The present study area covers Kurigram sadar, Ulipur and Chilmari Upazila of Kurigram district which is traversed by Brahmaputra (Jamuna) river (25˚48'26"; 89˚45'09" - 25˚33'32"; 89˚41'01"). The study area covers stable sand bars in an area of 30 km × 4 km from upstream to downstreamof the Brahmaputra River in the Kurigram district, north-eastern Bangladesh (Figure 1).

The present study covers about 30 km × 4 km area from which 14 successive locations were selected and numbered as DK₁-DK₁₄. From each location two samples were collected from sand bar (“chars”) and river bed found in the study area. The sample locations were recorded in terms of degree-minute-second (latitudinal and longitudinal position) using a hand-held global positioning system (GPS) (Model: Magellan-Map-410) unit. Each location is separated by a distance of approximately 2 km (approximately 1 minute in co-ordinate system). The samples were collected at the depth of 0 - 5 cm by plastic spade during summer period of 2014 and collected samples were packed in polyethylene bags. Each sample has the weight of about 2 kg. The collected samples were air dried at room temperature in open air.

Preparation and preliminary analysis of sediment samples were carried out as per the procedure given in Carver [7] and Solai et al. [8] . Heavy mineral separation was carried out using heavy liquid bromoform of 2.89 specific gravity as per the standard procedures outlined by Milner [9] . The total heavy mineral percentages (THM) were calculated. Using a hand magnet, magnetic minerals (magnetite) were separated from the heavy mineral fractions and the weight percentage was estimated. The samples were also examined to measure their granulometric fractions such as contents of sand, silt and clay. Using an ASTM sieve, about 100 g of sediment was taken for separation of sand, silt and clay fractions by sieving.

For mechanical separation 10 kg raw sand samples were dried in the sun to remove moisture. Separation of heavy and light minerals was carried out depending on specific gravity/density of the minerals using laboratory shaking table. For each sample, heavy minerals fraction obtained after density separation dried in burner. Magnetite, ilmenite, rutile and zircon mixed with others were separated using Induced Roll Magnetic Separator (IRMS) applying fixed magnetic intensity at 0.5 amp, 2.5 amp, 5.5 amp, and 7.5 amp condition, respectively.

Powder X-ray Diffraction (XRD) is a versatile technique that can be used to identify any crystalline substances, such as most minerals. It can also be used to quantify the proportions of different minerals or indeed many other substances when they are present in a mixture. Powder X-ray diffraction analysis is seemingly the perfect technique for crystalline-mixture analysis, since each component of the mixture produces its characteristic pattern independently of the others, making it possible to identify the various components by unscrambling their superposed patterns. An X-ray beam of known wavelength is focused on a powdered sample and X-ray diffraction peaks are measured using a detector. In the present study X-ray diffraction technique has been utilized to discern these properties. The phase identification was performed using Philips X’Pert Pro multipurpose X-ray diffractometer which uses copper Kα radiation. X-ray generator operated at 40 kV and 30 mA, and all scans were run at a speed of 2˚ per minute, from 10˚ to 70˚, with a continuous step size of 0.02.

The heavy mineral separation analysis has been carried out to know the total heavy mineral (THM) weight percentage in the present samples. Weight percentage of separated heavy minerals in most of the samples varied from 8% to 15%, while in one sample it reaches to 25%; similar THM value is reported for Brahmaputra river sand [10] . The THM percentage is varied greatly from site to site. It is minimum in DK₁₃ and it is maximum in DK₄ (Figure 2). Other sites are having intermediate values.

Average THM value in the sand samples of the study area is 12.40% while low (within10%) THM percentage is encountered in four locations. These values of THM are higher than the other river of Bangladesh as observed in someshari river [11] . The actual sorting and concentration of heavy minerals takes place due to the actions of two principal agents i.e. action of waves/currents and wind. Mechanical separation of raw sand using Induced Roll Magnetic Separator (IRMS) resulted to separate magnetite, ilmenite, rutile and zircon mixed with other minerals. Separation of samples DK₁, DK₈ and DK₁₂ reveals that ilmenite and rutile are main contributors of total heavy mineral (THM). Titanium bearing minerals ilmenite and rutile contributes 26.22% - 31.01% and 14.57% - 23.50% of THM.

Granulometric analysis has been carried out to know the content of sand, silt and clay (%) in the present sediments. Percentage of sand, silt and clay (%) content of the samples are presented in Figure 3. Sand is the main constituent in all the sampling locations, which varied from 91.87% to 99.74%. Silt content is the least constituent which varied from 0.03% to 7.98%. In the samples studied, the heavy minerals occur predominantly in the 60- to 200-mesh sieve grade (Table 1). 75% - 90% heavy minerals of THM occur in 60- to 120-mesh while some heavy mineral grains are finer but most of the samples contain negligible amounts of material of this grain size. In a given sample, the heavy minerals are primarily finer grained than the average grain size of the whole sample.

X-ray diffraction is a unique analytical tool for the analysis of geological samples. It is the most important and informative technique for identification and characterization of minerals in the sediments. The XRD spectra were recorded for all the sediment samples to determine the mineralogical composition. A representative diffractogram is shown in Figure 4 along with name of the minerals present.

The XRD pattern (Figure 4) indicates that the compositions are of single-phase structure since no ambiguous reflections other structure is evidenced. Analyzing the XRD patterns we notice that the position of the peaks comply with the reported value. All diffraction peaks of the studied samples are compared to the reported structure for relevant base sand samples. The observed peaks at 2 theta value of 18.37, 21.24, 23.86, 26.97, 29.01, 30.10, 35.42, 36.95, 38.43, 39.62, 40.64, 44.24, 46.10, 47.81, 50.53, 53.43, 55.30, 60.28, 63.68, 68.55 confirmed that the sand is mainly containing Quartz, Monazite, Feldspar, Uranium Fluoride, Rutile, Zircon, Hematite, Kayanite, Uranium Arsenide. Observed peak positions were matched against the International Centre for Diffraction Data (ICDD) Joint Committee on Powder Diffraction System (JCPDS) card database.

The Brahmaputra drains the Tibetan Plateau of China and is dominated by upland tributaries originating in the Himalayas. The Brahmaputra flows through various rock types including Precambrian metamorphics (high-grade schists, gneisses, quartzites, metamorphosed limestones), felsic intrusives, and Paleozoic-Mesozoic sandstones, shales and limestones [12] . From the heavy mineral assemblage study of the Brahmaputra-Jamuna River sand, it can be postulated that there are several types of source rock complexes are exist in the specific source area

mainly Lesser Himalayan and Indo-Burman Ranges with having little influence of Himalayan tributaries and north Himalayan [13] . Heavy minerals obtained in the study might have the similar source as described by the researchers.