Search

We designed a reflectometer for grazing-incidence X-ray measurements with a rotational feedthrough consisting of three welded bellows and a circle-type goniometer. This apparatus does not require differential pumping and is suitable for ultra-high-vacuum applications. With this reflectometer, we successfully performed high-precision synchrotron radiation X-ray reflectivity measurements on a thin iron-film under an ultra-high-vacuum condition.

Rice straw was added to a sewage sludge digester and its effects on methane production, dewatering characteristics, and microbial communities in the digested sludge were examined by a continuous digestion experiment under mesophilic conditions (35 °C). Stable gas generation was monitored in all digestion experiments. Methane yield from raw sludge, chopped rice straw and softened rice straw were estimated to be 0.27, 0.18 and 0.26 NL/g total solids load, respectively. The capillary suction time of digested sludge was decreased by the addition of rice straw. Archaeal and bacterial communities in the sludge were elucidated by PCR-DGGE (polymerase chain reaction – denaturing gradient gel electrophoresis) targeting 16S rRNA genes. The Shannon index of DGGE profiles indicated that bacterial diversity increased with the addition of softened rice straw. DNA sequences of significant bands of the digested sludge were most closely related to Methanosaeta concilii (97.4% identity) and Methanoculleus bourgensis (100% identity). Meanwhile, those in the co-digested sludge with rice straw were most closely related to Methanosarcina barkeri (98.4% identity) and Methanoculleus bourgensis (99.3% identity). Although both Methanosaeta spp. and Methanosarcina spp. metabolize acetate to methane, Methanosarcina spp. have a competitive advantage at acetate concentrations of >70 mg/L. Results suggested that the quantity of acetate produced during rice straw degradation may change the archaeal community.

Apatite layer was formed on polyethyleneterephthalate (PET) substrate by the following biomimetic process. The PET substrate was placed on granular particles of a CaO, SiO2-based glass in simulated body fluid (SBF) with ion concentrations nearly equal to those of human blood plasma to form apatite nuclei on their surfaces. The apatite nuclei was then grown into a continuous layer by subsequently soaking the substrate in SBF under air or CO2 atmosphere in which CO2 partial pressure in the ambient was adjusted to 14.8 kPa to increase the content of carbonate ion to a level nearly equal to that of blood plasma. The increase in the content of carbonate ions in SBF changed the Ca/P atomic ratio of the apatite from 1.51 to 1.63, content of CO(3)2- ions from 2.64 to 4.56 wt %, and lattice constants a from 94.32 to 94.23 nm and c from 68.70 to 68.83 nm, respectively. The Ca/P ratio and lattice constants of the apatite formed in SBF under CO2 atmosphere were approximately identical to those of bone apatite, i.e. Ca/P atomic ratio 1.65, content of CO(3)2- ion 5.80 wt % and lattice constants a 94.20 and c 68.80 nm. This indicates that an apatite with composition and structure nearly identical to those of bone apatite can be produced in SBF by adjusting its ion concentrations including the content of carbonate ions to be equal to those of blood plasma.

An apatite layer was formed on polyethyleneterephthalate (PET) substrates by the following biomimetic process. PET substrates were placed on granular particles of a CaO-SiO2-based glass in simulated body fluid (SBF) with ion concentrations nearly equal to those of human blood plasma to form apatite nuclei on their surfaces (first treatment). They then were soaked in modified SBFs, the ion concentrations of which were changed to give a variation in ionic activity product of apatite (IP), in order to make the apatite nuclei grow (second treatment). The Ca/P atomic ratio and the lattice constant c of the formed apatite decreased from 1.54 to 1.40 and from 6.880 to 6.838 A, respectively, with increasing ion concentrations from 0.75 to 2.00 times those of SBF, that is, with increasing IP from 10−96.6 to 10−91.9. This was attributed to an increase in the concentration of HPO42− ion substituting for the PO43− ion sites, which gave an increase in the Ca2+ in the apatite. Even the apatite formed in 1.00 SBF showed a Ca/P ratio of 1.51 and lattice constants a of 9.432 A and c of 6.870 A. The Ca/P ratio and lattice constant c were smaller and the lattice constant a was larger than those of the bone apatite; its Ca/P ratio and its lattice constants a and c, were 1.65, 9.419 A, and 6.88 A, respectively. This was attributed to the lower content (2.64 wt%) of the CO32− ion substituting for the PO43− ion sites of the apatite compared to that of the bone apatite (5.80 wt%). The lower content of the CO32− ion in the apatite might be caused by the lower concentration of HCO3− ion in 1.00 SBF compared to that in human blood plasma. © 1999 John Wiley & Sons, Inc. J Biomed Mater Res, 46, 228–235, 1999.

Single crystals of carbonate apatite were grown using a CaCO3 flux under an Ar gas pressure of 55 MPa. The crystals obtained were observed by scanning electron microscopy, optical microscopy and X-ray diffraction. Electron probe microanalyses and thermal analyses were performed. CO3 ions in planar triangle form replaced both OH sites and PO4 tetrahedral sites in the apatite structure: in particular, the OH sites were perfectly substituted by CO3 ions using this method.

Thin boron films were produced on Si substrates from a solid boron source and a hydrogen plasma. The plasma was generated using a 13.56 MHz generator and films were deposited with a forward radio frequency (RF) power of 2.0 kW. At pressures from 0.931–2.26×102 Pa under high hydrogen concentrations a capacitively coupled plasma (CCP) was observed whereas at low hydrogen concentrations an inductively coupled plasma (ICP) was observed. The films were predominantly deposited with an ICP but in one case a film was deposited using a CCP discharge. The deposited films consisted primarily of boron, but they also contained oxygen and silicon. The films were amorphous at 225 and 350°C, but revealed X-ray diffractions at 475°C. It was concluded that the hydrogen concentration, RF plasma power and surface temperature as well as the plasma-boron source interactions strongly influenced the film thickness and composition.

New artificial bone materials were prepared using calcium phosphates, hydroxyapatite and β-tricalicum phosphate, and copoly-L-lactide, CPLA. Calcium phosphate powder and CPLA were mixed at 453 K for 10 min with various mixing ratios. Scanning electron microscope observations indicated that the composites of β-tricalicum phosphate and CPLA were homogeneously dispersed and highly adhesive. Young’s modulus of the composites was the same as bone, and bending strength was over half that of bone. The improvement of Young’s modulus compared to the original two materials was due to a composite effect. The composites are expected to be usable as artificial bone materials.

The surfaces of single crystals and polycrystalline specimens of hydroxyapatite were measured by auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS). Compared to P content, Ca content was found to be enriched near the surface of the hydroxyapatite single crystal. A Ca-deficient and carbonate-containing phase was observed at the grain boundaries of the polycrystalline specimen. Ca content at fracture surface increased as a function of time. It is suggested that Ca ions in the hydroxyapatite crystals migrated from inside the crystal to the surface, and that carbonate ions were incorporated from the gas phase into the crystal interior.

A bellows assembly with a radio frequency (rf) shield has been designed and developed for the KEK B factory (KEKB). The rf shield is the usual finger type but has a special spring finger to press contact fingers onto the beam tube. A test of the mechanical performance using a trial model shows good results. Further experimental studies are focused on the two key points of the finger‐type rf shield, that is, the contact force of fingers and the length of slits between adjacent fingers. First, to reduce the excess abrasion at contact points, the necessary contact force is studied experimentally utilizing microwaves. Abnormal heating is checked by transmitting a 508 MHz cw microwave through a trial model in a coaxial line. Arcing at the contact point in vacuum is observed by transmitting a 2856 MHz pulse microwave in a rectangular waveguide equipped with a model piece of the rf shield. A contact force of 50 g/finger is found to be necessary. Second, the optimum slit length is studied with a strategy to minimize the higher order mode (HOM) power leaked from the beam tube into the inside of the bellows while keeping a sufficient sliding stroke. The coupling coefficient, β, of the rf shield is measured experimentally from the Q value of a TE mode resonance in a cylindrical cavity connected to the bellows assembly. Using the measured dependence of β on the slit length, a rough estimation of leaked HOM power is tried for the KEKB. The expected leaked power is about 6–18 W for a slit length of 20 mm, which is in the allowable range. The results obtained through these experiments are reflected to the design for the KEKB: the contact force of 80–100 g/finger will be adopted finally considering the manufacturing error of typically ±10 g/finger in our test and leaving a margin for higher frequencies. The nominal slit length will be set to be 20 mm, keeping the expansion/contraction of ±10 mm.

Distributed ion pumps (DIPs) for the TRISTAN accumulation ring (TAR) have been recently designed and experimentally studied. Effects on the pumping speed are investigated in some geometrical conditions, under the spatially non-uniform magnetic field and on the distance between cathodes and anodes. The pumping speed per cell depends on the cell diameter. Furthermore, the following new facts are determined: (1) the non-uniform magnetic field reduces an effective cell diameter and causes the decrease of the pumping speed; (2) although the pumping speed per cell of the perforated plate-type anode is greater than that of the cylinder-type anode, the preference between these two anodes should be determined by considering also the total cell number and the structure of the whole DIP; and (3) there exists the most suitable value of the distance between cathodes and anodes for a given cell diameter. DIP of the cylinder-type anode with a cell diameter of 15 mm and with the distance between cathodes and anodes of 7 mm is found to be the most appropriate one for TAR in the magnetic field of 3.6 ≈ 11.5 kG. A pumping speed greater than 120 l s−1 m−1 at a pressure of ≈ 10−9 torr can be obtained after baking.

An improved empirical formula giving a pumping speed of ion pumps in the high magnetic field (HMF) mode is proposed based on Hartwig's empirical formula. Empirical formulae or theories derived so far have implicitly assumed the Penning cell with an infinite length along the axis. In this report, the total charge is calculated in the cell with a finite length using the finite element method (FEM), and the factor representing the geometrical dependence of the total charge is introduced. The improved formula is obtained by adding this new factor to Hartwig's formula. This improved formula is able to estimate the effect on the pumping speed of the cell radius, the cell height and the distance between the anodes and cathodes, and reproduces the experimental results very well.

Correlation between state of dispersion and mechanical properties is obtained and failure mechanisms from falling weight impact tests are discussed. The state of dispersion is characterized for calcium carbonate filled polypropylene model composites by using a scanning electron microscope. A dispersion index is defined in terms of the fraction of agglomerates of the filler remained in the composites. Improvement of dispersion uplifts tensile strength and more prominently falling weight impact strength. It is pointed out that an initial failure in the falling weight impact test rises at the agglomerate existing near the loading point. Falling weight impact strength is found to be dependent upon the distance from loading point to failure initiation point where the agglomerate exists. This explains the dependency of impact strength on the state of dispersion and a large sample deviation often observed in the falling weight impact test.

An experimental technique based on small-angle light scattering for characterizing dispersion of particles and presence of agglomerates in compounds of minerals in thermoplastics is described. The technique is critically discussed in terms of the Mie and Rayleigh-Gans theories of scattering. An experimental study of dispersion in compounds of polypropylene and polystyrene with calcium carbonate by mixing variously on two roll mills, an internal mixer, and a modular intermeshing co-rotating twin screw extruder.

Crab cavities have been installed in the KEKB B-factory rings to compensate the crossing angle at the collision point and thus increase luminosity. The beam operation with crab crossing has been done since February 2007. This is the first experience with such cavities in colliders or storage rings. The crab cavities have been working without serious issues. While higher specific luminosity than the geometrical gain has been achieved, further study is necessary and under way to reach the prediction of simulation.

KEKB has been operating with crab cavities for the first time in the world. This paper represents beam operation with the crab crossing.

At KEKB, a dedicated machine experiment on crab crossing has been carried out for about 4.5 months. Some of the beam-beam effects observed with crab crossing, which include a beam lifetime issue, are discussed in comparison with those with a finite crossing angle of plusmn11 mrad.

This paper presents a new scheme of fault location on power system comparing actual measurements of voltage sag to fault current calculation results. Using a two-circuit transmission line fault analysis tool developed by CRIEPI for calculating voltage sag value, this scheme can locate a faulted point not depending on causes of the faults. By estimating fault point impedance, highly accurate fault location is expected. From a viewpoint of practical use, the presented scheme is applied to a real power system model of a Japanese electric power company using practical voltage sag measurement values at substations on subtransmission system. Results of this evaluation show that this scheme has a possibility of having practical accuracy of the fault location

The bellows assembly with an RF-shield has been developed for the KEK B-Factory (KEKB). The RF-shield is a usual finger-type but has a special spring-finger to press the contact-finger to the beam tube. The mechanical workings of the RF-shield is tested using a trial model and no mechanical problem is found except for the dust generation. The necessary contact force is obtained experimentally transmitting the 508 MHz microwave up to 80 kW through the trial model. The higher order modes power leaked into the inside of bellows assembly is also estimated measuring the coupling coefficient of the RF-shield.

Single crystals of carbonate apatite were grown using a CaCO3 flux under an Ar gas pressure of 55 MPa. The crystals obtained were observed by scanning electron microscopy, optical microscopy and X-ray diffraction. Electron probe microanalyses and thermal analyses were performed. CO3 ions in planar triangle form replaced both OH sites and PO4 tetrahedral sites in the apatite structure: in particular, the OH sites were perfectly substituted by CO3 ions using this method.

Apatite layer was formed on polyethyleneterephthalate (PET) substrate by the following biomimetic process. The PET substrate was placed on granular particles of a CaO, SiO2-based glass in simulated body fluid (SBF) with ion concentrations nearly equal to those of human blood plasma to form apatite nuclei on their surfaces. The apatite nuclei was then grown into a continuous layer by subsequently soaking the substrate in SBF under air or CO2 atmosphere in which CO2 partial pressure in the ambient was adjusted to 14.8 kPa to increase the content of carbonate ion to a level nearly equal to that of blood plasma. The increase in the content of carbonate ions in SBF changed the Ca/P atomic ratio of the apatite from 1.51 to 1.63, content of CO(3)2- ions from 2.64 to 4.56 wt %, and lattice constants a from 94.32 to 94.23 nm and c from 68.70 to 68.83 nm, respectively. The Ca/P ratio and lattice constants of the apatite formed in SBF under CO2 atmosphere were approximately identical to those of bone apatite, i.e. Ca/P atomic ratio 1.65, content of CO(3)2- ion 5.80 wt % and lattice constants a 94.20 and c 68.80 nm. This indicates that an apatite with composition and structure nearly identical to those of bone apatite can be produced in SBF by adjusting its ion concentrations including the content of carbonate ions to be equal to those of blood plasma.

At the high energy ring of the KEK B-factory (KEKB), it was found that some bellows near the movable masks were overheated due to the higher order mode (HOM) as increasing the beam current over 900 mA. To cope with this problem, a new mask was designed where the length of ramps beside the mask head was expanded from 30 mm to 400 mm. MAFIA T3 simulations showed that the loss factor for the new long mask was about a half of that for the previous short one. The power of the TE mode, on the other hand, which can easily couple with the bellows through the finger-type RF-shield, was expected to reduce to about 6 percent of that for the short one. During the summer shutdown in 2002, two long masks were installed as a test. In the following run the temperature rise of bellows near the long masks was about 20 percent of those near the short ones and the new design was found to be effective to reduce the HOM. The decrease in the temperature rise was larger than the reduction of the HOM power estimated from the loss factor. This result indicates that the overheating of the bellows is mainly due to the TE mode like HOM rather than the total HOM.

In a high luminosity lepton machine such as the KEK B-factory (KEKB), the vacuum components are likely to be annoyed by intense higher order modes (HOM) due to the high beam currents. A winged HOM damper equipped with SiC HOM absorbers was developed to absorb unnecessary HOM, especially TE mode like HOM with a power of several kW. Four dampers were installed in the KEKB ring near movable masks and relieved heating of bellows and pump elements effectively at the beam current up to 1.5 A.

Conceptual design of the vacuum system for the upgrade of KEKB (Super KEKB) is presented. For its large storage current, the synchrotron radiation (SR) power is much higher than ever before. The beam chamber will have an ante-chamber structure to reduce the irradiated SR power density. To accommodate the intense higher order mode (HOM) arising from the short bunch length, the careful R&Ds are required for bellows, movable masks (collimators), HOM absorbers and so on. The first step of R&D has just begun using the KEKB.

The vacuum system for a future high luminosity lepton rings are discussed citing the upgrade plans of KEKB and PEP-II as examples. The high luminosity means firstly the large beam currents as well as the short bunch length. The vacuum system has to cope with the resultant intense synchrotron radiation power and the high photon flux. Deliberate care should be paid for the beam impedance issues. The design will proceed based on the various experiences at present rings introducing novel ideas.

We have evaluated the absorption dose by simulation in a phantom irradiated by the X‐ray from the 6MeV electron synchrotron MIRRORCLE‐6X. Comparison with the dose by the X‐ray from a 100keV X‐ray tube was made. It shows that the averaged dose integrated over the phantom to make one flame of the X‐ray image in the case of MIRRORCLE‐6X is much smaller than that in the case of the X‐ray. We have also measured the dose using MIRRORCLE‐6X under the various conditions. We could confirmed this results experimentally by observing dose at the surface, although it was slightly higher than expected.

Non-destructive and accurate measurement of zero-dispersion wavelength variation in concatenated dispersion shifted fiber link by improved four-wave-mixing technique has been proposed. The improved technique utilizes the characteristics of the four-wave-mixing efficiency that the efficiency always takes the peak value when a pump wavelength meets the zero-dispersion wavelength, independent of the a probe wavelength, and that the period of the power oscillation is different for different probe wavelengths. The validity has been verified by numerical calculations and the experiment using this technique has been successfully demonstrated.

Theoretical expression has been derived for polarization-mode dispersion with random-mode coupling in optical fiber by simple means. The square-root dependence of polarization-mode dispersion not only on a fiber length but on a mean-coupling length has been experimentally verified for the first time. This characterization became possible only by the use of highly birefringent polarization-maintaining fiber and introduction of fusion splices with desired coupling lengths. >

A novel rf shield structure for bellows chambers or gate valves in high-current accelerators is proposed. The rf shield has no thin fingers, but nested comb teeth instead at the inner surface of the beam duct. The comb-type rf shield has a higher thermal strength structurally than the conventional finger-type one. A simulation also indicates its lower impedance. Especially, the leak of a TE modelike higher order mode through the axial slits, which causes mainly heating of these components, is almost suppressed due to thick teeth. An abrasion test in a vacuum using a test piece of the comb-type rf shield shows less dust generation compared to the finger type. Another preliminary test transmitting 508 MHz microwaves through a prototype rf shield in the atmosphere found no abnormal heating or arcing. A trial model of bellows chambers with the comb-type rf shield will be installed and tested using a real beam in the KEK B-factory positron ring this year.

Uniform and low polarisation-mode dispersion, 0.052 and 0.053 ps/√km on the average, has heen confirmed in 100 fibre ribbon slotted core optical cable before and after installation, respectively. The cable has been manufactured with standard singlemode fibres having nearly optimised coating structure determined through finite element method analysis conducted to minimise lateral force induced birefringence.

A trial model of a copper beam duct for the KEK B-Factory was fabricated and gas desorption rates were measured. The duct consists of a cooling channel, a beam channel and a pump channel, which were extruded separately and welded each other by the electron beam welding. The duct received only ultrasonic cleaning in a freon bath. The ultimate total pressure was about 4/spl times/10/sup -10/ Torr after baking at 150/spl deg/C for 48 hours and the thermal gas desorption rate was calculated to be about 1.5/spl times/10/sup -12/Torr. l/(s/spl middot/cm/sup 2/) in N/sub 2/ equivalent. The measurement of the photon stimulated gas desorption rate was performed using synchrotron radiation with a critical energy of 26.3 keV from the TRISTAN Accumulation Ring. The photo-desorption coefficients of the order of 10/sup -2/ molecules/photon were observed for main gases desorbed at the initial stage. These high gas desorption rates indicate that more careful surface preparation must be necessary. >

A non-linear dependence of vacuum pressure on the beam current has been observed in the KEKB positron ring. The behaviour changes with the bunch fill pattern, the bunch current and the external magnetic field, which is similar to that of the beam size blow up phenomenon. A simple simulation on the non-linear pressure rise is proposed assuming that the behaviour occurs by the electron multipactoring. The simulation well reproduces the observed non-linear pressure behaviours. The results support a suggestion that the beam size blow up is caused by the electron cloud around the positron beam.

International audience; The structure of the oxygen-deficient compound 7Bi2O3 · 2WO3, a fluorite-derivative phase considered a candidate for electronic applications because of its high ionic conductivity, is investigated by singlecrystal X-Ray diffraction employing Ag-Kα radiation (λ = 0.5608 Å) to minimize the effect of the absorption by the heavy metals. The space-group type is I41, the acentric subgroup of I41/a that was previously suggested from powder-diffraction data and precession-camera photographs. Lattice parameters are a = 12.513(2), c = 11.231(4) (Z = 2.5). The sample is twinned by syngonic merohedry, class I, with volume ratio of the individuals 0.58/0.42. The ordering of W partly confirms previous models, with one W fully occupying one of the sites on special position. However, the remaining W goes in a site on general position, which shares with Bi, resulting thus more diluted in the structure. The oxygen vacancies are partly ordered in three of the ten anion sites.

An apatite layer was formed on polyethyleneterephthalate (PET) substrates by the following biomimetic process. PET substrates were placed on granular particles of a CaO-SiO2-based glass in simulated body fluid (SBF) with ion concentrations nearly equal to those of human blood plasma to form apatite nuclei on their surfaces (first treatment). They then were soaked in modified SBFs, the ion concentrations of which were changed to give a variation in ionic activity product of apatite (IP), in order to make the apatite nuclei grow (second treatment). The Ca/P atomic ratio and the lattice constant c of the formed apatite decreased from 1.54 to 1.40 and from 6.880 to 6.838 A, respectively, with increasing ion concentrations from 0.75 to 2.00 times those of SBF, that is, with increasing IP from 10(-96.6) to 10(-91.9). This was attributed to an increase in the concentration of HPO4(2-) ion substituting for the PO4(3-) ion sites, which gave an increase in the Ca2+ in the apatite. Even the apatite formed in 1.00 SBF showed a Ca/P ratio of 1.51 and lattice constants a of 9.432 A and c of 6.870 A. The Ca/P ratio and lattice constant c were smaller and the lattice constant a was larger than those of the bone apatite; its Ca/P ratio and its lattice constants a and c, were 1.65, 9.419 A, and 6.88 A, respectively. This was attributed to the lower content (2.64 wt%) of the CO3(2-) ion substituting for the PO4(3-) ion sites of the apatite compared to that of the bone apatite (5.80 wt%). The lower content of the CO3(2-) ion in the apatite might be caused by the lower concentration of HCO3- ion in 1.00 SBF compared to that in human blood plasma.

New artificial bone materials were prepared using calcium phosphates, hydroxyapatite and beta-tricalicum phosphate, and copoly-L-lactide, CPLA. Calcium phosphate powder and CPLA were mixed at 453 K for 10 min with various mixing ratios. Scanning electron microscope observations indicated that the composites of beta-tricalicum phosphate and CPLA were homogeneously dispersed and highly adhesive. Young's modulus of the composites was the same as bone, and bending strength was over half that of bone. The improvement of Young's modulus compared to the original two materials was due to a composite effect. The composites are expected to be usable as artificial bone materials.

Gas desorption rates for several types of B(4)C resins were investigated using a throughput method. The investigation was particularly focused on determining the out gas composition, effects of dry air, grain size (density) effects on the gas desorption rates. It is found that water is the main component of out gas and that dry air can effectively reduce gas desorption.

Rhodobacter capsulatus has been known to possess two ferredoxins (I and II) with distinct physicochemical and structural properties: ferredoxin I is a 2[4Fe-4S] type and the other is a [3Fe-4S] [4Fe-4S] type. To analyze their possible functional differences, their genes (fdxN and fdxA) were cloned, sequenced, and subjected to interposon mutagenesis experiments. The former gene was adjacent to a gene encoding a chloroplast-type [2Fe-2S] ferredoxin (fdxC). Mutants with inactivated fdxN and/or fdxC were obtained, and they showed virtually no growth under nitrogen-fixing conditions. Complementation experiments confirmed that both fdxN and fdxC were required for nitrogen fixation. On the other hand, we have not been able to disrupt fdxA under the screening conditions surveyed, including conditions that do not require nitrogenase activity for growth, suggesting that ferredoxin II could have an unknown essential role(s). These indicate functional differences among multiple ferredoxins in one bacterium other than in cyanobacterial heterocysts and indispensability of certain ferredoxins in nitrogen fixation other than Rhizobium meliloti FdxN.

Two distinct ferredoxins were purified from Rhodobacter capsulatus SB1003. Their complete amino acid sequences were determined by a combination of protease digestion, BrCN cleavage and Edman degradation. Ferredoxins I and II were composed of 64 and 111 amino acids, respectively, with molecular weights of 6,728 and 12,549 excluding iron and sulfur atoms. Both contained two Cys clusters in their amino acid sequences. The first cluster of ferredoxin I and the second cluster of ferredoxin II had a sequence, CxxCxxCxxxCP, in common with the ferredoxins found in Clostridia. The second cluster of ferredoxin I had a sequence, CxxCxxxxxxxxCxxxCM, with extra amino acids between the second and third Cys, which has been reported for other photosynthetic bacterial ferredoxins and putative ferredoxins (nif-gene products) from nitrogen-fixing bacteria, and with a unique occurrence of Met. The first cluster of ferredoxin II had a CxxCxxxxCxxxCP sequence, with two additional amino acids between the second and third Cys, a characteristics feature of Azotobacter-[3Fe-4S] [4Fe-4S]-ferredoxin. Ferredoxin II was also similar to Azotobacter-type ferredoxins with an extended carboxyl (C-) terminal sequence compared to the common Clostridium-type. The evolutionary relationship of the two together with a putative one recently found to be encoded in nifENXQ region in this bacterium [Moreno-Vivian et al. (1989) J. Bacteriol. 171, 2591-2598] is discussed.