collectors for low rank coal & oxidized coal

Ž .Int. J. Miner. Process. 58 2000 99–118www.elsevier.nlrlocaterijminpro

An improved class of universal collectors for theflotation of oxidized andror low-rank coal

Renhe Jia, Guy H. Harris, Douglas W. Fuerstenau )

Department of Materials Science and Mineral Engineering, EÕans Hall, UniÕersity of California, Berkeley,Berkeley, CA 94720, USA

Received 1 December 1998; received in revised form 15 April 1999; accepted 1 June 1999

Abstract

Ash minerals, including pyrite, can be separated from coal by flotation, primarily making useof the natural hydrophobicity of the carbonaceous matter in coal. However, to overcome thedeleterious effect of oxygen functional groups on the coal surface, an organic collector is required.The most common industrial coal flotation collector is fuel oil, but the addition of oxygenatedfunctional groups to the collector molecule markedly enhances the flotation of lower rank andoxidized coals. This paper summarizes the results of detailed study of the flotation response of twohigh-sulfur coals, Illinois No. 6 coal and Pittsburgh No. 8, using different non-ionic oxygenatedsurfactants as the collector. The performance of these reagents is compared with that of two oilycollectors, namely dodecane and nonylbenzene, and mechanisms for the interaction of thesecompounds with coal are suggested. q 2000 Elsevier Science B.V. All rights reserved.

Keywords: oxidized coal flotation; oxygenated collectors; non-ionic surfactants; coal collection mechanisms;coal-collector interactions

1. Introduction

Coal is a solid combustible material that results from the geologic alteration ofvegetable matter, largely in the absence of air. Depending on the oxygen content of thecoal, there is a progression in carbon content and hydrophobicity, and hence, flotabilityof various coals. The mineral matter in raw coal is comprised of hydrophilic minerals,

Ž .mainly clays such as kaolinite and montmorillonite , quartz, carbonate minerals such ascalcite and dolomite, gypsum, and pyrite.

) Corresponding author. Tel.: q1-510-642-3826; Fax: q1-510-643-5792; E-mail:[email protected]

0301-7516r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0301-7516 99 00024-1

( )R. Jia et al.r Int. J. Miner. Process. 58 2000 99–118100

Overall, although not a major aspect of coal cleaning, froth flotation is an importantmeans of upgrading the fine fraction of mined raw coal, typically particles finer than 0.5mm. In addition to recovering the fine coal, it is an effective means for reducing thedeleterious pyrite and ash-forming minerals in the coal product. Coal flotation also playsan important role in removing coal fines from cleaning plant water to be discharged andin removing fine coal particles from the discharged plant refuse. In summary, flotationrecovers the valuable constituent in coal and separates it from the worthless ash-formingminerals.

Ž .Coal flotation is typically carried out with the natural fines -0.5 mm as feed atŽ .very low solids concentration in the slurry about 4% at natural pH. In the US, around

10% of the raw coal is processed with flotation, while on a worldwide basis, from 0 toŽ .40% of the coal is treated by flotation Aplan and Arnold, 1991 . Given that the 1997

Ž .coal production in the US was 988 million tons Hong, 1998 , this means that thetonnage of raw coal cleaned by flotation was indeed large. In his recent review of coal

Ž .flotation, Aplan 1999 pointed out that in 1995 there were 345 coal preparation plantsin the US, of which 110 employed flotation for treating some of their material.

Coal flotation makes use of the natural hydrophobicity of the carbonaceous matter incoal. To enhance the hydrophobicity of the coal particles, oily materials, such as diesel

Ž .oil and kerosene, are added industrially as collectors Brown, 1962; Harris et al., 1995 .For higher-rank coals, the reagent consumption in flotation is low because of the naturalhydrophobicity of the coal. However, for low-rank coals containing greater amounts ofoxygen, oily collectors will not spread on the surface of the coal particles, which leadsto poor flotation performance and large reagent dosages requirements even to obtain

Ž .moderate recovery of coal Aplan, 1976; Gutierrez-Rodriguez and Aplan, 1984 . Fur-thermore, most coals are susceptible to oxidation by weathering, which can occur at themine site or during storage and transportation. Weathering processes result in theformation of oxygen functional groups, most commonly, carboxyl, phenolic and car-bonyl functionalities on the coal surface, which reduce the hydrophobicity of the coalsurface by increasing the number of sites that hydrogen bond with water molecules. It is

Žthis that makes the coal more difficult to float with oily collectors Laskowski andMiller, 1984; Aplan, 1988; Fuerstenau et al., 1987, 1992; Fuerstenau and Diao, 1990;

.Harris et al., 1995 . The objective of this paper is to present the results of a study inwhich oxygenated polar groups are added to the collector molecule to provide a meansfor it to hydrogen bond with the oxygenated surface sites on the coal.

Ž .Previous work had shown that tetrahydrofurfuryl butyrate designated here as THF-3Ž .should be an effective collector for bituminous coal Harris et al., 1995 . Those results

suggested that the THF series of reagents might be effective alternatives as collectors inthe flotation of both low-rank and oxidized coal. The objective of our investigation wasto determine the efficacy of the collecting ability of THF compounds having varioushydrocarbon chain configurations to function as collectors for the flotation of unoxidizedcoals, laboratory-oxidized coals, and naturally weathered coals. The investigation cen-tered on two coals: the more hydrophilic Illinois No. 6 coal and the more hydrophobicPittsburgh No. 8 coal. The results are compared with those of other representativecollectors. It is expected that the oxygenated polar group of these non-ionic surfactantswill have strong polar interactions with the surface oxygenated sites on the coal.

( )R. Jia et al.r Int. J. Miner. Process. 58 2000 99–118 101

Table 1Ž .The proximate and sulfur analysis of the two coals in percentage Fuerstenau et al., 1992

Coal As received moisture Fixed carbon Volatile matter Ash Total sulfur Pyritic sulfur

Illinois No. 6 9.5 49.2 36.2 14.6 4.61 2.36Pittsburgh No. 8 2.3 52.6 35.7 11.7 4.10 2.78

Because of this, the surfactants can spread over the coal surface more readily than oilycollectors and should not act as wetting agents.

2. Experimental materials and methods

2.1. Materials

Ž .Two bituminous coals, namely Illinois No. 6 from Peabody Coal, Marissa, IL andŽ .Pittsburgh No. 8 from R&F Coal, Warnock, OH , were used in this investigation. The

proximate and sulfur analyses of these two coals are given in Table 1. As indicated byŽ .its high moisture content Table 1 , Illinois No. 6 coal is much less hydrophobic than

Pittsburgh No. 8 coal.Table 2 summarizes the results of a variety of experimental measurements that clearly

show the differences in wettabilityrflotability of Illinois No. 6 and Pittsburgh No. 8Ž .bituminous coals. A detailed investigation Fuerstenau et al., 1992 of the flotability of

these two coals showed that to achieve similar flotation response, the dosage ofdodecane collector was 2.59 kgrton for Illinois No. 6 but only 0.86 kgrton forPittsburgh No. 8 coal. For this standard flotation procedure, the dosage of MIBC frotherwas 0.53 kgrton and 0.13 kgrton, respectively, for the two coals. Using the sessile dropmethod, airrwater contact angles of samples that had been wet-polished in air werefound to be 518 and 648 for Illinois No. 6 and Pittsburgh No. 8 coal, respectivelyŽ . ŽFuerstenau et al., 1992 . Measurement of induction times on 104 mm=148 mm 100

.mesh=150 mesh particles showed that the time to achieve air bubblercoal particlecontact was 480 ms for Illinois No. 6 coal and 180 ms for Pittsburgh No. 8 coal.Induction times were also measured on 100 mesh=150 mesh particles taken from coalsamples that had been weathered for various time periods. After 3 or 4 monthsweathering, induction times had increased to 200 ms for Illinois No. 6 coal and 40 ms

Table 2Comparison of interfacial parameters related to the wettability and flotation of Illinois No. 6 and Pittsburgh

Ž .No. 8 coal Fuerstenau et al., 1992

Ž .Coal Contact Induction Standard reagent dosage kgrton HydrophobicŽ . Ž . Ž .angle 8 time ms fraction %Collector Frother

Illinois No. 6 51 480 2.59 0.53 56Pittsburgh No. 8 64 180 0.86 0.13 85


for Pittsburgh No. 8 coal. After weathering for 1 year, the respective induction timesincreased further to 1700 ms and 450 ms because of oxidation. Film flotation testsconducted on particles that had been prepared from coal which had been weathered for

Ž0.5 month showed that 56% of the Illinois No. 6 coal particles were hydrophobic that.is, floated on water during the film flotation experiment whereas 85% of the Pittsburgh

No. 8 coal particles remained hydrophobic. All of these results clearly point to IllinoisNo. 6 coal being considerably more hydrophilic than Pittsburgh No. 8 coal.

For the present work, all lump coal samples were first crushed to y6.35 mm, thensplit into 500-g subsamples, and stored in plastic bags under inert environment toprevent further oxidation before being used for each set of flotation experiments.

2.2. Chemicals

wThe main chemicals used in this work were a series of THF esters C H O–CH –4 7 2xOOC–R . Table 3 gives our designation, the chemical formula, and the source of these

reagents. Some of these chemicals were purchased, while the rest were synthesized inour laboratory. The THF esters were prepared from equal molar amounts of THF alcoholand various carboxylic acids with a catalytic amount of toluenesulfonic acid in refluxing

Table 3Chemical composition and source of reagents used in this work

Designation Chemical formula Source

Ž .THF-1 Tetrahydrofurfuryl acetate C H O–CH - MW 144 Aldrich Chemicals4 7 2

OOC–CH 3Ž .THF-3 Tetrahydrofurfuryl butyrate C H O–CH - MW 172 CTC Organics4 7 2

OOC–C H3 7Ž .THF-7 Tetrahydrofurfuryl octonoate C H O–CH - MW 228 Made in our lab4 7 2

OOC–C H7 15Ž .THF-11 Tetrahydrofurfuryl laurate C H O–CH - MW 284 Made in our lab4 7 2

OOC–C H11 23Ž .THF-17en Tetrahydrofurfuryl oleate C H O–CH - MW 366 Pfaltz and Bauer4 7 2

OOC–C H –CH5CHC H7 14 8 17Ž .THF-P Tetrahydrofurfuryl benzoate C H O–CH - MW 206 Pfaltz and Bauer4 7 2

OOC–C H6 5Ž .THF-1-P Tetrahydrofurfuryl phenyl acetate C H O–CH - MW 220 Oxford Organics4 7 2

OOC–CH –C H2 6 5Ž .THF-2-P THF-2-phenylproprionate C H O–CH - MW 234 Made in our lab4 7 2

OOC–C H –C H2 4 6 5Ž .THF-3-P THF-3-phenylbutyrate C H O–CH - MW 246 Made in our lab4 7 2

OOC–C H –C H3 6 6 5Ž . Ž .Nonylphenol Nonylphenol CH – CH –C H –OH MW 220 Rhone-Poulenc3 2 8 6 4

Ž .GH4 Polyethoxylated nonylphenol CH – MW 341 Dow Chemical3Ž . Ž .CH –C H –O CH O H2 8 6 4 2 4

Ž . Ž .Nonylbenzene Nonylbenzene C H – CH CH MW 204 Acros Organics6 5 2 8 3Ž .Dodecane Dodecane C H MW 170 Fisher Chemical12 26

Ž . Ž .MIBC 4-methyl-2-pentanol CH CH–CH – MW 102 Kodak3 2 2

CHOH–CH 3


toluene. Refluxing was continued for a half-hour after water had ceased being collectedin the Dean–Stark trap. The toluene was washed with aqueous solutions of sodiumbicarbonate, water, and NaCl which removed the catalyst and any possible remainingstarting materials. The filtrate from the anhydrous magnesium sulfate drying agent gavethe desired product after removal of the solvent in a rotary evaporator. The yields werealmost were quantitative.

Nonylphenol was obtained from Rhone-Poulenc and polyethoxylated nonylphenolfrom Dow Chemical.

2.3. Experimental procedure

In this investigation, all flotation tests were conducted following a standard flotationŽ .procedure developed in our laboratory Fuerstenau et al., 1992 . Briefly, to start, a

sample of the 500-g y6.35 mm coal was wet-ground in a laboratory rod mill to preparea product that was 95% passing 200 mesh and then split into four parts. Flotation testswere conducted with a 2-l Denver cell using 125 g of coal at a pulp density of 6.25%solids at the natural pH of each coal. Distilled water was used in all experiments. Theimpeller speed of the flotation machine was 1200 rpm and the airflow rate was 4 lrmin.In each flotation test, the pulp was first agitated in the flotation cell for 3 min, afterwhich the collector was added and the pulp conditioned for an additional minute. MIBC

Žfrother 0.52 kgrton for Illinois No. 6 and 0.13 kgrton for Pittsburgh No. 8 coal,.respectively was then added and the pulp was conditioned for an additional 3-min

period. After collecting the flotation product for 5 min, the concentrate and tailing werefiltered, dried, and weighed. The ash content of the sample was analyzed using a LECOMAC-400 Proximate Analyzer, the results of which were used to calculate combustiblematter recovery.

Laboratory-oxidized coal was prepared by first wet-grinding the two coals toy200-mesh following our standard grinding procedures. The ground Pittsburgh No. 8coal was then oxidized in slurry form in an oven for 120 h at 608C and additionaldistilled water was added periodically to keep the sample from drying out. On the otherhand, the ground Illinois No. 6 coal was first oxidized in an oven for 90 h at 1058C andthen stored for 3 days at room temperature. After undergoing the oxidation procedure,these coal samples were then used for flotation tests with dodecane and some of THFsurfactants. Dodecane and nonylbenzene were both used for comparative purposes. Theeffect of oxidizing the surface of a coal by heating is similar to using a coal of lower

Ž .rank Aplan, 1988 .

3. Results

3.1. Fine coal flotation

The flotation response of y200-mesh Illinois No. 6 and also Pittsburgh No. 8 coalwusing dodecane, nonylbenzene and representative non-ionic surfactants THF-17en,

Ž Ž Ž . .xnonylphenol and GH4 polyethoxylated nonylphenol, C H –C H –O– C H O H9 19 6 4 2 4 4


as the collector is presented in Fig. 1. The upper part of Fig. 1 gives the combustiblematter recovery for Illinois No. 6 after 5 min of flotation as a function of reagent dosagein mol per ton of coal and the lower part of the figure presents similar plots for theflotation of Pittsburgh No. 8 coal. The changes in the heteropolar nature of the collectormolecules used for these flotation experiments can be seen in Table 4, where A denotesthe polar functional group and B the nonpolar hydrocarbon chain.

The results given in Fig. 1 indicate that THF-17en is more effective than the othercollectors for both coals. The plots also show that surfactants with an oxygenated groupare generally more effective than those without one and that nonylbenzene is moreeffective than dodecane. This indicates that there might be two mechanisms for theinteraction between the surfactants and the coal surface. As we know, the surface of coalconsists of inherently hydrophobic areas and also sites containing oxygenated moietiesw x Ž .such as carboxyl, carbonyl, phenolic and ester groups Aplan, 1988; Laskowski, 1995 .The first mechanism of interaction between the surfactants and the coal surface appears

Fig. 1. Comparison of the collecting ability of the non-ionic THF-17en surfactant with that of dodecane,nonylphenol, nonylbenzene and GH4 for the flotation of 200-mesh Illinois No. 6 coal and Pittsburgh No. 8coal.


Table 4Ž . Ž .Molecular structure of the series of coal collectors, showing the polar group A and hydrocarbon chain B

configuration

to be through the polar groups of the reagent interacting with the oxygenated functionalgroups on the coal surface by hydrogen bonding. The second mechanism involves theinteraction of the nonpolar chain with the carbonaceous sites on the coal surface bydispersing water molecules from the coal surfaces. The interaction between an aliphaticchain and the coal surface is less pronounced than that of a benzene ring with thearomatic sites on the coal surface. This is due to strong p-bonding interaction between

Žthe aromatic component of the coal matrix and the benzene ring of the reagent Mattson.and Mark, 1971; Harris et al., 1995 . Compared to dodecane, reagents containing a

benzene ring, however, still have much better collecting ability. This indicates thatalthough the benzene ring has stronger interaction with the aromatic coal surface than analiphatic chain, the interaction is not as strong as that between the polar group of thereagents and the coal surface. This also suggests that hydrogen bonding of theoxygenated groups in the reagents is stronger than the van der Waals interaction of thealiphatic chains with the carbonaceous portions of the surface. It is for this reason thatreagents containing oxygen functional groups are much better collectors than a hydrocar-bon oil, such as dodecane, for the flotation of bituminous coals.

The main part of this research was concerned with the effectiveness of the THF seriesŽ .THF esters as flotation collectors for these two coals.

w xThe flotation response of y200-mesh y74 mm Illinois No. 6 coal using dodecaneand the various non-ionic surfactants as the collector is presented in Fig. 2. The upperpart of Fig. 2 gives the combustible matter recovery after 5 min of flotation as a functionof reagent dosage in mol per ton of coal and the lower plot presents the same reagentconsumption data in terms of kg per ton. These figures clearly show that at the samereagent dosage, all the non-ionic surfactants are more effective than dodecane for theflotation of Illinois No. 6 coal. For the surfactants having an aliphatic chain, to obtaincomparable combustible matter recovery, increasing the number of carbon atoms in the


Fig. 2. Comparison of the collecting ability of the non-ionic THF surfactants with that of dodecane for theŽ . Ž .flotation of 200-mesh Illinois No. 6 coal. Reagent dosages in molrton upper figure and kgrton low figure .

alkyl chain of the THF series of reagents lowers the necessary amount of reagent, andparticularly so in comparison to dodecane. These plots also show that at the same


reagent dosage, the combustible matter recovery obtained using aliphatic THF reagentsis much higher than those reagents containing a benzene ring. Since the reagents have

Fig. 3. Comparison of the collecting ability of the non-ionic THF surfactants with that of dodecane for theŽ . Žflotation of 200-mesh Pittsburgh No. 8 coal. Reagent dosages in molrton upper figure and kgrton low

.figure .


the same polar functionality, the significant difference between these two subgroups ofreagents is the benzene ring. This certainly indicates that the benzene ring has a role inthe interaction of the reagents with the coal surface.

As already discussed, nonylbenzene is a much more effective collector than dode-cane, also suggesting that there is a stronger interaction between the benzene ring andthe aromatic sites on the coal surface than between an aliphatic hydrocarbon chain andthe coal surface. This is due to strong p-bonding between the aromatic component of the

Žcoal matrix and the benzene ring of the reagent Mattson and Mark, 1971; Harris et al.,.1995 .

Fig. 3 presents the flotation response of Pittsburgh No. 8 coal with the THF series ofnon-ionic surfactants as collectors, again including dodecane and nonylbenzene forcomparison purposes. These results also indicate that the THF series of reagents are alsomore effective collectors than dodecane for this coal. It should be noted, however, thatthe reagent dosages of the various collectors for the flotation of the more hydrophobicPittsburgh No. 8 coal are considerably lower than required for the more hydrophilicIllinois No. 6 coal. The plots also show that surfactants with an aliphatic chain again aregenerally more effective than those containing a benzene ring, as was found for theflotation of Illinois No. 6 coal. This is in consonance with there being two mechanismsfor the interaction between the surfactants and coal surfaces, as will be discussed indetail in Section 4.3.

The results given in Fig. 2 show that the most effective collector for Illinois No. 6coal is THF-17, the long-chain THF oleate. On the other hand, from the plots given in

Fig. 4. Comparison of the collecting ability of THF-17 and THF-11 with that of dodecane in the flotation of200-mesh lab-oxidized Illinois No. 6 coal.


Fig. 3, it can be seen that THF-3 is better on the high-rank Pittsburgh No. 8 coal than theŽ .remaining surfactants in terms of kilograms per ton coal although it has the second

shortest aliphatic chain. The more detailed considerations of interaction mechanismspresented in the paragraphs that follow will help explain this.

3.2. Flotation of oxidized coals

Most coals are susceptible to weathering or oxidation, which, in turn, results in theincreased formation of oxygenated functional groups on the coal surface, resulting in

Žmore hydrophilic surfaces and hence reduced flotability Swann et al., 1972; Fuerstenau.et al., 1983, 1992; Philips et al., 1987; Laskowski, 1995 . As a result of the large

number of hydrophilic oxygenated functional groups present in oxidized coal, it isexpected that the THF series might be even more effective as flotation collectors whenthe coal is more oxidized. To test this hypothesis, both Illinois No. 6 and Pittsburgh No.8 coals were oxidized in the laboratory. After undergoing the oxidation processdescribed in the experimental procedure, these coal samples were then subjected toflotation tests using dodecane and some of the THF surfactants as collectors. Figs. 4 and5 present plots of the combustible matter recovery as a function of collector dosage forthe flotation of lab-oxidized Illinois No. 6 coal and Pittsburgh No. 8, respectively. Theresults presented in these plots show that the combustible matter recovery is very low

Fig. 5. Comparison of the collecting ability of THF-17, THF-11 and THF-7 with that of dodecane in theflotation of 200-mesh lab-oxidized Pittsburgh No. 8 coal.


when dodecane is used as the collector, but with THF collectors, a much lower dosage isrequired for the same combustible matter recovery. These results indicate that thesurfactants used have the ability to restore the flotability of oxidized coal. At highercollector concentrations, the flotability of oxidized Illinois No. 6 coal is reduced, whichsuggests that the collector may be at the surface as bilayers, with the outer layer havingits oxygen functional groups oriented towards the aqueous phase, thereby reducing theflotability of the coal.

4. Discussion

In mineral flotation, most of the collectors are water-soluble, heteropolar organicŽ .compounds soaps, sulfonates, amines, xanthates, etc. , and they impart hydrophobicity

by adsorption at the mineral–water interface. The polar head group interacts withsurface sites, thereby orienting the hydrocarbon chain of the adsorbed collector towardsthe water. However, it should be pointed out that for the flotation of naturally

w xhydrophobic molybdenite MoS , a hydrocarbon oil is added to enhance its hydropho-2

bicity — similar to coal flotation. In Molybdenite or coal flotation, the oily collectormust act by spreading rather than adsorption.

4.1. Collector spreading

To be effective, the insoluble oils used as collectors for the flotation of coal mustspread over the coal surface. If the coal is somewhat oxidized, water molecules will tendto interact with the coal surface, reducing the oilrwater contact angle, thereby reducingthe tendency of the oil to spread over the coal surface. The spreading coefficient, S ,orw

Ž .for an oil phase displacing water on a solid surface is given by Berg, 1993 :

S sg cosuy1Ž .o r w ow

where g is the oilrwater interfacial tension and u is the contact angle measuredow

through the oil phase. The spreading coefficient is determined by both the contact angleand the oilrwater interfacial tension. Polar oils, such as the THF esters, will tend toinfluence both parameters.

For Illinois No. 6 coal, the oilrwater contact angles, measured across the oil phase,were found to be 1438 for dodecane and 818 for THF-7. The interfacial tension for thedodecanerwater system is 51 mNrm and that for THF-7rwater was measured by thependant drop method to be 10 mNrm. Thus, the spreading coefficient for dodecane inthe dodecanerwaterrcoal system is y91 and that for the THF-7rwaterrcoal system isy10, indicating a much greater tendency for THF-7 than dodecane to spread on IllinoisNo. 6 coal particles that are immersed in water. In the case of Pittsburgh No. 8 coal, thesame two contact angles were found to be 1158 for dodecane and 668 for THF-7. Thespreading coefficient for dodecane on Pittsburgh No. 8 coal is y72 and that for THF-7is y6. These numbers are in agreement with the greater hydrophobicity and greaterflotability of Pittsburgh No. 8 coal.


The equilibrium thickness, h, of an oil film on a substrate can be calculated from theŽ .spreading coefficient Davies and Rideal, 1961 :

Ssy1r2 D r gh2Ž .

Where D r is the difference in density between the oil and water, and g is thegravimetric constant. Clearly, polar oils will tend to spread as a thin film, whereas thenonpolar oil may tend to contract to tend towards the equilibrium thickness, h. Themagnitude of D r for dodecanerwater is 0.24 grcm3 and for THF-7rwater is 0.04grcm3. The calculated equilibrium film thickness for dodecane on Pittsburgh No. 8 coalin water is 0.78 cm and that for THF-7 is 0.22 cm. The equilibrium film thickness islarge because of the extremely low density differences between the oil and water phase.However, substances that reduce the interfacial tension will reduce the equilibrium filmthickness on the surface and should promote spreading.

Table 5ŽAqueous solubility and boiling points of some four-carbon oxygenated compounds Lange, 1961; Aldrich,

.1995


4.2. Structural effect of reagent on the flotation response

In considering details of the action of the non-ionic surfactants studied here, it shouldŽ .be noted that they are bifunctional: they possess oxygenated functional groups A and a

Ž .hydrocarbon chain B . In the case of the THF series of reagents, the oxygenatedŽfunctional groups themselves have two components: the tetrahydrofurfuran a ring-struc-

.tured saturated ether and a methyl ester. Hydrogen bonding plays an important role indetermining the physical properties of such compounds. Table 5 summarizes theaqueous solubility and the boiling point of various oxygenated four-carbon compounds.

Considering the simple oxygenated compounds, all have significant solubility in therange of 8 parts per 100 parts water, which indicates strong hydrogen bonding with

Ž .water. Except for diethyl ether which cannot hydrogen bond with itself , the relativelyhigh boiling points indicate internal hydrogen bonding of molecules with each other.Reducing the number of carbons by one in the case of butanol to propanol and ethylmethyl ketone to acetone leads to compounds that are completely miscible with water.

Furan, which is a ring-structured four-carbon unsaturated ether, is insoluble in waterand has a boiling point similar to that of diethyl ether. The double bonds in the moleculestrongly attract the electrons away from the oxygen, thereby eliminating its ability tohydrogen bond. On the other hand, tetrahydrofurfuran, which is simply comprised of aring-structured saturated alkane, is miscible with water and boils at 668C, both indicatingstrong hydrogen bonding capacity. THF acetate, which is the basis of this series ofreagents, is also miscible in water but boils at a high temperature, 1948C. The strongtendency of both the ether group in the THF molecule and the ester group to hydrogen

Fig. 6. Comparison of the dosages required of the different reagents for the flotation of Illinois No. 6 coal at70% combustible matter recovery, showing the effect of chain structure of the THF esters on theireffectiveness as collectors.


bond must be responsible for hydrogen bonding action with polar sites on the coalsurface. In order to reduce their solubility and to give the surfactants an oily character,hydrocarbon chains of various configurations were attached to the ester functional groupon the THF series of reagents.

To see the effect of the chemical structure of the collectors on the flotation responseof the coal, the reagents for 70% combustible matter recovery were taken from theresults given in Figs. 2 and 3. The reagent requirements expressed in terms of molrtonfor Pittsburgh No. 8 are given in the left part of Fig. 6 and for Illinois No. 6 coal in theright part of that figure. The effectiveness of dodecane and nonylbenzene are includedfor comparative purposes. With Illinois No. 6 coal, the amount of dodecane required foradequate flotation is 50 times that of the appropriate THF compounds. The aromaticnonylbenzene is considerably more effective than dodecane. For flotation of the morehydrophobic Pittsburgh No. 8 coal, only about a tenth of the amount of dodecane isrequired in comparison to the performance of Illinois No. 6. In assessing collectorinteraction with bituminous coals, it should be noted that surface of these coals is

Žcomprised of aromatic carbonaceous sites plus oxygenated functional groups and.contained mineral matters . This figure shows that the results exhibit distinctly different

behavior, depending on the hydrophobicity of the coal and whether the reagents arealiphatic or aromatic.

For both coals, the results given in Fig. 6 show the following for reagent effective-ness:

w x w xaromatic hydrocarbon oil ) aliphatic hydrocarbon oilw x w xoxygenated non-ionic surfactants ) hydrocarbon oilw x w xoxygenated aliphatic surfactants ) oxygenated aromatic surfactants .In the case of the THF series of collectors, their interaction with the coal surface is

strongly influenced by hydrogen bonding phenomena associated with the oxygen atomsin the THF ester part of the molecule. With Illinois No. 6 coal, the most effectivecollectors are the aliphatic THF esters, with the aromatic collectors being less effective.Possibly, with the aromatic collectors, interaction of the benzene ring with the coalsurface competes with hydrogen bonding at the other end of the collector molecule, andthereby affects the orientation of the molecules at the interface. The role of all thereagents in the flotation of the more hydrophobic Pittsburgh No. 8 coal is lesspronounced than for Illinois No. 6. Interestingly, THF-3 appears to be the most effectiveof the shorter chained reagents, possibly due to the lower ratio of hydrophilic-to-hydro-phobic sites on Pittsburgh No. 8 coal.

4.3. Collection mechanisms

The attachment of the THF series of reagents to the coal surface can take place bythree mechanisms:1. Hydrogen bonding of oxygen atoms in the THF ester polar part of the molecule with

oxygenated surface sites on the coal.2. Hydrophobic bonding of the aliphatic hydrocarbon chain with the hydrophobic sites

on the coal surface.


3. p-Bonding of the benzene ring on the hydrocarbon chain of the collector witharomatic sites on the coal surface.The interplay of these three mechanisms determines the effectiveness of the various

THF reagents as a coal flotation collector. For example, with low-rank coals that containa large fraction of oxygen functional groups, such as Illinois No. 6, attachment of thecollector is through hydrogen bonding with the polar oxygen atoms of the collector. Thecollector chain will be oriented away from the coal surface, with the effectiveness of thecollector increasing as the aliphatic hydrocarbon chain length is increased. Even though

Ž .THF-1 has the shortest alkyl chain only one CH group and is miscible with water, it3

must adsorb through a hydrogen bonding mechanism that imparts some hydrophobicityto the coal surface. Fig. 3 shows that it is about as effective as dodecane. The aromaticTHF collectors as a group are less effective, probably because of competition of thebenzene ring for the surface in relation to the oxygenated head group on the collector.This might lead to some orientation of the polar head of the collectors towards the water,possibly requiring a bilayer type of situation to achieve flotation response equivalent tothat for aliphatic reagents. Without any oxygenated functional groups on the molecule,dodecane has limited affinity for a hydrophilic coal surface, thereby requiring a muchlarger dosage for equivalent flotation. On the other hand, the aromatic hydrocarbon

Fig. 7. Schematic representation of interaction between surfactant molecules and coal surface, showing A aspolar head and B as hydrocarbon chain.


nonylbenzene exhibits significantly greater collecting ability because of the interactionof the benzene ring with aromatic groups in the coal surface. These interactionmechanisms are illustrated schematically in Fig. 7. In this schematic representation, theaffinity of the individual functional group of a reagent for the coal surface is representedas a .

4.4. Ash rejection

Fig. 8 presents a plot of ash rejection as a function of the combustible matterrecovery for the flotation of y200-mesh Illinois No. 6 coal with the various collectors.The results show that in terms of ash rejection, similar selectivity can be achieved withthese non-ionic surfactants as compared to dodecane. The same trend was observed forPittsburgh No. 8 coal.

4.5. Surface properties of coal on the performance of collectors

In order to reduce their solubility and to give the surfactants an oily character, ahydrocarbon chain was attached to the THF functional group on the THF series ofreagents. Reagent performance was compared with two hydrocarbons: dodecane andnonylbenzene. Nonylbenzene was selected on the basis of a benzene ring beingconsidered equivalent to a three-carbon alkyl chain in surface chemistry phenomenaŽ .Shinoda et al., 1963 . As already stated, the results given in Figs. 1 and 2 indicate that

Fig. 8. Ash rejection as a function of combustible matter recovery in the flotation of 200-mesh lab-oxidizedIllinois No. 6 coal with various collectors.


the aromatic nonylbenzene is more effective as a collector than the alkane, dodecane.Thus, there must be some interaction of the benzene ring with aromatic sites on the coalsurface.

Table 2 shows that Pittsburgh No. 8 bituminous coal is considerably more hydropho-bic than Illinois No. 6 coal. This means that Pittsburgh No. 8 coal will have a greaterhydrocarbon content than Illinois No. 6 coal. The surface of Illinois No. 6 coal willconsist of aromatic carbonaceous sites plus oxygenated functional groups, whereas thesurface of Pittsburgh No. 8 coal will consist of hydrocarbon and aromatic carbonaceousgroups, but fewer oxygenated functional groups. Clearly, the interaction of the variouscollectors will be different for the two coals.

For Illinois No. 6 coal, the THF reagents are very effective flotation collectors, withtheir effectiveness increasing as the aliphatic hydrocarbon chain length is increased. Thearomatic collectors are less effective, possibly because of competition of the benzenering for the surface in relation to the oxygenated head group on the collector.

From the plots given in Figs. 2 and 3, we can see that the improvement in theflotation performance with these non-ionic surfactants over dodecane for the morehydrophobic Pittsburgh No. 8 coal is not as significant as for the more hydrophilicIllinois No. 6 coal. The most effective collector for Pittsburgh No. 8 coal is THF-3,indicating an optimum in the balance between the hydrophobic and hydrophilic part ofthe collector molecule for the more hydrophobic coal surface. While THF-1 has theshortest alkyl chain, it is miscible with water and even if the hydrogen bondingmechanism dominates, its short chain is not long enough to make the coal surface morehydrophobic. The flotation results for both coals indicate that the main role of the polargroup of the THF collectors is interaction with polar sites on the coal surface, and thisinteraction may be enhanced if the surface of the coal contains numerous oxygenatedgroups, as in the case of Illinois No. 6 coal, or if the coal has been weathered. Becauseof this, the non-ionic surfactants can spread on a hydrophilic coal surface more readilythan an oily hydrocarbon collector, thereby creating a hydrophobic surface. Therefore,the improvement over dodecane in flotation performance with these surfactants is more

Ž .pronounced for the hydrophilic coal Illinois No. 6 . In the case of a more hydrophobiccoal such as Pittsburgh No. 8 coal, the spreading of dodecane on the hydrophobicsurface takes place quite readily and, therefore, the improvement in flotation perfor-mance with THF surfactants is less pronounced than for Illinois No. 6 coal. This is ofsignificance when we consider that approximately 23% of the world’s economicallyrecoverable coal reserves are subbituminous or lignite, and these portions are even

Ž .higher in Canada and the US Mikula, 1991 .

5. Summary and conclusions

For the flotation of Illinois No. 6 and Pittsburgh No. 8 coals, non-ionic THF esterŽ . Ž .surfactants THF series are more effective collectors than dodecane an oily collector

for both oxidized and unoxidized coals. For comparable combustible matter recovery,these non-ionic surfactants require substantially lower dosages than dodecane withsimilar flotation selectivity in terms of ash rejection and pyrite sulfur rejection. The


reagents function by hydrogen bonding with surface oxygenated sites on the coal surfaceand through hydrophobic bonding of the hydrocarbon chain of the collector withhydrophobic carbonaceous sites on the coal. Depending on the rank of the coal and itshydrophobicity, one or the other of these interactions mechanisms may be moredominant. Experiments on the flotation of laboratory-oxidized coal show that, while it isdifficult to float oxidized coal with dodecane, the THF series of reagents have thecapability of restoring the flotability of oxidized coal through their ability to hydrogenbond to oxygen functional groups on the oxidized coal surface.

Nonylbenzene was found to be a better flotation collector than dodecane for thesetwo coals, indicating strong interaction of the benzene ring with aromatic sites on thecoal surface. This type of interaction also has an influence on the performance of THFreagents when a benzene ring is incorporated into the collector chain. In particular,flotation is less effective with aromatic THF reagents than with the alkyl analogs,probably because of molecular orientation that results from competition between thebenzene ring and the oxygenated functional groups for different sites on the coal surface.

References

Aldrich Catalog, 1994–1995. Handbook of Fine Chemicals.Ž .Aplan, F.F., 1976. Coal flotation. In: Fuerstenau, M.C. Eds. , Flotation. A.M. Gaudin Memorial Volume,

Vol.2. AIME, New York, pp. 1235–1264.Ž .Aplan, F.F., 1988. How the nature of raw coal influences its cleaning. In: Klimpel, R.R., Luckie, P.T. Eds. ,

Industrial Practice of Fine Coal Processing. SME, Littleton, CO, pp. 99–111.Aplan, F.F., 1999. The historical development of coal flotation in the United States. In: Parekh, B.K., Miller,

Ž .J.D., Eds. , Advances in Flotation Technology. SME, Littleton, CO, pp. 269–287.Ž .Aplan, F.F., Arnold, B.J., 1991. Flotation. In: Leonard, J.W., III, Hardinge, B.C. Eds. , Coal Preparation.

SME, Littleton, CO, pp. 450–485.Ž .Berg, J.C., 1993. Role of acid–base interactions in wetting and related phenomena. In: Berg, J.C. Ed. ,

Wettability. Marcel Dekker, New York, pp. 75–148.Ž .Brown, D.J., 1962. Coal flotation. In: Fuerstenau, D.W. Ed. , Froth Flotation, 50th Anniversary Volume.

AIME, New York, pp. 518–538.Davies, J.T., Rideal, E.K., 1961. Interfacial Phenomena. Academic Press, New York, pp. 59–52.Fuerstenau, D.W., Diao, J., 1990. Surface Properties of Coal and Their Role in Coal Beneficiation, Final

Report, Project No. DE-FG22-86PC90507.Fuerstenau, D.W., Rosenbaum, J.M., Laskowski, J.S., 1983. Effect of surface functional group on the flotation

of coal. Colloids and Surfaces 8, 153–174.Fuerstenau, D.W., Yang, G.C.C., Laskowski, J.S., 1987. Oxidation phenomena in coal flotation: Part I.

Correlation between oxygen functional group concentration, immersion wettability and salt flotationresponse. Coal Preparation 4, 161–182.

Fuerstenau, D.W., et al., 1992. Coal Surface Control for Advanced Fine Coal Flotation, Final Report, ProjectNo. DE-AC22-88PC88878.

Gutierrez-Rodriguez, J.A., Aplan, F.F., 1984. The effect of oxygen on the hydrophobicity and flotability ofcoal. Colloids and Surfaces 12, 27–51.

Harris, G.H., Diao, J., Fuerstenau, D.W., 1995. Coal flotation with nonionic surfactants. Coal Preparation 16,135–147.

Hong, B.D., 1998. Coal. In: Annual Review 1997. Mining Engineering, Vol. 50. No. 5, pp. 49–120.Lange, N.A., 1961. Handbook of Chemistry. 10th edn., McGraw-Hill, New York.

Ž .Laskowski, J.S., 1995. Coal surface chemistry and its effects on fine coal processing. In: Kawatra, S.K. Ed. ,High Efficiency Coal Preparation: An International Symposium. SME, Littleton, CO, pp. 163–176.


Ž .Laskowski, J.S., Miller, J.D., 1984. New reagents in coal flotation. In: Jones, M.J., Oblatt, R. Eds. , Reagentsin the Mineral Industry. The Institution of Mining and Metallurgy, pp. 145–154.

Mattson, J.S., Mark, H.B., Jr., 1971. Activated Carbon, Surface Chemistry and Adsorption from Solution.Marcel Dekker, New York, pp. 209–222.

Ž .Mikula, R.J., 1991. Cleaning low-rank coal. In: Leonard, J.W., III, Hardinge, B.C. Eds. , Coal Preparation.SME, Littleton, CO, pp. 450–485.

Philips, K.M., Glanville, J.D., Wightman, J.P., 1987. Heat of immersion of Virginia-C coal in water as afunction of surface oxidation. Colloids and Surfaces 21, 1–8.

Shinoda, K., Nakagawa, T., Tamamushi, B., Isemura, T., 1963. Colloidal surfactants. Chap. 1, AcademicPress, New York.

Swann, P.D., Allardice, D.J., Evans, D.G., 1972. Low-temperature oxidation of brown coal: 1. Change ininternal surface due to oxidation. Fuel 53, 85–87.

collectors for low rank coal & oxidized coal

Documents