ABSTRACT

Use of the densitographic method shows that the iron in sediments forms complexes, chiefly of a physical nature, with organic matter. The complex which is formed contains, in 80 percent of the cases, about three parts by weight of organic matter and one part of iron oxide. There are notable exceptions. Beach sand does not contain any appreciable amount of organic matter, and in certain "hardpans" the iron content is much higher than that of organic matter. In other sediments the organic content is greatly in excess of the iron. It is suggested that other metals, for example uranium in black shales, may form similar complexes.

MATERIALS AND METHODS

The materials used are listed in a previous paper in this journal (Baas Becking and Moore, 1959). They comprise nearly one hundred samples of estuarine, marine and freshwater sediments and a few samples of sedimentary rocks. These samples were submitted to densitometric analysis. All samples were examined microscopically. If magnetite, pyrite or marcasite were detected, the iron was extracted by treatment with 60% perchloric acid. In other cases 2N hydrochloric acid sufficed. The iron was then determined colorimetrically in a Lange photometer or a Spekker adsorptimeter, using the thiocyanate method.

Organic matter was determined either in the wet way (Piper, 1947, p. 214), by digestion in 20% hydrogen peroxide or, in the majority of cases, by ignition, subtracting 6% of the weight of clay minerals present. The figures for organic matter are less reliable than those for iron.

IRON CONTENT OF THE SEDIMENTS

The iron occurs in the samples as hydroxide; there may be phosphate, magnetite, siderite, and sulphide in addition. In this paper it is expressed as FeO(OH). In shales (Clarke, 1916) the ratio ferric/ferrous iron is about 1.6, in sandstone it is around 3.3, and no ferrous iron is reported from limestones. In foetid limestones nearly all the iron should be in the ferrous state, so that the above results may be ascribed to insufficient sampling. Figure 1 shows the iron content of the materials used by the authors. The distribution is given in table 1; it appears to be logarithmic. The mean is almost identical with that obtained by Rochford (1952) for the average composition of estuarine muds. Clark's results are shown in table 2. The sediments shown in table 1 comprise 60% shales, 24% s ndstones, and 16% limestones (estimation from a previous paper by the authors (Baas Becking and Moore, 1959)). Using Clarke's figures for iron hydroxide content we obtain an average of 5.03% FeO(OH), which is in close agreement with our own figure of 4.71%. The influence of sulphate-reduction on a water depends primarily upon its sulphate content. For seawater SO₄⁼=6.210^-2N, and excess of hydrogen sulphide will be produced at complete sulphate-reduction if Fe⁺⁺ is less than 6.210^-2N, that is, less than 0.17% Fe⁺⁺ or 0.27% FeO(OH). As nearly all this iron is available only in the mud surface only a fraction of it will be ble to bind hydrogen sulphide. Moreover, "calamitous" sulphate-reduction will take place when a sulphaterich water becomes isolated. The depth of the isolated water masses is variable. In cases observed by the authors in estuaries it may vary from 1 to 10 metres. Iron oxide content of the mud lower than 2% could lead to the formation of "foetid black mud" with a complete exhaustion of the watermass

End_Page 454------------------------

FIG. 1. Relation of organic matter to iron hydroxide in muds.

End_Page 455------------------------

in oxygen. About 30% of the determined iron values are of this order, two-thirds of them from marine and estuarine muds. On the other hand, very high iron values were obtained from a hardpan in coral rubble from Isle Charron, New Caledonia, and in a gravel layer from the bottom of a 10 foot core from the freshwater Lake George, New South Wales.

ORGANIC CONTENT OF THE SEDIMENTS

The nature of the organic matter in the sediments may be ascertained from microchemical reactions. In recent sediments this material may be pectic, chitinous, cellulosic or lignic. In fossil sediments dehydration may give rise to "caramels" and phytomelanins rich in carbon; dehydration plus de-oxygeneration will lead to the lignite-coal-anthracite series, whereas decarboxylation might cause the accumulation of hydrocarbons.

Pectic materials are found in estuarine muds in all stages of decomposition. The authors believe that the pectinoids are the chief fuel for microbiological reactions in estuarine mud. Chitinous materials are abundant and are also decomposed by bacteria, aerobically as well as anaerobically, although the turnover is much slower than in the pectinoids. Cellulose, present chiefly in plant remains, and to a lesser extent in Tunicates, is very slowly decomposed anaerobically in seawater. The slowest turnover is shown by the lignic materials. Decomposition, even in an alkaline marine environment, takes a very long time, probably at least a few years. In artificial muds described in a previous paper (1959) the rates of turnover of the materials mentioned above wer clearly apparent. The organic materials are even more diversified than are the iron compounds, and generalizations are, therefore, dangerous.

In the samples studied by us the organic matter varied between zero and 26 percent. Much higher values are recorded in the literature, especially from freshwater sediments, and much lower values are recorded for the marine environment outside estuaries. According to Rochford (1952) the organic content of estuarine mud ranges from 8.5-19% with an average of 11.76% which is very close to the average of our results. The distribution described in table 2 still shows considerable irregularity, in spite of the large frequency classes. Analytical errors play a subordinate role here, as shown by trial determinations using different methods. It must be that heterogeneity of the organic matter is the chief cause of this irregularity.

RELATION BETWEEN IRON AND ORGANIC MATTER IN MUDS

Since the densitograms described in a previous paper (1959) showed usually the absence of free iron compounds, it was assumed that iron enters into combination with other components of sediments. There is ample evidence in the literature to support this view. As far back as 1897 Spring has shown that "the organic matter of natural waters is incompatible with iron, the two substances separating out as a flocculent precipitate." According to Spring one part of colloidal ferric oxide (presumably FeO(OH)) will remove ten parts of humus from solution. The use of iron salts or alum in filtration plants is an application of this principle. Although Clarke (1916) claims that minerals in finely divided form are carried down mechanically in this coagulation,

TABLE 1. Iron content of sediments

TABLE 2. Iron in sediments, after Clarke, 1916

End_Page 456------------------------

there is much evidence to challenge this opinion, as the iron might well be adsorbed by both quartz and clay minerals.

Early in our work it was observed that the average density of a sediment decreases with increasing iron content. The highest density determined for normal sediments was 2.82. From this the following relations could be expressed by:

       density=2.82-0.065 (FeO(OH))%
    or density=2.82-0.022 organic%
     organic %= 2.88 FeO(OH)%

These relations are no more than tendencies. These tendencies no longer exist if there is a great preponderance of either component. However, they are explicable if we assume that the iron enters into combination (either physical or chemical) with the organic matter. In the case of the pectinoids such a combination may be partly chelate (Baas Becking and Mackay, 1955).

Figure 1 shows the relation between FeO(OH) content and organic content of 77 sediment samples; twenty of these results are taken from Rochford (1952). Table 3 shows the distribution of the relation between iron and organic matter. Exceptions exist in the region of a very low value of the quotient in iron-rich tills and in beach sand without organic matter. High values for the quotient, approaching those cited by Spring (1897), are found in Lake MacQuarie (estuarine) muds, in Cawley Bay (also estuarine) and in a deep-sea mud from the Tasman Sea. Coal represents, of course, an extreme case. The data of Rochford (1952) also indicate that the quotient between organic matter and iron oxides cannot be far from 3.0.

If there exists an organo-iron complex its specific gravity can now be approximated. Allowing a range between 2.5 and 3.5 parts of organic matter to one part of FeO(OH), the density range for iron-pectin complexes is 2.19 to 2.02, and for iron-cellulose complexes it is 2.24 to 2.08. Though sufficient analytical figures are not yet available, it appears that most of the iron is usually present in the sediments in the density fraction 2.0 to 2.3.

In two sediments from Port Hacking River the density fraction 1.87 to 2.37 was isolated. Table 4 shows the results of analyses of this fraction. In an artificial mud containing calcium carbonate, quartz and pectin, the fraction with density less than 2.37 was isolated after four months incubation. This fraction contained no quartz or carbonate, only organic matter and iron oxide, in the proportion 2.15:1.

DISCUSSION

Use of the densitographic method shows that most of the iron in sediments exists in complexes, chiefly with organic matter. Experience with artificial muds has shown that sulphate-reduction promotes the formation of such compound structures. Experiments with other metals, with and without iron, are under way. Though it is too early to report

TABLE 3. Organic content of sediments

TABLE 4. Ratio between organic matter and iron

TABLE 5. Organic matter and iron in Pt. Hacking River

End_Page 457------------------------

results, it may be that uranium also gathers into an organo-complex with pectin derivatives; the proportion 500 organic to 1 uranium has been observed in uraniferous black shales in the United States (Tourtelot, 1956). It may be that a further study of this matter will lead to interesting and unexpected results.

The authors are indebted to Professor G. W. Leeper (Melbourne University) for criticism and to Mr. D. Izard for much of the analytical work.