Dating of Pore Waters with 129I: Relevance for the Origin of Marine Gas Hydrates

See allHide authors and affiliations

Science  29 Sep 2000:
Vol. 289, Issue 5488, pp. 2332-2335
DOI: 10.1126/science.289.5488.2332


Pore waters associated with gas hydrates at Blake Ridge in the Atlantic Ocean were dated by measuring their iodine-129/iodine ratios. Samples collected from sediments with ages between 1.8 and 6 million years ago consistently yield ages around 55 million years ago. These ages, together with the strong iodine enrichment observed in the pore waters, suggest that the origin of iodine is related to organic material of early Tertiary age, which probably is also the source of the methane in the gas hydrates at this location.

Gas hydrates are a potentially large source of energy and of greenhouse gases (1). Recent surveys suggest that they are ubiquitous features of continental slopes and that the energy equivalent of the methane in hydrates could potentially surpass that of all the known reservoirs of crude oil and natural gas combined. Marine gas hydrates commonly consist of methane trapped in a lattice of water ice (2) and are found in a specific depth range in marine sediments, typically between 200 and 500 m below the seafloor (mbsf). On seismic profiles, the bottom of the gas hydrate zone is indicated by a strong reflector, the bottom-simulating reflector (BSR). Below the BSR, free methane gas is common; its concentration is, however, considerably lower than the methane bound as gas hydrates.

The origin and formation of gas hydrates have been the focus of a growing number of studies (1–3). Although it is generally assumed that gas hydrates are related to the deposition and subsequent diagenesis of organic matter in marine sediments, no consensus exists on whether the methane in the hydrates formed at their present location or migrated from different source areas. To address this question, we used the 129I system to date the origin of these hydrocarbons.

The cosmogenic radioisotope 129I (half-life of 15.7 million years) is produced by the spallation of Xe isotopes in the atmosphere and by spontaneous fission of 238U in the crust. Both of these mechanisms contribute similar amounts of natural 129I to surface reservoirs (4). Iodine moves quickly through most surface reservoirs and is considered isotopically homogeneous at the surface of Earth, including the oceans and shallow sediments. The isotopic equilibrium of I in surface reservoirs has been disturbed recently by releases from weapons tests and reprocessing plants, which have increased the concentrations of 129I in fresh waters, soils, surface ocean water, and shallow sediments by several orders of magnitude (5–8).

Iodine has one stable isotope, 127I, and its isotopic composition is reported as 129I/I. Recent sediments without anthropogenic I have 129I/I of 1.5 × 10−12 (5, 9). This ratio, which is used as a starting value for 129I-based age calculations, is more than two orders of magnitude above the detection limit of accelerator mass spectrometry (AMS), the method of choice for the detection of cosmogenic radioisotopes. Although the generally low concentrations of iodine in many environments often make it difficult to collect enough sample material, the very high concentrations of iodide found in the pore waters at Blake Ridge (10) allowed the preparation of an AMS sample from the 3 to 5 ml collected from the sediment cores. The half-life of 129I, together with the observed input ratios and the detection limit for 129I/I, allows applications of this dating method within a range of about 80 million years ago (Ma). The biophilic nature of iodine suggests that this isotopic system is particularly useful for the dating and tracing of organic material (11–15).

The samples used in this study were collected during Leg 164 of the Ocean Drilling Program (ODP) from Site 997 on Blake Ridge. Blake Ridge is a large drift deposit that extends about 400 km in a southeast direction from the continental rise of the passive margin of the southeastern United States. The top of the ridge is at depths of 2500 to 3000 m (Fig. 1). The sedimentary package in this area is up to 10 km thick, consisting of sediments deposited continuously from the Jurassic to the present (16). The presence of gas hydrates at Blake Ridge is identified by an essentially continuous BSR at depths of 450 to 500 mbsf. The upper boundary of the hydrate layer is not as well defined, with estimates varying between 240 mbsf (17) and 24 mbsf (10,18). Free methane is found below the BSR, with dissociation of gas hydrates occurring above the gas hydrate layer. Core recovery from this site was excellent, and pore waters were collected from 0 to 750 mbsf, covering the full range of gas hydrate occurrence.

Figure 1

Location map of Blake Ridge with ODP Site 997 and the USGS line 32. Shaded area indicates the occurrence of gas hydrates, based on the presence of a BSR. Water depth at Site 997 was 2770 m; the BSR was penetrated at 451 mbsf.

For this study, we selected seven samples from the collection of pore water samples from ODP Site 997 (Table 1) (19). The pore water samples had between 3 and 5 ml of fluid with iodide concentrations between 1.19 and 1.72 mM. The lowest concentration was found in the top sample, whereas concentrations in the other six samples fell into a quite narrow range between 1.5 and 1.72 mM. The pore waters show a marked increase in iodide concentrations from levels below 0.001 mM close to the sediment-water interface to values up to 1.5 mM at depths below 100 mbsf. This strong increase in iodide levels is accompanied by a parallel increase in bromide concentrations and is in contrast to the behavior of chloride concentrations, which decrease from a surface value of 560 mM to values close to 520 mM throughout the interval with some variations in the area occupied by gas hydrates (Fig. 2). These values can be compared with seawater concentrations of 533 mM for chloride, 0.839 mM for bromide, and 0.00045 mM for total dissolved iodine (20). Although chloride concentrations in these sediments are close to seawater concentrations, iodide in these pore waters is enriched by more than three orders of magnitude and bromide by a factor of 4. Comparison with seawater concentrations indicates that the main source of chloride is probably connate seawater but that iodide (and bromide to a lesser degree) must have an additional source in order to explain the observed strong enrichment. Given the biophilic character of iodine, it is likely that the enrichment reflects the release of iodide from the decomposition of organic matter in marine sediments (21) and is thus directly related to the origin of the gas hydrates. Bromide enrichment is also commonly related to the release from organic sources. A good correlation exists between Cl concentrations and I and Br concentrations (Fig. 3). The concentrations of Br and I in these pore waters are much higher than in seawater, in contrast to the Cl values, which essentially reflect seawater concentrations, diluted perhaps by low-chlorinity water advected from below (18) or by the dissociation of gas hydrates.

Figure 2

Concentration profiles for chloride (diamonds), bromide (squares), and iodide (triangles) at Site 997. Solid symbols indicate the samples used for the 129I/I determinations.

Figure 3

Br (squares) and I (triangles) concentrations increase linearly with Cl concentration. Pore fluids have Cl concentrations close to but slightly lower than seawater [solid square, marine Br; solid triangle, marine I, plotted at Cl seawater concentrations (20)] but are enriched in Br by a factor of 4 and in I by a factor of 4000. Although the degree of enrichment is very different, the close correlation between Br and I concentrations in the pore fluids suggests a common diagenetic source as cause for the enrichment.

Table 1

Concentrations, isotope ratios, and ages of pore waters. Concentrations have 1σ uncertainties of 3.3% (Cl), 2.3% (Br), and 0.48% (I). Sed. age, depositional age of sediments containing hydrates; fluid age, sediment age corrected for expulsion of pore waters because of compaction; min. age, age based on decay of initial ratio of 129I/I = 1500 × 10−15; corr. age, age corrected for in situ buildup of 129I in deep sediments.

View this table:

Iodine extracted from the pore water samples was used to prepare samples for the determination of 129I/I ratios by AMS (22). All the 129I/I values are considerably below the value for recent marine sediments (1.5 × 10−12), indicating that old iodine is indeed present in these samples. With the use of this ratio as the input ratio for marine iodine, ages can be calculated on the basis of the decay constant of129I (4.4 × 10−8 year−1). Except for the deepest and most shallow samples, all the ages are close to 50 Ma; the other two ages are lower by about 5 Ma.

The ages found for the pore waters need to be put into the context of the depositional environment. Figure 4compares the iodine ages with the ages determined for the deposition of the sediments in Site 997, which cover a range between 1.8 and 5.5 Ma. During the deposition of sediments and subsequent compaction, waters are squeezed out and forced into the new, additional layers on top. Therefore, pore waters are considered to be older than the age of the associated sediments. The difference between these ages has been modeled by assuming constant rate of deposition and no migration of pore waters (23). Even considering this correction (Fig. 4), iodine ages indicate that iodine was not deposited together with these sediments and had to have migrated into the sediments sometime after their deposition.

Figure 4

Age-depth profile for sediments (triangles), sediment pore fluids after correction for compaction (circles), minimum iodine ages (diamonds), and iodine ages corrected for in situ production of129I from uranium (squares).

Two potential corrections for the iodine ages need to be considered, correction for the addition of in situ–produced129I and for the introduction of anthropogenic129I. Because both of these processes would have added129I to the pore waters, the calculated iodine ages are minimum values. The element distribution found for the pore water samples makes contamination by surface components highly unlikely, so that correction for anthropogenic 129I does not seem necessary here. The addition of in situ–produced 129I could have occurred either in the source formations or in the sediments where it was found, the reservoir formations. Time-dependent in situ addition is mainly a function of the concentration of uranium and the escape probability of the fissiogenic 129I (24).

If it came from the reservoir formations, the amount of129I added can be calculated with the observed uranium concentrations, porosities, and the ages of the formations. If we assume that all the 129I produced by the spontaneous fission of 238U in the reservoir sediments has found its way into the pore waters, an upper limit for this correction results. Uranium concentrations in these sediments are between 2 and 3 parts per million (ppm), with porosities around 50%. Given these figures, we can calculate additions both for sediment ages (between 1.8 and 5.5 Ma) and pore water ages (between 4.6 and 10.3 Ma). The maximum correction based on these assumptions is a lowering by 12 × 10−15 of the ratio for the deepest sample. Because this change in ratio corresponds to an increase in age of only slightly more than 1 Ma, it is well within the error limits of the determinations and therefore will be ignored in further discussion.

A larger correction could be made on the basis of the residence time of the pore waters in deeper layers. Because these fluids must have resided somewhere in the sediment pile for the time determined before, it is quite likely that fissiogenic 129I has been added during that time. Although the specific location of residence for these waters is not known, uranium concentrations and porosity values probably were not markedly different from the values found in the shallow sediments. As a first approximation, we can therefore apply the same approach as before, using, however, 3 ppm of uranium and the ages determined by the 129I/I ratios. Because longer times are involved in this calculation, the corrections are more substantial than those determined for the shallow sediments. The average age of the pore waters increases from 50.5 to 56.5 Ma; individual corrected ages are also listed in Table 1. Because in all these cases we assumed that all of the fissiogenic 129I escaped from the sediments into the pore waters, the results of these calculations are probably too high. Although corrections for in situ production can be substantial, they do not change the overall range of results, nor do they seem to influence the differences between the individual results.

The most striking result of these determinations is that all the ages for the pore waters are consistently and substantially older than the age of the host sediments. The ages indicate that the source of iodine in the pore waters is from the early Tertiary. According to a cross section of this area [U.S. Geological Survey (USGS) line 32], sediments of this age are present in this area, generally between 1 and 3 km below the sea floor (16). Although the ages determined here reflect strictly only the origin of the iodine in the pore waters, it seems reasonable to assume that iodine migrated together with the fluids into their current positions. Given the geochemical behavior of iodine, the most likely cause for the observed strong enrichment is the release of iodine during the diagenesis of organic material of early Tertiary age. A large amount of organic material is needed to explain the enrichment of the waters by factors of more than 4000 over the typical concentrations in seawater. This organic material likely was also the origin of the methane found together with the pore waters in the form of gas hydrates. This suggests that the organic parent material of the gas hydrates is of early Tertiary age and that gases and waters migrated together to their current positions from depths between 1 and 3 km below the sea floor.

Iodide concentrations are not routinely measured in pore waters, so it is not known to what extent the observed correlation between high iodide concentrations and hydrate occurrence is representative. If such a correlation is common, the 129I method could be used on a broad scale to determine the age of the sources of gas hydrates. The age found for this deposit, close to 55 Ma, is one in which large depositions of marine organic material have been observed, perhaps associated with a large-scale change in oceanic circulation (25). Source ages determined with the 129I system in the Gulf Coast region (12) are also close to those of this study. If iodine ages determined in pore fluids associated with other gas hydrate deposits fall into the same time range, the case would be strengthened that the period around 55 Ma was responsible for large-scale deposition of organic matter and subsequent hydrocarbon generation in the oceans.

  • * To whom correspondence should be addressed. E-mail: fehn{at}


View Abstract

Stay Connected to Science

Navigate This Article