Examination of Leg 207 black shales indicates that abundant material with very good to excellent preservation of mid-Cretaceous foraminifers can be analyzed for paleoceanographic data. However, past attempts to separate clean tests from the organic-rich mudstones using a Calgon-hydrogen peroxide solution have proved time consuming and generally unsatisfactory. We describe here a series of cleaning experiments performed to determine which solvents and steps are most effective at disaggregating organic mudstone clasts, separating foraminifer tests from the matrix, and removing dark matter from the surface of foraminifers. For our purposes, it was most important that the oxygen and carbon stable isotope ratios are not affected by the cleaning procedure.
Experiments were performed on splits of a large (~60 g dry weight) black shale sample (Sample 207-1258B-51R-2, 13–20 cm) that remained after aliquots were taken for lipid analyses. Samples from nearby composite depths in this hole that were measured shipboard contained 9.5–12.4 wt% total organic carbon (Shipboard Scientific Party, 2004a). The sediments are Cenomanian in age and are assigned to planktonic foraminiferal Zone KS19. The freeze-dried, crushed sample was homogenized to minimize biases among subsequent splits. Cleaning tests were performed on ~5-g sample splits in 50–60 mL of solvent. The four cleaning procedures are given in Table AT1. Washing was done with warm tap water (pH = 6) through a 63-µm sieve. Wet samples were dried in an oven at 45°C for 1–2 hr.
After final drying, 15–20 individual whole specimens of Whiteinella baltica were picked from the >150-µm size fraction from each procedure. In general, foraminiferal preservation in Sample 207-1258B-51R-2, 13–20 cm, is moderate to good. This particular interval does not exhibit the very good and excellent preservation that has been observed in other Demerara Rise Cenomanian and Turonian samples, but it was the only sample available to us that had sufficient volume for multiple procedures. An attempt was made to select the cleanest, best preserved examples of W. baltica from each procedure. These were then divided to yield five stable isotope measurements from each cleaning procedure. Oxygen and carbon isotope ratios were measured on a Finnigan MAT 252 mass spectrometer with an automated Kiel carbonate device at WHOI. Instrument precision is ±0.07 for
18O and ±0.03‰ for
13C. The solvent in Procedure 1 is a 3% Calgon-hydrogen peroxide solution that has been used for soaking most Cretaceous and Paleogene carbonate sediment samples processed in the WHOI paleoceanography group's laboratory. For that reason, foraminifers picked from the Procedure 1 sample are taken as our control for comparison of isotopes.
Our qualitative assessment of the relative effectiveness of the cleaning procedures runs in reverse order to the procedure numbers: Procedure 4 yielded the cleanest sample; Procedure 1 retained the most organic matter. Procedure 1 also resulted in the most flakes of clay and organic compounds adhering to foraminifers and significantly higher dry weights than Procedure 4. Visual examination of the tests revealed no major difference in preservation quality, although sonicated samples tended to have a higher proportion of broken specimens and samples washed with the Calgon-peroxide solution were more prone to foraminifer breakage because of the need to physically rub the clay chips though the screen. In contrast, samples soaked in bleach did not require physical disaggregation during washing, particularly if two cycles of soaking, washing, and drying were employed.
Results of the 20 stable isotope measurements on specimens from Procedures 1–4 are given in Table AT2 and are reported relative to the Vienna Peedee belemnite (VPDB) isotope standard. The differences among the mean
18O or
13C values are not significant at the 95% confidence level (Fig. AF1). The variance in oxygen isotope values from Procedure 1 (our control) is an order of magnitude greater than that for Procedures 2, 3, and 4. However, additional tests would be required to determine if this is a robust observation and whether the difference in variance is significant.
For Procedures 1 and 3, many large (>3 mm) grains remained even after the second soak (15 hr) and wash. A second sonication disaggregated most of these grains, but the final coarse fraction percentages for Procedures 1 and 3 (Table AT1) are lower than those for the procedures using bleach. This may be the result of fragmentation during sonication or the very long second soak. Although sonication quickly produced very clean foraminifers, it is our opinion that sonication should be avoided because of possible breakage. Under the binocular microscope, it appears that the smallest grains remaining in Procedure 1 are primarily individual chambers. In contrast, the smallest size fraction for Procedure 4 (no sonication) contains abundant whole shells. The only sample that disaggregated well without sonication was that in Procedure 4 (100% bleach).
In addition to the 2-hr/1-hr soak in 100% bleach (Procedure 4), we also tested a 1-hr/1-hr soak procedure with 100% bleach, in order to see if processing time could be decreased. However, cleaning results were not as satisfactory as those obtained with Procedure 4. We note that, regardless of the solvent used, the intermediate drying stage is critical in obtaining the greatest disaggregation possible: most of the disaggregation occurs in the second soak/wash cycle.
It should be noted that hydrogen peroxide, an ingredient in the solution used in Procedure 1, is strongly corrosive to calcium carbonate (Pingitore et al., 1993). While fragmentation during sonication may help account for the very low coarse fraction remaining in the Procedure 1 sample, it is also possible that the acidity (pH = 6) of the Calgon (sodium hexametaphosphate) and hydrogen peroxide solution causes dissolution of some grains. To better identify the primary cause of low final material weight and to test the procedures on a more carbonate rich black shale interval, we performed a set of experiments similar to Procedures 1 and 4. The Calgon-peroxide and 100% bleach experiments were repeated. For both solvents, samples were soaked for 3 hr, washed, dried, soaked for 1 hr, washed, dried, and weighed (Table AT3). The black shale samples used in this case were taken from Hole 1259C, containing 6–9 wt% total organic carbon (Shipboard Scientific Party, 2004b). Sample 207-1259C-11R-4, 130–150 cm, is from the Santonian, near the top of the Leg 207 black shale sequence. Preservation is good to very good. Foraminifer tests are opaque, but there is no evidence of mineral infilling or overgrowths. Sample 207-1259C-17R-1, 116–136 cm, is Turonian in age and occurs below the distinctive glauconitic claystone seen in the Turonian interval at Sites 1259, 1260, and 1261. This Turonian sample contains several thin carbonate stringers and preservation of foraminifers ranges from poor to moderate. Calcite spar infilling is common in this sample.
Again, the 100% bleach solvent produced a clean, well-disaggregated sample without sonication (Fig. AF2). Because many large mudstone grains still remained in the samples soaked in Calgon solution, these sediments were wetted with Calgon solution for 5 min, sonicated for 90 s, washed, dried, and reweighed. The reduction in coarse fraction weight percent (Table AT3) for the Calgon-peroxide solution in this second experiment was not nearly as great as that observed in Procedure 1 (Table AT1), either before or after sonication. This suggests that the very low coarse fraction in Procedure 1 may have resulted from soaking in the Calgon-peroxide solution overnight. To check the effect of long-term soaking in Calgon-peroxide, the Hole 1259C samples were soaked overnight (15 hr) in the solution, washed, dried, and weighed a third time (Table AT3). For both samples, the percent reduction in coarse fraction after the 15-hr soak was approximately equal to that caused by sonicating for 90 s.
It is clear from our second set of experiments that different black shales will exhibit varying degrees of disaggregation with the same procedure. For the Hole 1259C Santonian sample, this final soak (15 hr) resulted in an acceptably clean sample, one from which many clean, well-preserved tests could be picked fairly easily. However, in the Hole 1259C Turonian sample, which had a higher percentage of secondary carbonate cement ("stringers"), the sample left was predominantly shale and cement fragments. Besides a few foraminifers visible encased within shale matrix, almost no tests remained. In fact, this Turonian sample did not disaggregate as well as either the Hole 1258B Cenomanian shale or the Hole 1259C Santonian shale in either the Calgon or bleach solvents. However, for all three shales, the bleach soaking procedure produced a more than sufficient number of loose, clean foraminifers for isotope work. We saw no samples in which it was necessary to soak overnight in bleach to obtain a well-disaggregated sample, perhaps because the intermediate drying stage we used is so effective at causing shale grains to fall apart. However, the very high pH of the bleach suggests that no loss of carbonate material is likely to occur through dissolution if samples are soaked overnight in bleach.
In some regions, Clorox brand regular bleach may not be available. In the United States, for example, many stores carry a less expensive bleach that is 5.25%, rather than 6% sodium hypochlorite. We therefore repeated Procedure 4 in a side-by-side test using the Hole 1258B Cenomanian sample with Clorox bleach (6% NaOCl), America's Choice bleach (5.25% NaOCl) and Stop & Shop bleach (5.25% NaOCl). The difference in cleaning power between the two concentrations of sodium hypochlorite is readily apparent, with the Clorox bleach producing a cleaner sample. Because we do not know the composition of the 94.75% inert ingredients listed for the weaker bleaches, we can not say if it is the concentration of NaOCl or the inert ingredients themselves that is critical to obtaining a clean sample with bleach. While we recommend that the 6% NaOCl bleach be used, if it is available, the 5.25% bleach produced a sample that was acceptably clean for picking foraminifers.
There is reason to think that use of a 20% bleach solution (Procedure 2) should be avoided because diluting bleach lowers its pH. Gaffey and Bronnimann (1993) examined the textural effects of bleach on modern biogenic carbonates and determined that the increase in concentration of hypochlorous acid in dilute bleach solutions caused minor pitting of skeletal material. We did not bother to examine our Procedure 2 samples for pitting because this procedure yielded the second worst result in terms of organic matter removal and has no other advantages.
As with any chemical, laboratory personnel should be familiar with material safety information for bleach. If undiluted bleach is used in the laboratory, care should be taken to avoid reaction with agents that will produce chlorine gas. Having the beakers in a hood while soaking was sufficient to prevent bleach odors in the laboratory. Prolonged contact of bleach with some metals in the laboratory may cause pitting and discoloration. We noticed no effects of the beach on stainless steel (including sieves) in the laboratory when processing the black shale samples. However, we observed that ODP Leg 207 nannofossil chalks that contain small amounts of oxidized iron (such as Sample 207-1258B-39R-5, 115–135 cm) caused minor amounts of iron oxides to be deposited on the sieve wire mesh during drying in the oven. This occurred if the chalks were soaked in bleach but not when they were soaked in the Calgon-peroxide solution. We believe that the higher solubility of iron oxides in the high-pH (>11) bleach allowed dissolution of oxides in the beaker while soaking; these oxides were reprecipitated on the sieve while drying, even though the sediments were washed thoroughly. These nannofossil chalk samples disaggregated much better in the Calgon-peroxide solution than in the bleach, apparently because of dissolution of intergranular carbonate cements in the pH 6 solution. Based on these experiments, we caution that soaking in bleach should be performed only on organic-rich mudstone samples and not on chalk samples.
There are no significant differences among the oxygen or carbon isotope ratios obtained for Whiteinella baltica from the four cleaning procedures given in Table AT1. However, Procedure 4 yielded the best result in terms of cleaning foraminifer tests. Soaking in 100% bleach produced a well-disaggregated sample, with little organic matter adhered to tests and the lowest amount of fragmentation. The low mass of material remaining at the end of Procedure 1, as well as the near loss of all foraminifers in a second test, suggests that corrosion of biogenic carbonates may occur with prolonged soaking in a Calgon and hydrogen peroxide solution. Although frequently used for disaggregating carbonate-rich samples, this solution was the least effective in removing organic matter from the Cenomanian black shale sample.