A novel approach for reduction roasting of iron ore slime using cow dung

    A novel approach for reduction roasting of iron ore slime using cow dung Swagat S. Rath, Danda S. Rao, Barada K. Mishra PII: DOI: Reference:

S0301-7516(16)30248-4 doi:10.1016/j.minpro.2016.11.015 MINPRO 2987

To appear in:

International Journal of Mineral Processing

Received date: Revised date: Accepted date:

25 April 2016 12 October 2016 14 November 2016

Please cite this article as: Rath, Swagat S., Rao, Danda S., Mishra, Barada K., A novel approach for reduction roasting of iron ore slime using cow dung, International Journal of Mineral Processing (2016), doi:10.1016/j.minpro.2016.11.015

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT A Novel Approach for Reduction Roasting of Iron Ore Slime using Cow Dung

IP

T

Swagat S. Rath*, Danda S. Rao, Barada K. Mishra CSIR-Institute of Minerals and Materials Technology, Bhubaneswar-751013, India

SC R

Abstract

The present study encompasses the first ever attempt of the usage of cow dung as a reductant in the reduction roasting of an Indian iron ore slime containing 56.2% Fe.

NU

The resultant reduced mass generated a concentrate of ~64% Fe with a weight recovery ~66% after being subjected to low intensity magnetic separation (LIMS).

MA

The optimum conditions of roasting as determined by the Taguchi statistical design were found to be temperature: 700º C, time: 45 min and reductant to feed ratio: 0.25:1. Under similar conditions, the conventional reductant charcoal (fixed carbon:

D

93.5%, volatile matter: 1.2%) could result in a product of ~66% Fe at ~35% weight

TE

recovery proving cow dung (fixed carbon: 8.66%, volatile matter: 41.27%) to be a better reductant. The cow dung cake and some of the reduced products were

CE P

subjected to integrated instrumental techniques such as X-ray Diffraction (XRD), Scanning Electron Microscope attached with an Energy Dispersive Spectroscopy (SEM-EDS) and Electron Probe Micro-Analysis (EPMA), which revealed the

AC

formation of new phases such as hercynite, corundum, iron aluminium silicate along with magnetite and wustite, at different levels of operating parameters, thereby explaining the role of different reduction variables. Key words: Iron ore slime, reduction roasting, cow dung, low intensity magnetic separation, EPMA

Corresponding author email id: [email protected]; Ph. No. +91-674-2379147, Fax No. +91-674-2567160

ACCEPTED MANUSCRIPT 1. Introduction The exploration, production and utilization of renewable energy resources for various applications are necessitated by the synergistic effect of alarming climatic

T

changes and gradual depletion of various fossil fuels. In this context, the use of

IP

biomass such as agricultural waste, municipal waste, plant material, sewage or food

SC R

waste is of global interest considering they are renewable, freely available and they put forth a very small carbon foot print in many cases. Cow dung, the undigested residue of plant matter passed through the cattle’s gut is one such biomass. On an

NU

average, a mature well-fed cow can produce 10-15 Kg of cow dung every day that contains around 28% water in its fresh state and 34% ash when calcined [Olusegun

MA

and Sam, 2012]. The world cattle population is estimated at 964.6 million in 2015. India has the highest cattle inventory followed by Brazil and China [World cattle inventory, 2015]. India alone produces over 83 MTPA of dry dung-cake, which is

D

mostly used by the rural households as a domestic fuel for cooking and warming

TE

purposes [Dikshit and Birthal, 2010]. Traditionally used as a fertilizer, fuel, insect repellent and thermal insulator in the rural parts of India, cow dung found its major

CE P

application in the energy sector in the form of biogas [Singh, 1973]. The generation of energy from cow dung is a topic of general interest to the global scientific community [Song et al., 2012; Corro et al., 2013; Jaiganesh et al. 2013]. Cow dung

AC

slurry is composed of 1.8-2.4% N2, 1.0-1.2% P2O5, 0.6-0.8% potassium, and 50-75% organic humus. During anaerobic fermentation, the biogas generated from cow dung contains 55-65% methane, 30-35% carbon dioxide, with some hydrogen, nitrogen and other traces. Its calorific value is about 600 B.T.U.'s per cubic foot. However, the aerobic fermentation produces mostly carbon dioxide and ammonia along with a huge amount of heat [Milono et al., 1981; Centre for Application of Renewable Energy, 2013]. Other than this, the modern-day researchers are also exploring the applications of cow dung in areas like specialty chemicals [Vijayaraghavan et al., 2016; Vijayaraghavan et al., 2015], solid waste disposal [Rahman et al., 2014] and biological strains [Adebusoye et al., 2015; Kanso et al., 2014].

2

ACCEPTED MANUSCRIPT The growing demand for iron and steel in tandem with the gradual depletion of high grade ores calls for the utilization of the low grade resources. Iron ore slimes are one of such low grade iron resources. The fine sized slimes are the rejected stream of the

T

washing and sizing of the iron ore, which account for around 15-25% of the mined

IP

ore and are discarded into the tailing ponds at the rate of 10 MTPA. The utilization

SC R

of these slimes is still a formidable task considering their low iron content (50-58%) [Mishra et al., 2007]. As a result, they are stockpiled consuming huge land space and creating environmental hazards. Accidents related to the tailing dam-breaks have

NU

been reported in the literature [Chao et al., 2010]. Moreover, these slime tailings contain heavy metals like lead and cadmium that contaminate the surrounding areas [Ghose and Sen, 2001; Mohapatra et al., 2009]. Consequently, these heavy metal ions

MA

enter the food chain through continuous accumulation in the bio-organisms. Considering the ever increasing demand of iron and steel, the mining and

D

processing of iron ore will keep on increasing thereby creating a huge volume of

TE

slimes. Thus, it is the need of the hour to develop proper utilization strategies for these slimes. Over and above, coking coals are one of the most important raw

CE P

materials for the reduction of iron ores. So far as the Indian scenario is concerned, the country is mostly dependent on import for fulfilling its demand of coking coal.

concern.

AC

Therefore, the search for cheap alternatives for coal based reductants is also of prime

Reduction roasting is now a subject of interest in treating many low grade ores [Yu et al., 2014; Li et al., 2013; Jiang et al., 2013, Zhang et al., 2012]. During the process of reduction roasting, the hematite and goethite phases are reduced to magnetite, which is easily separated by the low intensity magnetic separation unit. This process is best suitable when the ore doesn’t respond to physical beneficiation techniques. The present study is focused on looking at the possibilities of using cow dung as a reductant for the reduction roasting studies of an Indian iron ore slime sample. Though there have been reports of biomass being used as reductant in iron making [Ooi et al., 2008; Honaker et al., 2004; Matsuda et al., 2006; Strerzov, 2006], to the best of the authors’ knowledge, the beneficiation of any low-grade iron resources has not been carried out utilizing cow dung as the reductant. The iron ore slime sample 3

ACCEPTED MANUSCRIPT collected from the beneficiation plant of the Barsua iron ore mines has been considered for this work. Several physical beneficiation studies to upgrade the Barusa slimes have been reported. However, only the latest ones are being discussed

T

here. Hydrocyclone studies were carried out by Mohanty and Das [2010] where a

IP

maximum iron grade of 65% Fe at 60% recovery was predicted using the empirical

SC R

models developed by factorial design. Very recently, a flow sheet, comprising of screening, hydrocyclones and two stage magnetic separations was developed to produce an iron concentrate containing 63% Fe at 71% weight recovery from the

NU

slime [Jena et al., 2015]. Singh et al. [2015] obtained a concentrate of 62.6% Fe at 72% recovery using the colloidal magnetite assisted magnetic separation.

MA

The key motivation behind the present work is to propose a simple flow sheet involving reduction roasting and magnetic separation while using a biomass

TE

2. Materials and methods

D

resource that is abundantly available in India.

2.1 Materials

CE P

Iron ore slime with 56.2% Fe was collected from the existing slime pond of the Barsua mines. Around 25% by weight of the sample had a size of +500 µm while 46% of the samples reported below 45 µm. Cow dung cakes used as the reductant in this

AC

study were collected from the nearby rural areas. They had a diameter of ~150 mm and were powdered to below 1 mm and used in roasting. 2.2 Reduction roasting The reduction roasting experiments were carried out in a laboratory muffle furnace. A batch of 400 g of the iron ore slime sample was distributed evenly in four crucibles; a desired amount of the cow dung powder was added and mixed thoroughly with the sample. The samples were kept inside the furnace after setting the desired temperature and residence time. On completion of the experiment, the roasted masses were taken out and immediately water quenched. The cooled and dried mass was pulverized to -150 µm particle size and subjected to LIMS having a magnetic intensity of ~1800 Gauss. The magnetic and nonmagnetic fraction obtained 4

ACCEPTED MANUSCRIPT from the LIMS were dried and analyzed for the total iron content. The magnetic fraction was considered as the product and the percentage weight recovery of the product was calculated based on the weight of the initial feed slime sample (400g).

T

For comparison purposes, some reduction roasting experiments were carried out

IP

with activated charcoal (3.0% moisture, 2.5% ash, 1.0% volatile matter and 93.5%

separation was adopted for these experiments. 2.3 Characterization techniques

SC R

fixed carbon) as the reductant. The same procedure of roasting and magnetic

NU

Characterization studies were undertaken for the reductant and some of the reduced products. XRD was carried out using a Philips X-ray diffractometer with Cu-Kα

MA

radiation (PANalytical, X’pert) operated at 40 kV and 30 mA. A Zeiss SEM equipped with a Bruker EDS detector was used for the SEM-EDS studies. EPMA was done

D

using the Jeol JXA 8200 model. The reflected light microscopic and stereo

TE

microscopic studies were carried out using Leitz instruments. The Thermogravimetric (TG) and Differential Scanning Calorimetry (DSC) studies were carried

CE P

out using the Netzsch Sta 449 C equipment. The slime samples were cold mounted in resin with a hardener and then polished for SEM-EDS and EPMA studies while powder samples were subjected to XRD analysis. The polished samples were carbon

AC

coated for SEM and EPMA studies. The polished sections of the lumpy pieces of the cow dung cakes were characterized using reflected light microscopy, whereas the stereomicroscopic studies were conducted after grinding them to below 500 µm particle size. 2.4 Statistical design With no prior information available on the parametric levels for the reduction roasting experiments using cow dung as a reducant, the Taguchi statistical design [Taguchi and Konishi, 1987] was adopted in this work. The experiments in this method are designed in the form of orthogonal arrays, and the whole parameter space is investigated by conducting less number of experiments. This method deals with a systematic approach to optimize the process performance with proper screening of factors and selection of the optimum levels. The reduction roasting 5

ACCEPTED MANUSCRIPT experiments in this work were conducted using the three level L9 design where the reductant to feed ratio, roasting temperature and roasting time were considered as the factors, whereas the corresponding iron grade and weight recovery of the

T

magnetic product obtained from LIMS were treated as the response variables. The

SC R

factors with their selected levels are given in Table 1.

IP

statistical studies were carried out using MINITAB 15.0 software. The experimental 3. Results 3.1 Characterization of the reductant

NU

The cow dung cakes in the rural parts of India are generally prepared in a very crude way where the discharged dung falls on the ground and gets added up with soil,

MA

sand and rice straw (used as cow’s feed). Therefore, the cow dung cakes used in this work were characterized under the microscope and depicted in Fig. 1. Under reflected light microscope, the silicates (quartz) appear as granular white masses

D

(Fig. 1 (a & b)) while the black coloured patches correspond to the carbonaceous

TE

undigested humus materials. The stereo-microscopic studies reveal the presence of straw along with fine coagulated carbonaceous cow dung materials (Fig. 1 (c & d)).

CE P

These microscopic studies clearly indicate the presence of higher amount of silica bearing minerals in the cow dung cakes.

AC

The cow dung cake as well the cow dung ash (prepared by calcining it at 850ºC for 1 hr) were subjected to XRD analysis and presented in Fig. 2. The XRD pattern of cow dung reveals one amorphous phase (bumpy peak) accounting for the non-crystalline carbonaceous humus materials while quartz and orthoclase show up as the crystalline mineral phases. The presence of quartz and orthoclase phases in the cow dung cake could be due to the association of siliceous materials like sand during the discharge of cow dung on the earth. However, the XRD pattern of the cow dung ash reveals only quartz and orthoclase without any amorphous bumpy peaks indicating the fact that the carbonaceous materials have burnt off during the rise in temperature. The proximate and ultimate analyses of the cow dung cakes as given in Table 2 show a high ash percentage (41%) confirming the microscopic findings. The

6

ACCEPTED MANUSCRIPT calorific value of the cow dung cake as determined by a Bomb calorimeter was found to be 2837 cal/g.

T

3.2 Mineralogical analysis of the slime sample

IP

The Barsua slime sample contained hematite: 47.89%, and goethite: 23.99%. The gangue minerals associated with the sample were impure goethite, gibbsite, quartz

SC R

and kaolinite. Corundum was observed occasionally. The hematite phase was always found to be contaminated with some amounts of Al2O3, SiO2, and P2O5 (ranging from 0% to 0.93%). The alumina content in goethite grains varied from

NU

12.88- 27.34%; silica content from 6.86% to 13.53% and they contain some amount of P2O5 (ranging from 0.09% to 0.24%) as well. The limonite mineral had a high content

MA

of SiO2 (11.2% to 25.46%) and Al2O3 (22.59% to 27.55%) in comparison to goethite and hematite. A detailed mineralogical study of the sample is available in our

D

previous work [Jena et al., 2015].

TE

3.3 Reduction roasting and magnetic separation

CE P

3.3.1 Statistical modeling

The Taguchi design makes use of the signal to noise (

) ratio approach in order to

estimate the variation of the performance of a response parameter. Here, the signal

AC

represents the desirable value of the output, and the word noise refers to the undesirable value of the output. The

ratio is used to quantify the quality

characteristic deviating from the desired value. The

ratio characteristics can be

broadly divided into three categories: the larger-the-better, the smaller-the-better and the nominal-the-better. Allowing for the maximization the iron grade and weight recovery resulting out of the process, the larger-the-better characteristic was employed in this work. The

where

ratio is quoted in dBi units and can be written as

is the mean-square deviation for the response variable. The

for

the larger-the-better quality characteristic can be defined as 7

ACCEPTED MANUSCRIPT

where

is the value of the quality characteristic determined by a trial; and n and N

IP

T

correspond to the test number and the total number of data points respectively. The optimum combination of the experimental factors is determined using the ratio in the orthogonal array.

SC R

parametric levels of the highest

The experimental results in terms of the Fe percentage and weight percentage of the

NU

magnetic products as obtained from the LIMS of the roasted products for the different levels of roasting parameters as per the L9 design are given in Table 3.

MA

For a better understanding of the regions where maximum Fe and weight recovery are obtainable, contour diagrams were drawn and displayed in Fig. 3. It is observed that the highest grade of Fe can be obtained at a temperature of ~ 900ºC and time 45-

D

70 min (Fig. 3(a)). However, the region represented by a temperature of ~700ºC and

TE

time 55-70 min gives the best weight recovery (Fig. 3(b)). Similarly, Fig. 3 (c) indicates that the combination of high reductant to feed ratio (~0.75) and high

CE P

temperature (~ 900ºC) favors a high grade of Fe though the highest weight recovery is attainable at a temperature of 700-750 ºC irrespective of the reductant to feed ratio

AC

(Fig. 3 (d)).

Though the contour plots gave an indication of the preferable regions of operation to obtain the best product, the effects of individual factors were assessed based on the ratios obtained for each experiment. The marginal means of the

ratios of

each factor were plotted against the factor at different levels, and shown in Fig. 4. The corresponding delta values (the difference between the highest and lowest ratio) of the factors and the ranks assigned accordingly are shown in the figure. These ranks are an indication of the relative significance of the factors in the process. It is evident that the

ratio response decreases significantly up to two levels with

the increasing reductant to feed ratio and up to three levels with the increasing temperature. On the contrary, it increases with an increase in the reduction time. It is observed that, temperature having the maximum delta value (7.84) plays the most 8

ACCEPTED MANUSCRIPT important role followed by the reductant to feed ratio (delta value: 4) and time (delta value 3.05). The overall best experimental conditions as per the design are reductant to feed ratio: 0.25, temperature: 700ºC and time: 90 min.

T

As revealed by the TGA analysis (discussed later in section 4), most of the volatile

IP

matter in the cow dung burns off by the time it reaches 700 ºC. Therefore, some more

SC R

roasting experiments were carried out in order to study the reduction behaviour at the temperatures of 600 and 700ºC. The roasting time was varied while the reductant to feed ratio was kept fixed at 0.25. The results of the reduction roasting and

NU

magnetic separation experiments as displayed in Fig. 5 indicate that, at a temperature of 700ºC and time of 45 min, it is possible to obtain an iron ore

MA

concentrate of 64.8% Fe at 66.2% weight recovery. At a higher time of 90 min, the Fe grade enhances up to 66.5% though the weight recovery decreases to 53%. The results obtained at the roasting temperature of 600 ºC are, however, not impressive.

D

Though some reduction occurs at this temperature, the Fe grade is limited to 59-60%,

TE

which is an indication of the fact that the temperature of 700ºC is suitable to obtain a

CE P

good iron ore concentrate.

3.3.2 Comparison with activated charcoal Further, reduction roasting and magnetic separation experiments were carried out to

AC

compare the performance of cow dung cake with a high purity activated charcoal. The temperature and time were fixed at 700ºC and 45 min while the reductant to feed ratio was varied between 0.05-0.25. The results as displayed in Fig. 6 suggest that the weight recoveries of the iron ore concentrates obtained using cow dung as the reductant are better compared to that using activated charcoal. At the highest reductant to feed ratio of 0.25, though the grade in case of the charcoal aided roasting is higher (66.2% Fe) compared to the one using cow dung (64.2% Fe), the corresponding weight recovery of 35% is very low when compared to the weight recovery (66.2%) of the product obtained using cow dung as the reductant. At a reductant to feed ratio of 0.1, it is possible to obtain a concentrate of 63.1% Fe at 48% weight recovery in case of cow dung while the corresponding grade and weight recovery are 63.9 and 18.7% respectively with charcoal as the reductant. 9

ACCEPTED MANUSCRIPT This gives a clear indication that, the cow dung cake used in this work performs better than an established reductant like activated charcoal under the similar reduction roasting conditions. Our earlier work [Rath et al., 2014] involving

T

activated charcoal as a reductant in the reduction roasting and magnetic separation

IP

studies of an iron ore with 51.6% Fe reported the best product at a temperature of

SC R

950ºC, which further supports that cow dung is a better reductant at 700 ºC. 3.4. Mineralogical studies of the reduced product

NU

3.4.1 SEM-EDS studies

The morphological, as well as the mineralogical changes in the roasted product were

MA

analyzed using SEM. Three experimental conditions specified as experiment no. 1, 5 and 9 in Table 3 were considered. Various phases in these figures were identified, marked with numbers and were subjected to analyses by EDS attached to the SEM.

D

The details of these phases (Figs. 7-9) along with their semi-quantitative mineral

TE

chemistry are reported in Table 4.

The back scattered electron (BSE) images obtained by SEM show spongy/porous

CE P

textures (Fig. 7) and cracked/fractured mineral grains (Figs. 7-9) predominantly occurring in the iron rich regions, whereas the ring structures show the hematite phase encircling quartz and corundum. These features are an indication of the

AC

reduction of goethite and hematite to magnetite. Fig. 7 (a) shows the development of magnetite even at the lowest temperature adopted (700ºC) in this work. The newly formed magnetite phase not only shows wide particle size distribution but also shows wide morphological varieties. Other than magnetite, two new phases like corundum (Fig. 7) and hercynite (Fig. 9) are observed in the reduced samples. The development of hercynite is due to the reaction of kaolinite and iron phases during the process of reduction while the formation of corundum may be attributed to the phase conversion of kaolinite and gibbsite, which were present in the unreduced feed sample. The corundum as well as hercynite phases, thus developed in these reduced samples occur as independently isolated phases within magnetite. 10

ACCEPTED MANUSCRIPT At places, these corundum, as well as hercynite phases, are encircled

by hematite

and or magnetite phase at its borders. The iron-aluminium silicate (IAS) as observed in Fig.7 and Fig.9, possibly almandine-pyrope solid solution series of garnet, has

T

developed by the reaction of silicates, alumina and iron phases. These iron-

IP

aluminium silicate phases show wide compositional variation with respect to iron,

SC R

aluminium and silicon contents (Table 4). 3.4.2 EPMA studies

The characterization of the reduced samples was also carried out by EPMA in order

NU

to gain further knowledge on the impacts of the roasting products on downstream processes such as LIMS, and to decipher the textural complexity of the phases in the

MA

products, which could not be revealed by SEM-EDS. Two experimental conditions specified as experiment no. 5 and 9 in Table 3 were considered for the elemental The EPMA of the reduced product obtained from

D

mapping under EPMA.

TE

experiment no. 5 as given in Fig. 10 shows traces of magnesium and that too only in the iron bearing phases, which is likely to have come from the bio-mass. The

CE P

corundum phase shows wide particle size variation and complicated association as inclusions within the magnetite phases (Fig.10). The corundum phase does not show any sharp boundary line indicating that the phase has developed during the

AC

reduction process. The BSE image along with the mapping indicates that the fine grained corundum phase may report in the magnetic fraction as unliberated particles, whereas the coarser ones would report to the non-magnetic fraction because of their existence as liberated particles. The EPMA of the reduced product obtained from experiment no. 9 is depicted in Fig. 11 and 12. The elemental mapping in Fig. 11 indicates the presence of extreme fine sizes of medium to coarse grains of hercynite (10-70 µm) and magnetite in the size range of 5-70 µm. The wustite phase (Fig. 11) is seen encircling the aluminium silicate phase. This confers that the extremely fine sized silica and corundum are not liberated, hence, would be locked within the magnetic fractions of the LIMS process. This may result in a low weight recovery in the magnetic product. Fig. 12 indicates the presence of the crystalline silica phase having perfect boundaries along with 11

ACCEPTED MANUSCRIPT rounded/globular silica bearing phases. The crystalline phases might be from the iron ore slime while the biomass might have contributed the globular silica phases. The

other

unidentified

phase

in

the

biomass

could

be

the

unburnt

T

carbon/carbonaceous material which could not be mapped as the samples were

IP

coated with carbon. Traces of iron reported in this figure could be from the slimes.

SC R

3.4.3 XRD studies

The magnetic and non-magnetic products as obtained from the LIMS, from the exp. no. 5 and 9 were subjected to XRD studies in order to confirm the newly developed

NU

phases like hercynite and corundum. The XRD pattern of the magnetic products as shown in Fig. 13 indicates the presence of magnetite and hematite as the major

MA

constituents. Quartz even though being nonmagnetic in nature, reports in the magnetic fraction, which is due to the locking and non-liberation from iron phases. However, the major amount of quartz is reported in the non- magnetic fraction. This

D

is revealed from the highest intensity peak of magnetite in the magnetic fraction and

TE

quartz in the non-magnetic counterpart. Corundum and hercynite, which are

CE P

developed as the new phases during the course of this reduction process report only in the non-magnetic fraction. The presence of wustite is only observed in the nonmagnetic fraction of experiment no. 9, which indicates either the temperature (900ºC)

AC

or the reductant to feed ratio (0.75) is very high facilitating the further conversion of magnetite to wustite. This explains why the Fe grade in experiment no. 9 is as high as 71% accompanied by a very poor weight recovery of 14.3%. The corundum phase has developed as a result of the conversion from kaolinite and or gibbsite at high temperature. Hercynite resulted from the reaction of alumina phases with iron at high temperatures. The characterization studies suggest that the reduction at 700º C accompanied by less reductant to feed ratio and roasting time produces mostly magnetite in the magnetic product and quartz in the non-magnetic product. On the contrary, a good portion of the iron phase gets converted to wustite and hercynite leading to a poor weight recovery in the magnetic fraction for the experiments no. 5 and 9 that involve higher roasting temperature, time and reductant to feed ratio. 12

ACCEPTED MANUSCRIPT 4. Discussion As evident from the results, cow dung is found to be a better reductant than activated charcoal. The presence of volatile matter to the tune of 41.27% in cow dung

T

testifies that it contains significant amount of short and long chain hydrocarbons,

IP

including aromatic compounds [Kiyasudeen et al., 2015]. The combustion of the

SC R

volatile matter raises the actual reaction temperature much above the set furnace temperature. Moreover, the incomplete combustion of the hydrocarbons generally leads to the formation of reducing gases like CO and H2. These facts explain why

NU

cow dung is able to reduce the iron oxides to the desired level and produce an optimum magnetic product even at a furnace temperature of 700°C. On the contrary,

MA

activated charcoal needs a higher temperature for displaying similar reducing behavior as it has very little VM. Considering the probability of solid state reduction of iron oxides at such a low temperature is insignificant, a good magnetic product is

D

not obtainable at a temperature of 700°C while using charcoal as a reductant. The

TE

comparative TG and DSC plots of charcoal and cow dung as shown in Fig. 14 further substantiate the above statements. The TG analysis (Fig. 14 a) of cow dung cake

CE P

reveals a mass loss of 8% till 250ºC; 40% till 500ºC and 47% till 1100ºC. It is noteworthy that the maximum amount of volatile matter loss is reported below 700 ºC. In contrast, the mass loss in the case of activated charcoal is not more than 20-

AC

22% even at the highest temperature. The DSC plots also support the TG data indicating more heat generation in the case of cow dung compared to charcoal especially in the temperature range of 600-900ºC. 5. Conclusions The present study confirms the application of cow dung cake as a reductant of iron ore slime. The slime sample with 56.2% Fe was roasted at different temperatures, reductant to feed ratios and time periods; followed by LIMS to obtain iron rich magnetic concentrates. The statistically designed experiments followed by some additional experiments suggested that it is possible to obtain an iron ore concentrate of 64.3% Fe at a weight recovery of 66.2% with roasting temperature: 700ºC, reductant to feed ratio: 0.25 and reduction time: 45 min. Under similar conditions, 13

ACCEPTED MANUSCRIPT the use of activated charcoal as a reductant ended up with poor weight recoveries. The generation of heat from the organic volatile matter in the cow dung cake, which is as high as 41.27% can be held responsible for the process of reduction at a furnace

T

temperature of 700ºC, which is an advantage over activated charcoal where the

IP

complete burning of the fixed solid carbon needs higher temperature and time

SC R

duration. The possible generation of reducing gases such as CO and H2 from the combustion of the hydrocarbon content of cow dung has a key role to play in the reduction process.

NU

The extensive characterization studies of the roasted product by SEM-EDS, EPMA and XRD indicated that at 700ºC, magnetite is a predominant product, which

MA

explains why iron rich magnetic products were obtained from LIMS at this temperature. As the temperature increased to 800ºC and 900ºC accompanied by a higher reductant to feed ratio and a higher roasting time, the weight recovery of the

D

magnetic concentrate decreased. This is due to the formation of hercynite and iron

TE

alumino silicate at both 800ºC and 900ºC, and wustite at 900ºC. Both these products being non-magnetic in nature report in the non-magnetic fraction leading to a poor

CE P

weight recovery in the magnetic portion during the LIMS. The present process for utilization of the Barsua iron ore slime using cow dung

AC

enabled reduction roasting followed by magnetic separation is superior in comparison to the processes reported earlier [Jena et al., 2015, Singh et al., 2015]. The process using a complex flow sheet comprising of multistage hydrocyclone and magnetic separation [Jena et al., 2015] gave rise to a product having 63% Fe with 70.7% weight recovery while the magnetic separation using selective colloidal magnetite coating could enhance the slime to 62.6% Fe with 72% iron recovery [Singh et al., 2015]. Moreover, the present study involves grinding the reduced product to below 150µm while the earlier workers ground the sample to below 75µm, which is a more expensive affair. Above all, the present process uses an environment-friendly, renewable, abundantly available and economically viable reductant that is suitable for the reduction roasting of complex and low grade iron ores. With high ash and high volatile 14

ACCEPTED MANUSCRIPT content, cow dung is not supposed to replace coke as a reductant. However, it can be beneficial for a typical application like reduction roasting followed by magnetic separation, where the ash content gets separated from the product during magnetic

T

separation. The present work has generated the first hand information on cow dung

IP

as a reductant and the studies on several aspects related to kinetics, thermodynamics

SC R

and issues like smoke, shoot and tar formation is open for further investigations. Acknowledgements

The authors are thankful to the central characterization cell of CSIR-IMMT for their

NU

support in XRD and EPMA studies. The authors also thank Govt. of India for their

MA

funding.

References

D

Adebusoye, S. A., Okpalanozie, O., Nweke, N. C., 2015. Isolation and

TE

characterization of bacterial strains with pyrene metabolic functions from cow

CE P

dung and Terminalia catappa phylloplane. Biocatalysis and Agricultural Biotechnology. 4(4), 778-783. Chao, L., Henghu, S., Jing, B., Longtu, L., 2010. The recovery of iron from iron ore

AC

tailings using magnetic separation after magnetizing roasting, Innovative methodology for comprehensive utilization of iron ore tailings: part 1. J. Hazard. Mater. 174 71–77. Corro, G., Paniagua, L., Pal, U., Bañuelos, F., Rosas, M., 2013. Generation of biogas from coffee-pulp and cow-dung co-digestion: Infrared studies of post combustion emissions. Energy Conversion and Management. 74, 471-481. Data by Centre for Application of Renewable Energy, http://www.careindia.com/gobar-gas.html, 2013 (accessed on 09.02.2016).

15

ACCEPTED MANUSCRIPT Dikshit, A. K., Birthal, P. S., 2010. Environmental value of dung in mixed croplivestock systems, Indian Journal of Animal Sciences. 80 (7), 679–682.

T

Ghose, M.K., Sen, P.K., 2001. Characteristics of iron ore tailing slime in India and its

SC R

IP

test for required pond size. Environ. Monit. Assess. 68 (1), 51–61. Honaker, R.Q., Taulbee, D., Parekh, B.K., Tao, D., Patil, D.P. 2004. Premium fuel production from coal and timber waste. Miner. Metall. Proc. 21, 183–188.

NU

Jaiganesh, V., Nagarajan, P.K., Geetha, A., 2014. Solid state bio methane production

MA

from vegetable wastes Current state and perception, Renewable and Sustainable Energy Reviews. 40, 432-437.

D

Jena, S. K., Sahoo, H., Rath, S. S., Rao, D. S., Das, S. K., Das, B., 2015. Characterization

TE

and Processing of Iron Ore Slimes for Recovery of Iron Values. Mineral

174-182.

CE P

Processing and Extractive Metallurgy Review: An International Journal, 36(3),

AC

Jiang, M., Sun, T., Liu, Z., Kou, J., Liu, N., Zhang, S., 2013. Mechanism of sodium sulfate in promoting selective reduction of nickel laterite ore during reduction roasting process. Int. J. Miner. Process. 123, 32-38. Kanso, S., Dasri, K., Tingthong, S.A., Watanapokasin, R. Y., 2011. Diversity of cultivable hydrogen-producing bacteria isolated from agricultural soils, waste water sludge and cow dung. International Journal of Hydrogen Energy. 36 (14), 8735-8742.

16

ACCEPTED MANUSCRIPT Kiyasudeen, S. K., Ibrahim, M.H., Ismail, S.A. 2015. Characterization of fresh cattle wastes using proximate, microbial and spectroscopic principles. American-

T

Eurasian J. Agric. & Environ. Sci. 15 (8), 1700-1709.

IP

Li, G., Zhang, S., Rao, M., Zhang, Y., Jiang, T., 2013. Effects of sodium salts on

ore. Int. J. Miner. Process. 124, 26-34.

SC R

reduction roasting and Fe–P separation of high-phosphorus oolitic hematite

NU

Matsuda, T., Takekawa, M., Hasegawa, M., Ikemura, Y., Wakimoto, K., Ariyama, T.,

MA

Iwase, M. 2006. Utilization of waste wood for production of iron, carbon monoxide and dhydrogen without generating carbon dioxide, Process

D

Metallurgy—Ironmaking, Steel Res. Int. 77, 774–784.

TE

Milono, P., Lindajati, T., Aman, S., 1981. Biogas production from agricultural

CE P

residues, The First ASEAN Seminar-Workshop on Biogas Technology, Working Group on Food Waste Materials, pp. 52–65.

AC

Mishra, B. K., Reddy, P. S. R., Das, B., Biswal, S. K., Prakash, S., Das, S. K., 2007. Issues relating to characterization and beneficiation of low grade iron ore fines, Steel World, 34–40. modelling, Renewable energy 31, 1892–1905. Mohanty, S., Das, B., 2010. Optimization studies of hydrocyclone for beneficiation of iron ore slimes. Mineral Processing and Extractive Metallurgy Review. 31(2), 86-96.

17

ACCEPTED MANUSCRIPT Mohapatra, M., Rout, K., Mohapatra, B.K., Anand, S., 2009. Sorption behavior of Pb(II) and Cd(II) on iron ore slime and characterization of metal ion loaded

T

sorbent, J. Hazard. Mater. 166 (2–3), 1506-1513.

IP

Olusegun, A.A., Sam, S.A., 2012. Methanol from Cowdung. Journal of Environment

SC R

and Earth Sciences. 2 (7), 9-16.

Ooi, T.C., Aries, E., Ewan, B.C.R., Thompson, D., Anderson, D.R., Fisher, R., Fray, T.,

NU

Tognarelli, D. 2008. The study of sunflower seed husks as a fuel in the iron ore

MA

sintering process. Miner. Eng. 21, 167–177.

Rahman, M. A., Jalil, M.A., Ali, M. A., 2014. Transformation of arsenic in the

D

presence of cow dung and arsenic sludge disposal and management strategy in

TE

Bangladesh. Journal of Hydrology. 518, Part C, 486-492.

CE P

Rath, S. S., Sahoo, H., Dhawan, N., Rao, D. S., Das, B., Mishra, B. K., 2014. Optimal Recovery of Iron Values from a Low Grade Iron Ore using Reduction Roasting

AC

and Magnetic Separation. Separation Science and Technology. 49 (12), 19271936.

Singh, R.B., Bio-gas plant: designs with specifications, Ram Bux Singh Gobar Gas Research Station, Ajitmal, Etawah, 1973. Singh, S., Sahoo, H., Rath, S.S., Sahu, A. K., Das, B., 2015. Recovery of iron minerals from Indian iron ore slimes using colloidal magnetic coating. Powder Technology. 269, 38–45. Song, Z.X., Wang, Z. Y., Wu, L.Y., Fan, Y.T., Hou, H.W., 2012. Effect of microwave irradiation pretreatment of cow dung compost on bio-hydrogen process from 18

ACCEPTED MANUSCRIPT corn stalk by dark fermentation. International Journal of Hydrogen Energy 37(8), 6554-6561.

IP

T

Strezov, V. 2006. Iron ore reduction using sawdust: experimental analysis and kinetic

SC R

Taguchi, G., Konishi, S., 1987. Taguchi methods, orthogonal arrays and linear graphs, tools for quality American supplier institute, American Supplier Institute, pp. 8-35.

NU

Vijayaraghavan, P., Kalaiyarasi, M., Vincent, S.G.P., 2015. Cow dung is an ideal

MA

fermentation medium for amylase production in solid-state fermentation by Bacillus cereus. Journal of Genetic Engineering and Biotechnology. 13(2), 111-

D

117.

TE

Vijayaraghavan, P., Vincent, S.G.P., Dhillon, G.S., 2016. Solid-substrate bioprocessing

CE P

of cow dung for the production of carboxymethyl cellulase by Bacillus halodurans IND18. Waste Management. 48, 513-520. cattle

Inventory:

ranking

of

countries

by

Rob

Cook

AC

World

http://beef2live.com/story-world-cattle-inventory-ranking-countries-0-106905, 2015 (accessed on 09.02.2016) Yu, W., Sun, T., Cui, Q., 2014. Can sodium sulfate be used as an additive for the reduction roasting of high-phosphorus oolitic hematite ore? Int. J. Miner. Process. 133, 119-122. Zhang, Y.B., Li, G.H., Jiang, T., Guo, Y.F., Huang, Z.C. 2012. Reduction behavior of tin-bearing iron concentrate pellets using diverse coals as reducers. Int. J. Miner. Process. 110-11, 109-116. 19

ACCEPTED MANUSCRIPT Figure captions

T

Fig. 1 Photomicrographs of cow dung cake by reflected light microscopy (a, b) and stereomicroscopy (c & d)

IP

Fig. 2 XRD spectra of the cow dung cake and its ash

SC R

Fig. 3 Contour diagrams showing the variation of (a) Fe grade (b) weight recovery with respect to temperature and time, (c) Fe grade and (d) weight recovery with respect to temperature and reductant to feed ratio Fig. 4 Main effects plot for the S/N ratios of different roasting factors

NU

Fig. 5 Effect of roasting time on the Fe grade and weight recovery at a temperature: 700ºC and reductant to feed ratio: 0.25

MA

Fig. 6 Comparison of performance of cow dung cake and activated charcoal at different reductant to feed ratio with temperature: 700ºC and roasting time: 45 min

TE

D

Fig. 7 BSE images of the reduced samples of experiment number 1 (a) Porous/spongy structure (b) Hematite (2) along the corundum (1) (c) Hematite rims around the silicate grains (d) Unreacted cow dung mass at the centre (Magnetite (M) in b and c)

CE P

Fig. 8 BSE images of the reduced samples of experiment number 5 (a) Liberated quartz (3) and iron aluminium silicate (IAS) (4) and magnetite (5) (b) Development of IAS (6 and 7) and magnetite (8) (c) Quartz (9) and amorphous silica (10) and carbonaceous material (11) (d) Fibrous type of carbonaceous material (centre of the photomicrograph) along with magnetite (white particles)

AC

Fig. 9 BSE images of the reduced samples of experiment number 9 (a) Hercynite (13) along the borders of corundum (12). Hercynite containing magnetite crystals (14 & 15) (b) Cracked magnetite (17) crystals enclosing and enclosed by IAS (18) and hercynite (19) (c) Contraction cracks in magnetite (d) Unreacted cow dung mass at the centre Fig.10 BSE images of the reduced sample of experiment number 5 with elemental mapping for Fe, Al, and Mg indicating association of corundum (C) within the magnetite (M) Fig.11 BSE images of the reduced sample of experiment number 9 with elemental mapping for Fe, Al, Si, Mg and O. Aluminium silicate (AS; arrow marked), hercynite (H) and magnetite (M) are depicted. Fig.12 BSE images of the reduced sample of experiment number 9 with elemental mapping Fe, Al and Si. It shows the presence of unreacted biomass. Presence of colloidal silica, traces of iron and aluminum within the biomass 20

ACCEPTED MANUSCRIPT Fig. 13 XRD of the magnetic and non-magnetic fractions of the roasted products of exp no. 5 and 9

AC

CE P

TE

D

MA

NU

SC R

IP

T

Fig. 14 (a) TG and (b) DSC plots of cow dung and activate charcoal

21

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

Fig. 1

22

ACCEPTED MANUSCRIPT

Q Q

T

Or

Q

10

20

Or Or Or Q

30

40

Q

Cow dung

Or Q Q

Q

50

60

NU

Or Q

Q

SC R

Or Or

IP

Cow dung ash Or Or Or OrOr Or Or Q Q Q Q Q O Or

Q

70

Q

80

Angle, 2 Theta

AC

CE P

TE

D

MA

Fig. 2 XRD spectra of the cow dung cake and its ash

23

ACCEPTED MANUSCRIPT

Contour Plot of Fe, % vs Temperature, ºC, Time, min 900

Contour Plot of Weight, % vs Temperature, ºC, Time, min 900

Contour Plot of Fe, % vs Temperature, ºC, Time, min

Contour Plot of Weight, % vs Temperature, ºC, Time, min Fe, % < 62 62 - 64 64 - 66 66 - 68 850 68 - 70 > 70

(a) 700 30

800

750

50

60

T (b)

750 40

Weight, % < 20 20 - 30 30 - 40 40 - 50 50 - 60 60 - 70 > 70

Weight, % < 20 20 - 30 30 - 40 40 - 50 50 - 60 60 - 70 > 70

IP

750

800

SC R

800

Fe, % < 62 62 - 64 64 - 66 850 66 - 68 68 - 70 > 70

Temperature, ºC ºC Temperature,

850

Temperature, ºC

Temperature, ºC ºC Temperature,

900

70

80

700 30

90

40

50

60

70

80

90

Time, min Time, min min min ºC, Reductant to feed ratio Contour Plot of Weight, % vsTime, Temperature, Contour Plot of Fe, % Time, vs Temperature, ºC, Reductant to feed ratio 900

900

50

850 60

70

80

Fe, % < 62 62 - 64 64 - 66 66 - 68 68 - 70 > 70

90

Time, min

(c) 0.3

750

TE

700

800

D

Temperature, ºC

750

MA

850 800

Temperature, ºCºC Temperature,

40

Temperature, ºC ºC Temperature,

900

NU

Plot of Weight, %Fe, vs% Temperature, ºC, Reductant to feed ratio r Plot of Fe, % vs Temperature, ºC, ReductantContour to feed700 ratio 30 40 50 60 70 80 90

0.4 0.5 0.6 Reductant to feed ratio

< 62 64 66 85068 >

62 64 66 68 70 70

Time, min

750

(d) 700 0.7

0.3

CE P

0.4 0.5 0.6 Reductant to feed ratio

0.7

0.4 0.5 0.6 Reductant to feed ratio

0.7

Reductant to feed ratio 0.4

0.5

0.6

Fig.Reductant 3 to feed ratio

0.7

AC

0.3

0.3

Weight, % < 20 20 - 30 30 - 40 40 - 50 50 - 60 60 - 70 > 70

800

Reductant to feed ratio 700

Weight, % < 20 20 - 30 30 - 40 40 - 50 50 - 60 60 - 70 > 70

24

ACCEPTED MANUSCRIPT

37

T IP

33

SC R

31 29

27

Delta: 4.0 Rank: 2

Delta: 7.84 Rank: 1

NU

Means S/N ratios, dB

35

25 A1

A2

A3

B1

B2

B3

Delta: 3.05 Rank: 3

C1

C2

C3

MA

Factor levels

AC

CE P

TE

D

Fig. 4

25

68

66.5

66

66

64

T

67

62

IP

65 64.5

60

SC R

Fe, %

65.5

64 63.5 63 62

54

Weight, %

52

20

40 60 Time, min

50 80

100

MA

0

56

Fe, %

NU

62.5

58

Weight, %

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

Fig. 5

26

ACCEPTED MANUSCRIPT

67

70

66.5

60

T

66 65 64.5

40

64 63.5

30

Fe, % (with charcoal) Fe, % (with cow dung) Weight, % (with charcoal) Weight, % (with cow dung)

63

NU

62.5 62 0.05

0.1 0.15 0.2 Reductant to feed ratio

0.25

20 10 0

0.3

MA

0

Weight, %

IP

50

SC R

Fe, %

65.5

AC

CE P

TE

D

Fig. 6

27

Fig. 7

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

28

Fig. 8

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

29

Fig. 9

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

30

Fig.10

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

31

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

Fig. 11

32

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

Fig.12

33

ACCEPTED MANUSCRIPT

Q

M

Exp No. 5 - Magnetic

IP

T

Exp No. 5 - Non-Magnetic

H C

He

SC R

Q

H

H

H

M

0

10

20

Q HH H

H

HQ

30

40

50

H M

H

60

Or

Q M

Q

M

70

80

90

0

10

He

Q Q Or

20

Exp No. 9 - Magnetic Q

H Or H Q

He

30

40

Or C Q

50

C Q H H H Or Or Or Q C

60

Q

He W

HQ

10

20

30

40

D

Q Q H

H

50

60

He C

M

He

M

80

Q

Q C

M

70

90

W

H

90

0

10

20

C

30

40

C Q Q CC Q

50

60

Q C

70

80

90

Angle , 2 Theta

Angle , 2 Theta

AC

0

H

Q

CE P

M

Q H

TE

M

80

Exp No. 9 - Non-Magnetic

Q Q

70

W

MA

M

M

NU

M

Fig. 13

34

SC R

IP

T

ACCEPTED MANUSCRIPT

(a)

NU

(b)

AC

CE P

TE

D

MA

Fig. 14

35

ACCEPTED MANUSCRIPT

Table 1 Experimental factors and their levels

T

3 0.75 900 90

AC

CE P

TE

D

MA

NU

SC R

A: Reductant to feed ratio B: Temperature, ºC C: Time, min

1 0.25 700 30

Levels 2 0.5 800 60

IP

Factors

36

ACCEPTED MANUSCRIPT

Table 2 Proximate and ultimate analysis

Ultimate analysis (on air dried basis) H, % 2.68

Fixed carbon, % 8.64

N, % 1.09

S, % 0.21

AC

CE P

TE

D

MA

NU

C, % 26.86

Ash, % 41.43

IP

Volatile Matter, % 41.27

SC R

Moisture, % 8.66

T

Proximate analysis

37

ACCEPTED MANUSCRIPT

Table 3 Results of the statistically designed experiments Reductant to feed ratio

Temperature, ºC

Time, min

Weight, %

Fe, %

1 2 3 4 5 6 7 8 9

0.25 0.25 0.25 0.5 0.5 0.5 0.75 0.75 0.75

700 800 900 700 800 900 700 800 900

30 60 90 60 90 30 90 30 60

64.5 55.0 48.3 70.5 38.0 12.0 67.0 45.0 14.3

62.8 65.0 62.2 63.0 63.8 66.5 60.5 63.2 71.7

AC

CE P

TE

D

MA

NU

SC R

IP

T

S. No.

38

ACCEPTED MANUSCRIPT

IP

T

Table 4 The SEM-EDS semi-quantitative mineral chemistry of various phases marked with numbers on the photomicrographs.

Weight percentage of various elements S. No.

O

Al

Fe

Si

K

Ca

Mg

Ti

1

1

50.15

44.2

3.1

1.91

0.2

0.44

ND

1

2

31.04

1.76

66.48

0.53

ND

ND

5

3

57.45

0.36

0.36

41.73

0.1

5

4

31

11.02

51.89

5.02

5

5

27.68

0.5

71.48

5

6

49.07

21.46

5

7

31.33

5

8

5

S

Cl

ND

ND

ND

ND

Corundum

0.19

ND

ND

ND

ND

Hematite

ND

ND

ND

ND

ND

ND

Quartz

0.19

ND

0.41

0.47

ND

ND

ND

Fe-Al-Silicate

0.14

ND

ND

0.2

ND

ND

ND

ND

Magnetite

14.22

15.12

ND

ND

0.14

ND

ND

ND

ND

Fe-Al-Silicate

4.14

61.46

2.84

ND

ND

0.22

ND

ND

ND

ND

Fe-Al-Silicate

22.55

0.3

76.6

0.05

ND

ND

0.31

0.18

ND

ND

ND

Magnetite

9

58.74

ND

0.54

39.34

0.39

ND

0.99

ND

ND

ND

ND

Quartz

5 5

10 11

61.13 88.57

ND ND

3.58 1.15

30.47 8.78

1.13 0.25

1.74 0.38

0.85 0.45

ND ND

0.66 0.13

0.15 0.09

0.29 0.2

Quartz Carbonaceous material

9

12

48.5

44.55

6.26

0.36

0.11

ND

0.24

ND

ND

ND

ND

Corundum

9

13

44.29

29.61

25.81

0.07

0.03

ND

0.19

ND

ND

ND

ND

Hercynite

9

14

31.79

ND

68.08

0.04

ND

ND

0.09

ND

ND

ND

ND

Hematite

9

15

31.79

ND

68.08

0.04

ND

ND

0.9

ND

ND

ND

ND

Hematite

9

16

44.12

21.15

34.54

ND

ND

ND

0.19

ND

ND

ND

ND

Hercynite

9

17

15.3

3.68

78.7

0.11

ND

ND

0.47

1.53

0.21

ND

ND

Magnetite

9

18

49.01

43.39

6.00

1.1

0.06

ND

0.33

0.11

ND

ND

ND

Fe-Al-Silicate

9

19

36.1

28.59

30.91

3.83

0.02

ND

0.25

0.31

ND

ND

ND

Hercynite

NU

MA

D

CE P

AC

SC R

P

TE

Exp. No.

Phases inferred

39

AC

CE P

Graphical abstract

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

40

ACCEPTED MANUSCRIPT Highlights

T

IP

SC R NU MA D TE CE P

 

First ever usage of cow dung as a reductant Iron recovery from iron ore slime using reduction roasting and magnetic separation Product with ~64% Fe at a weight recovery ~66% from a slime of 56.2% Fe Formation of magnetite, wustite, hercynite and iron aluminosilicate phases

AC

 

41