Clay particle aggregation affects a number of environmental processes, such as contaminant sorption/desorption, particle movement/deposition, and sediment structure and stability, yet factors that control clay aggregation are not well understood. This study was designed to investigate how microbial reduction of Fe(III) in clay structure, a common process in soils and sediments, affects clay-particle aggregation. Microbial Fe(III) reduction experiments were conducted with Shewanella putrefaciens CN32 in bicarbonate buffer with structural Fe (III) in nontronite as the sole electron acceptor, lactate as the sole electron donor, and AQDS as an electron shuttle. Four size fractions of nontronite (D5–D95 of 0.12–0.22 µm, 0.41–0.69 µm, 0.73–0.96 µm and 1.42–1.78 µm) were used to evaluate size-dependent aggregation kinetics. The extent of Fe(III) bioreduction and the amount of exopolysaccharide (EPS), a major biopolymer secreted by CN32 cells during Fe(III) bioreduction, were measured with chemical methods. Nontronite particle aggregation was determined by photon correlation spectroscopy and scanning electron microscopy. The maximum extent of Fe(III) bioreduction reached 36% and 24% for the smallest and the largest size fractions, respectively. Within the same time duration, the effective diameter, measured at 95% percentile (D95), increased by a factor of 43.7 and 7.7 for these two fractions, respectively. Because there was production of EPS by CN32 cells during Fe(III) reduction, it was difficult to assess the relative role of Fe(III) bioreduction and EPS bridging in particle aggregation. Thus, additional experiments were performed. Reduction of Fe(III) by dithionite was designed to examine the effect of Fe(III) reduction, and pure EPS isolated from CN32 cells was used to examine the effect of EPS. The data showed that both Fe(III) reduction and EPS were important in promoting clay mineral aggregation. In natural environments, the relative importance of these two factors may be dependent on local conditions. These results have important implications for understanding factors in controlling clay particle aggregation in natural environments.

Fe-reducing micro-organisms can change the oxidation state of structural Fe in clay minerals. The interactions with complex clays and clay minerals in natural materials remain poorly understood, however. The objective of this study was to determine if Fe(III) in loess was available as an electron acceptor and to study subsequent mineralogical changes. The loess samples were collected from St. Louis (Peoria), Missouri, USA, and Huanxia (HX) and Yanchang (YCH), in the Shanxi Province of China. The total Fe concentrations for the three samples was 1.69, 2.76, and 3.29 wt.%, respectively, and Fe(III) content was 0.48, 0.69, and 1.27 wt.%, respectively. All unreduced loess sediments contained Fe (oxyhydr)oxides and phyllosilicates. Bioreduction experiments were performed using Shewanella putrefaciens CN32 with lactate as the sole electron donor and Fe(III) in loess as the sole electron acceptor with and without anthraquinone-2, 6-disulfonate (AQDS) as an electron shuttle. Experiments were performed in non-growth (bicarbonate buffer) and growth (M1) media. The unreduced and bioreduced solids were analyzed by X-ray diffraction, Mössbauer spectroscopy, diffuse reflectance spectroscopy, and scanning electron microscopy/energy dispersive spectroscopy. Despite many similarities among the three loess samples, the extent and rate of Fe(III) reduction varied significantly. In the presence of AQDS the extent of reduction in the non-growth experiment was 25% of total Fe(III) in HX, 34% in Peoria, and 38% in YCH. The extent of reduction in the growth experiment was 72% in HX, 94% in Peoria, and 65% in YCH. The extent of bioreduction was less in the absence of AQDS. Overall, AQDS and the M1 growth medium significantly enhanced the rate and extent of bioreduction. Fe(III) in (oxyhydr)oxides and phyllosilicates was bioreduced. Siderite was absent in control samples, but was identified in bioreduced samples. The present research suggests that Fe(III) in loess sediments is an important potential source of electron acceptors that could support microbial activity under favorable conditions.

Electron energy-loss spectroscopy (EELS), energy-filtered transmission electron microscopy (EFTEM), and high-resolution transmission electron microscopy (HRTEM) have been applied in mineralogy and materials research to determine the oxidation states of various metals at high spatial resolution. Such information is critical in understanding the kinetics and mechanisms of mineral–microbe interactions. To date, the aforementioned techniques have not been applied widely in the study of such interactions. In the present study, the three techniques above were employed to investigate mineral transformations associated with microbial Fe(III) reduction in magnetite. Shewanella putrefaciens strain CN32, a dissimilatory metal-reducing bacterium, was incubated with magnetite as the sole electron acceptor and lactate as the electron donor for 14 days under anoxic conditions in bicarbonate buffer. The extent of bioreduction was determined by wet chemistry and mineral solids were investigated by HRTEM, EFTEM, and EELS. Magnetite was partially reduced and biogenic siderite formed. The elemental maps of Fe, O, and C and red-green-blue (RGB) composite map for residual magnetite and newly formed siderite were contrasted by the EFTEM technique. The HRTEM revealed nm-sized magnetite crystals coating bacterial cells. The Fe oxidation state in residual magnetite and biogenic siderite was determined using the EELS technique (the integral ratio of L3 to L2). The integral ratio of L3 to L2 for magnetite (6.29) and siderite (2.71) corresponded to 71% of Fe(III) in magnetite, and 24% of Fe(III) in siderite, respectively. A chemical shift (~1.9 eV) in the Fe-L3 edge of magnetite and siderite indicated a difference in the oxidation state of Fe between these two minerals. Furthermore, the EELS images of magnetite (709 eV) and siderite (707 eV) were extracted from the electron energy-loss spectra collected, ranging from 675 to 755 eV, displaying different oxidation states of Fe in the magnetite and siderite phases. The results demonstrate that EELS is a powerful technique for studying the Fe oxidation-state change as a result of microbial interaction with Fe-containing minerals.

Shewanella putrefaciens CN32 reduces Fe(III) within two illites which have different properties: the Fithian bulk fraction and the <0.2 µm fraction of Muloorina. The Fithian illite contained 4.6% (w/w) total Fe, 81% of which was Fe(III). It was dominated by illite with some jarosite (∼32% of the total Fe(III)) and goethite (11% of the total Fe(III)). The Muloorina illite was pure and contained 9.2% Fe, 93% of which was Fe(III). Illite suspensions were buffered at pH 7 and were inoculated with CN32 cells with lactate as the electron donor. Select treatments included anthraquinone-2,6-disulfonate (AQDS) as an electron shuttle. Bioproduction of Fe(II) was determined by ferrozine analysis. The unreduced and bioreduced solids were characterized by Mössbauer spectroscopy, X-ray diffraction and transmission electron microscopy. The extent of Fe(III) reduction in the bulk Fithian illite was enhanced by the presence of AQDS (73%) with complete reduction of jarosite and goethite and partial reduction of illite. Mössbauer spectroscopy and chemical extraction determined that 21–25% of illite-associated Fe(III) was bioreduced. The extent of bioreduction was less in the absence of AQDS (63%) and only jarosite was completely reduced with partial reduction of goethite and illite. The XRD and TEM data revealed no significant illite dissolution or biogenic minerals, suggesting that illite was reduced in the solid state and biogenic Fe(II) from jarosite and goethite was either released to aqueous solution or adsorbed onto residual solid surfaces. In contrast, only 1% of the structural Fe(III) in Muloorina illite was bioreduced. The difference in the extent and rate of bioreduction between the two illites was probably due to the difference in layer charge and the total structural Fe content between the Fithian illite (0.56 per formula) and Muloorina illite (0.87). There may be other factors contributing to the observed differences, such as expandability, surface area and the arrangements of Fe in the octahedral sheets. The results of this study have important implications for predicting microbe-induced physical and chemical changes of clay minerals in soils and sediments.