quartz (95%) with minor amounts of feldspar and ferromagnetic minerals. They also report that the particles in the coatings consist of a variable mixture of Fe oxides, smectite and polycrystalline material enriched in Al, Fe and Si. They proposed that most of the particles in the coatings were derived by weathering of feldspar and accessory minerals in the sediments.

Ryan and Gschwend (1992) studied an unconfined sandy aquifer composed mostly of coated sand in the Lebanon State Forest (central New Jersey). They suggest that coatings were formed as the result of the illuviation or "mechanical infiltration" of clay colloids particles of authigenic clay particles. However, an aluminous-rich mineral as a source for authigenic clay, for example feldspar is not abundant in the Cohansey Sand.

In this study we found the clay mineral assemblage of the coating particles to be similar to the clay assemblage in the Cretaceous sediments. Therefore, the Cretaceous sediments could be the main source of clay minerals in the coating on Long Island. Exactly how the particles would be attached to the sand and silt grains is not clear.

We called "coatings" all fine grained material available after 30 minutes of sonication of coated sand or silt. After sonication the quartz grains lost most of their reddish-brown color. However, some fraction of strongly bonded material may be still left on the surface. Also, the material called coatings contains some fraction of fine particles not bonded to the mineral surface but which just filled the space between grains. It is not clear how one could sample the fine-grained interstitial material without including the particles in the coatings.

Did the particles in the coatings on the glacial sediments form before or after glacial deposition? The particles have irregular shapes and vary in size, which may be evidence for the physical crushing of the particles during glacial transportation or deposition. The uncoated surfaces of quartz have etch pits. These etch pits could have formed during weathering of the source rock or sediments or after these sediments were redeposited by the glacier. Only one pseudo-hexagonal crystal was identified by SEM. This may be evidence for authigenic clay growth. If so, some clay may have formed authigenically. The data, which we have, are not sufficient to determine whether the clays have an authigenic or detrital origin.

Return to Table of Contents