Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

On the interactions between naphthenic acids and inorganic minerals Nodwell, Maximilian 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2011_fall_nodwell_maximilian.pdf [ 1.05MB ]
Metadata
JSON: 24-1.0059078.json
JSON-LD: 24-1.0059078-ld.json
RDF/XML (Pretty): 24-1.0059078-rdf.xml
RDF/JSON: 24-1.0059078-rdf.json
Turtle: 24-1.0059078-turtle.txt
N-Triples: 24-1.0059078-rdf-ntriples.txt
Original Record: 24-1.0059078-source.json
Full Text
24-1.0059078-fulltext.txt
Citation
24-1.0059078.ris

Full Text

On the Interactions between Naphthenic Acids and Inorganic Minerals by Maximilian Nodwell B.A.Sc. Metals and Materials Engineering, University of British Columbia, 2003 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of Applied Science in THE FACULTY OF GRADUATE STUDIES (Chemical and Biological Engineering) The University Of British Columbia (Vancouver) August 2011 c©Maximilian Nodwell, 2011 Abstract Naphthenic acids are a family of carboxylic acids that are found in oil sands bi- tumen. These compounds partition to the aqueous phase during extraction and refining and are toxic to various biota. The removal of these acids from solution is difficult due to their low concentrations, complexity of the mixture and poor un- derstanding of the behaviour of the mixed compounds. In particular, partitioning of these organic acids to solid surfaces is not well understood. Knowledge of this equilibria would be helpful for potential process development. The research presented here describes the adsorption of two surrogate naph- thenic acids onto inorganic minerals (copper sulphide and copper hydroxide). De- canoic acid and cyclohexane pentanoic acid were found to be insoluble in water at pH 3, leading to hydrophobic adsorption onto the minerals and the reaction vessel surfaces. At pH 8.5, both acids formed insoluble copper-carboxylate complexes when mixed with the minerals. The hypothesized 2:1 acid:copper stoichiometry was confirmed. The mechanism of complexation varied with the reaction con- ditions; both chelating and bridging complexes were observed in the resultant metallo-organic solids. The relative hydrophobicity of the two NA surrogates was also found to contribute to the different adsorption trends. During the pH 8.5 reactions, the solution pHs were found to drop. The uncon- trolled decreases in pH had significant effect on the water-solid partition and on the apparent mineral loading of the organics. It appears that soluble copper cations have a higher extent of reaction with the carboxylate anions than does copper con- tained in the mineral solids. Quantification of these reactions is difficult; however this research does enable conclusions about how the organic acids and inorganic minerals interact and sets the stage for future research. ii Table of Contents Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.1 Naphthenic Acid Literature . . . . . . . . . . . . . . . . . . . . . 4 2.1.1 Treatment methods . . . . . . . . . . . . . . . . . . . . . 4 2.1.2 Biological/toxicological studies . . . . . . . . . . . . . . 6 2.1.3 Analytical detection of naphthenic acids . . . . . . . . . . 7 2.1.4 Geohydrology . . . . . . . . . . . . . . . . . . . . . . . 8 2.2 Metal Carboxylates . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2.1 Metallorganic bond types and other interactions . . . . . . 9 2.2.2 Solubility . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2.3 Infrared spectroscopy . . . . . . . . . . . . . . . . . . . . 10 2.3 Other Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.3.1 Surface chemistry . . . . . . . . . . . . . . . . . . . . . . 14 2.3.2 Flotation chemistry . . . . . . . . . . . . . . . . . . . . . 16 2.3.3 Wood preservers . . . . . . . . . . . . . . . . . . . . . . 17 iii 3 Research Hypotheses . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.1 Low pH Reactions . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.2 High pH Reactions . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.2.1 Copper sulphide at high pH . . . . . . . . . . . . . . . . 20 3.2.2 Copper hydroxide titration . . . . . . . . . . . . . . . . . 20 3.2.3 Copper hydroxide co-precipitation . . . . . . . . . . . . . 21 3.2.4 Coordination complexes . . . . . . . . . . . . . . . . . . 21 3.2.5 Stoichiometry of reaction . . . . . . . . . . . . . . . . . . 22 4 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.1 Adsorptive Equilibria . . . . . . . . . . . . . . . . . . . . . . . . 23 4.1.1 Minerals . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.1.2 Surrogate naphthenic acids . . . . . . . . . . . . . . . . . 24 4.1.3 Adsorption equilibria . . . . . . . . . . . . . . . . . . . . 26 4.1.4 Total organic carbon (TOC) . . . . . . . . . . . . . . . . 27 4.2 Solid Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 5 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 32 5.1 Adsorption Results . . . . . . . . . . . . . . . . . . . . . . . . . 32 5.1.1 Low pH adsorption . . . . . . . . . . . . . . . . . . . . . 34 5.1.2 High pH adsorption and coprecipitation . . . . . . . . . . 34 5.1.3 Adsorption effects . . . . . . . . . . . . . . . . . . . . . 38 5.2 Infrared Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . 43 5.2.1 Copper sulphide at low pH . . . . . . . . . . . . . . . . . 44 5.2.2 Copper sulphide at high pH . . . . . . . . . . . . . . . . 46 5.2.3 Copper hydroxide at high pH . . . . . . . . . . . . . . . . 48 5.2.4 Copper hydroxide co-precipitation . . . . . . . . . . . . . 51 5.2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . 54 5.3 Process Implications . . . . . . . . . . . . . . . . . . . . . . . . 54 5.4 Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 iv List of Tables Table 2.1 Referenced infrared spectral bands associated with carboxylic acids and their metallic soaps . . . . . . . . . . . . . . . . . . 12 Table 2.2 Referenced infrared spectral bands associated with various in- organic compounds . . . . . . . . . . . . . . . . . . . . . . . 13 Table 3.1 Experimental matrix, short form . . . . . . . . . . . . . . . . . 20 Table 4.1 Properties of the selected naphthenic acid surrogates (from man- ufacturer’s MSDS) . . . . . . . . . . . . . . . . . . . . . . . . 25 Table 4.2 Shimadzu TOC-V CPH instrument settings . . . . . . . . . . . 29 Table 5.1 Summary of Langmuir isotherm fitting parameters for DA and CHPA at high pH . . . . . . . . . . . . . . . . . . . . . . . . 35 Table 5.2 Summary of Langmuir isotherm fitting parameters for DA up- take with respect to initial acid concentration . . . . . . . . . . 38 Table 5.3 Summary of key spectral bands for CuS mixed with DA and CHPA at low pH . . . . . . . . . . . . . . . . . . . . . . . . . 44 Table 5.4 Summary of key spectral bands for CuS mixed with DA and CHPA at high pH . . . . . . . . . . . . . . . . . . . . . . . . 46 Table 5.5 Summary of key spectral bands for Cu(OH)2 mixed with DA and CHPA at high pH . . . . . . . . . . . . . . . . . . . . . . 50 Table 5.6 Summary of key spectral bands for Cu(OH)2 co-precipitated with DA and CHPA at high pH . . . . . . . . . . . . . . . . . 53 v List of Figures Figure 2.1 Example of an ionic sodium carboxylate compound. Note that the single double bond to oxygen is now a shared resonant bond between the two oxygens. . . . . . . . . . . . . . . . . 9 Figure 2.2 Three-dimensional rendering of copper carboxylate complexes. Images are adapted from Mehrotra and Bohra [1983] . . . . . 13 Figure 2.3 Typical linear, Langmuir and Freundlich adsorption isotherms 16 Figure 5.1 Naphthenic acid surrogate mineral loading as equilibrium acid concentration . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Figure 5.2 Decanoic acid mineral loading as equilibrium acid concentra- tion - high acid:metal ratios . . . . . . . . . . . . . . . . . . 36 Figure 5.3 Decanoic acid loading as function of initial acid concentrations 37 Figure 5.4 Naphthenic acid surrogate partitions from water phase as a function of experimental pH change . . . . . . . . . . . . . . 39 Figure 5.5 Naphthenic acid surrogate fractional uptake to mineral as a function of final solution pH . . . . . . . . . . . . . . . . . . 41 Figure 5.6 Solubility of copper cation and decanoate anion species with pH 42 Figure 5.7 IR spectrograph of naphthenic acid surrogates mixed with CuS at low pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Figure 5.8 IR spectrograph of naphthenic acid surrogates mixed with CuS at high pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Figure 5.9 IR spectrograph of naphthenic acid surrogates mixed with Cu(OH)2 at high pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Figure 5.10 IR spectrograph of naphthenic acid surrogates co-precipitated with Cu(OH)2 . . . . . . . . . . . . . . . . . . . . . . . . . . 52 vi Acknowledgments My sincere thanks go to Dr. Sue Baldwin for giving me the support and space needed to complete this project. Many thanks go to BioteQ Environmental Technologies for their generous mul- tidimensional support for the duration of this research. In particular, Dr. David Kratochvil has been a wonderful mentor for many years. I am also grateful to NSERC for providing funding for my studies. To Angela and Maxine, maholo for everything. vii Chapter 1 Introduction Let’s have a little talk about tweetle beetles, what do you know about tweetle beetles? — Dr. Seuss (1965) Naphthenic acids (NAs) are a complex mixture of cyclic or acyclic aliphatic organic compounds characterized by a carboxyl reactive group [Brient et al., 2000]. The general formula isCnH2n+zO2, where n is the carbon number and z is a negative multiple of two, indicative of the number of rings of the compound. In general, n ranges from 5 to 30, and z ranges from 0 to -12 (z = -12 would be a six ring structure). The distribution of n and z values of naphthenic acids varies depending on the source, and no one mixture can be considered representative of the family of compounds [Rogers et al., 2002a]. In the Athabasca oil sands surface mining operations, bitumen ore is excavated using loaders and dump trucks. This ore is taken to the upgrading facility, where the bitumen is washed from the sand using hot caustic water of approximately pH 8.5. The oil and water are separated and the bitumen is upgraded to light sweet synthetic crude prior to pumping to a final refinery. The water is either recycled or stored in tailings impoundment [Headley and McMartin, 2004]. Due to the polar carboxyl group, naphthenic acid compounds partition to the water phase. This partition is encouraged by the high pH. Even at 85 - 95% water recycle rates, oil sands bitumen extraction requires a large water withdrawal from local surface water sources, primarily the Athabasca river. The fresh water to oil ratio is, on average, 3m3 H2O:1m3 oil [Allen, 2008] - the total water use, includ- 1 ing recycled process water, is much higher. Consequently, substantial volumes of water are contaminated with these compounds. Typical concentrations of naph- thenic acids in oil sands tailing ponds are on the order of 100mg/L [Headley and McMartin, 2004]. Naphthenic acids at these concentrations are toxic to aquatic and mammalian life [Headley and McMartin, 2004]. Because of this toxicity, oil sands opera- tions have a zero liquid discharge policy to date. As a result, vast quantities of oil sands process-affected water (OSPW) are stockpiled on the oil sands surface mining leases, with a total volume exceeding one billion cubic metres [Han et al., 2009] covering a total area greater than 130km2 [Kunzig, 2009]. Although this storage prevents the export of toxins to the nearby environment, holding ponds of this size effectively create a local environment that can import nearby species [Ronconi, 2006]; migratory birds are particularly vulnerable. Over and above the scale of the water pollutant issue is the regulatory frame- work in which oil sands producers operate. The only water regulation in the oil sands industry is zero-discharge with no water quality guidelines for eventual dis- charge [Gosselin et al., 2010]. In addition, naphthenic acids are not listed as reg- ulated or monitored substances in either provincial (Alberta Environmental Pro- tection and Enhancement Act, Alberta Water Act) or federal (Canadian Environ- mental Protection Act, Canada Water Act, Fisheries Act, Chemicals Management Plan) legislation. This lack of target clarity renders effluent treatment difficult to navigate. These tailings ponds remain one of the most challenging aspects to oil sands surface mining. Regulatory hurdles aside, the future discharge of water and site reclamation requires improved understanding of the naphthenic acid compounds to create an effective method of detoxification or removal. A survey of the literature can re- veal how little was known about NAs up until several years ago; much of the NA research to date focuses on the analytical methods for measuring and characteriz- ing the substances in aqueous solution. These analytical methods, in turn, have helped the study of biodegradation, as researchers can measure the distribution of NA compounds and their selective degradation during treatment. Although these studies have made great advances, they are still in early stages and must be expanded. In particular, the aqueous/solids interactions are of interest. 2 This is so, as an understanding of the absorptivity of the compounds on various soil and mineral substrates could allow modelling of the movement of NAs in ground- water. Such interactions may also be relevant for biodegradation or even direct removal processes for treatment of NA contaminated water. However, the naph- thenic acid - environmental solids partition research to date tends to describe the results of equilibria instead of a reaction mechanism. For example, the sorption studies examine the interactions between bulk naphthenic acid samples (native or commercial) with natural soil samples, or model acids with bulk soils, or model soils with bulk acids. The overall effects of these studies cannot be well general- ized. This research attempts to address some of these shortcomings. In particular, this research observes the adsorption of singular model or surrogate naphthenic acids onto inorganic minerals (metal sulphides and hydroxides). The objective is to describe: 1. how the different naphthenic acid compounds interact with inorganic miner- als; and 2. what is the mechanism of naphthenic acid sorption onto inorganic minerals. 3 Chapter 2 Literature Review 2.1 Naphthenic Acid Literature The literature on naphthenic acids spans several fields. The key areas of related re- search with respect to the work presented herein include potential treatment meth- ods, toxicology, current analytical methods in industry and the hydrological trans- port of the compounds in ecosystems. Peripherally, there is also research in the areas of metal carboxylates, surface chemistry, mineral flotation and wood preser- vation that are relevant to the research hypotheses. 2.1.1 Treatment methods Oil sands producers have been aware since early in the industry that naphthenic acids are the most toxic agents in the tailings water [MacKinnon and Boerger, 1986]. Although there is much research into removal of naphthenic acids from oil phase (as a means of upgrading and managing corrosion), the studies into removal of these acids from water phase are fewer. One early study into the remediation of surface mine tailings ponds [MacKin- non and Boerger, 1986] showed two different treatments that resulted in toxicity decreases of tailings water. The first method involves acidification of the water to pH 4, with or without coagulents and flocculants. This re-associates the naphthenic acids, as the pKa of carboxylic acids is typically around 4.9. In this state, NAs are no longer soluble in water, and partition to the surface of the clay and silt particles 4 in the tailings. This decreases the acid content of the water, as well as effectively flocculating the solid particles. The resultant clarified water, upon re-neutralization to pH 7, showed zero acute toxicity, with an EC50 > 100 (by 15 minute Microtox assay). This treatment process was never adopted by industry, in spite of this ap- parent success. Although no OPEX analysis was considered in the study, one may speculate that the cost of acidification and re-neutraulization of such large volumes of water was the fatal flaw. MacKinnon and Boerger [1986] also demonstrated that long term aeration and storage of OSPW results in a decrease in acute toxicity. The EC50 of the treated and stored water was elevated from 30% to approximately 90% after 10 months storage. The detoxification was due to natural oxidation and physical or biological degradation of the NA compounds, either to complete mineralization or simply to more benign materials. This was seen as evidence that the produced waters could treat themselves if stockpiled in ponds. However, after the initial detoxification period of 10 months, no water quality improvements were observed even with a further 6 months storage. Two decades later, Holowenko et al. [2002] demonstrated that the high-molecular- weight (HMW) fraction of naphthenic acids (carbon number greater than 22) are refractory to biodegradation. However they tend to contribute less to the overall toxicity of the contaminated water [Frank et al., 2008] (more details on NA toxic- ity will be discussed later). Consequently, alternative treatment methods for tailings water were sought. Ozone has been shown to successfully oxidize NAs, either to complete solution non-toxicity [Scott et al., 2008b] or to the point that biodegradation is enhanced [Martin et al., 2010]. However, neither of these two ozone studes addressed the economics or energy consumption of the application, which is known to be an energy intensive process [Bijan and Mohseni, 2005]. As such, biodegradation is still considered the most economical method by which to remove the water toxicity. Quagraine et al. [2005] has suggested that biodegradation of the HMW fraction of NAs may be enhanced by prior sorption to the surface of some immobilizing material, such as clay or polymer beads. This would serve to concentrate the NAs as well as the bacterial populations that may consume NAs as a substrate. Given that the naphthenic acid salts found in the 5 OSPW are dissociated and readily water soluble, and considering that the clay, sand and silt in the OSPW does little to remove NAs, it is likely that an alterna- tive medium for solid-surface concentration must be found for this biodegradation method. This is, in part, a motivation for this current research. 2.1.2 Biological/toxicological studies Toxicology studies are concerned with the impact of naphthenic acid compounds in the biosphere. The research includes toxicity of the compounds in various or- ganisms (including aspen trees, fish, zooplankton, rats and microbes [Clemente and Fedorak, 2005]), as well as elucidating some possible mechanisms for the NA bio-activity. Rogers et al. [2002b] showed that rats fed sub-acute portions of NAs had enlarged livers (35% increase) compared to control groups. Although the rats’ hearts, kidneys and testes were also enlarged (by smaller increases), the authors concluded that the liver is the specific organ target of naphthenic acids. However, Frank et al. [2009] countered this suggestion by noting that, beyond a carboxyl moiety, there are no other functional groups on NAs that would target a specific physiological receptor. Considering that four rat organs were enlarged by NA ingestion [Rogers et al., 2002b], the claim of specificity may be over confident. Frank et al. [2009] suggest that NA toxicity is likely due to cell narcosis, a non- specific mechanism whereby the hydrophobic portion of the molecule can enter the lipid bylayer of a cell and disrupt membrane function [Roberts, 1991]. A con- sequence of this mechanism would be an increase in toxic effect with increasing molecular weight [Protic and Sabljic, 1989] - the larger HMW fraction NAs would be expected to cause more disruption to cell membranes, create more physiological disruption and manifest as the more harmful compounds. However, as suggested in several of the aforementioned studies, it is actually the smaller, lower molecular weight (LMW) fraction NAs (carbon number less than 21) that demonstrate higher toxicity [Frank et al., 2008]. This apparent contradiction between theory and observation may be explained by hypothesizing the presence of multi-carboxyl group HMW NAs in natural sam- ples. Such compounds would be more hydrophilic in spite of their large alkyl group, and thus less disruptive to cell membranes [Frank et al., 2009]. While this 6 may be a plausible explanation of toxic effects, it does not explain why these HMW NAs are less easily biodegraded. 2.1.3 Analytical detection of naphthenic acids These papers are concerned with the analytical techniques for determining the dis- tribution of compounds in naphthenic acid mixtures, and are applied to native and commercial samples. The principle methods include: Gas-chromatography/mass-spectrometry (GC-MS) This method can determine the molecular weight of the various compounds in a mixture. Prior to analysis, the organic acids must be extracted from the aque- ous phase using dichloromethane (DCM). After extraction and before analysis, the samples require derivatization with N-methyl-N-(t-butyldimethylsilyl)trifluoroacetamide, such that the compounds are more thermally stable [Headley et al., 2009]. The t- butyldimethylsilyl derivatives are analyzed from the DCM phase directly. While the GC-MS method offers the ability to resolve the constituent compounds, this method can only give relative quantities instead of absolute concentrations [Holowenko et al., 2002]. Fourier-transform infrared spectrometry (FTIR) This method is the industry standard for determination of the bulk concentration of naphthenic acids in oil sands waste water. The acids are extracted from the aqueous phase with DCM and dried. The residue is reconstituted in a known quantity of DCM prior to injection into the infrared spectrophotometer. The carboxylic acids are measured at the spectral peaks at 1743 and 1705cm−1. This method may give artificially high readings due to naturally occurring, non-toxic organic acids in the water [Scott et al., 2008a]. High-performance liquid chromatography (HPLC) The use of HPLC to evaluate naphthenic acids has been explored by Clemente et al. [2003] and Yen et al. [2004]. This method requires prior derivitization with 2- nitrophenylhydrazine (NPH) and 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide 7 hydrochloride (EDC) such that the compounds elute effectively. The nitrophenyl- hydrazine derivatives are detectable by 400mn UV-visible absorption, and the total area under the chromatograph is proportional to the total naphthenic acid concen- tration. According to Yen et al. [2004], this method can measure concentrations as low as 5mg/L. While HPLC can accurately measure total concentrations, it cannot distinguish the various compounds like the GC/MS method. 2.1.4 Geohydrology This research focuses on the transport of naphthenic acids in environmental com- partments, primarily water and soils. They typically attempt to characterize the motion of naphthenic acids from pond or spill points to groundwater, and any re- tardation due to soil adsorption. Most adsorption research articles focus on the parameters of linear isotherms and on correlating the water-soil partition coeffi- cient to the organic carbon content of the soils [Peng et al., 2002, Janfada et al., 2006, Piwoni and Banerjee, 1989, Bayard et al., 1998, Banerjee et al., 1997, Zhou et al., 1997]. Surface adsorption models will be discussed later. According to Banerjee et al. [1997], the temperature dependence of the NA-soil adsorption indicates that the reactions are exothermic; activation energies of the reaction equilibrium constants were also computed. Zhou et al. [1997] measured the enthalpy of adsorption of naphthenic acids from toluene onto clays and found that the interaction follows two steps - removal of toluene from the clay surface, then adsorption of the NAs to the solid. Bayard et al. [1998] and Connaughton et al. [1993] considered the desorption of organics from solid surfaces. Bayard et al. [1998] used distilled water to remove naphthalene from carbonaceous soils. Connaughton et al. [1993] used a gas purge method with similar substances. The results showed hysteresis related to the con- taminant contact duration, in that the adsorption and desorption quantities were not symmetric; the desorption of the sorbates did not go to completion, and the extent to which the organics remained on the soils after rinsing or gas purging was a func- tion of the adsorption contact time. These results are not likely to be relevant to this proposed research as the materials, reactions and kinetic focus are so different. 8 OO R Na + Figure 2.1: Example of an ionic sodium carboxylate compound. Note that the single double bond to oxygen is now a shared resonant bond between the two oxygens. 2.2 Metal Carboxylates Naphthenic acids fall under the category of carboxylic acids and are expected to behave as such when interacting with metals in solution or in solid mineral form. There is a substantial literature on the synthesis, detection and properties of metal carboxylates. 2.2.1 Metallorganic bond types and other interactions Interaction or reaction between chemical species implies that some type of bond has formed. The two ideal bonds typically described (adapted here from Atkins [1997]) are ionic bonds and covalent bonds. Ionic bonds occur when species of largely differing electronegativities meet and the more electronegative species strips an electron away from the more electropositive element (eg: chlorine can strip away an electron from sodium). The result is that the electron-poor ion is posi- tively charged and electrostatically attracted to the electron-rich negatively charged ion. The sodium salt of any carboxylic acid falls under this category; an example of this type of bonding is shown in Figure 2.1. A covalent bond occurs when the difference in electronegativities is less pro- nounced, and the two chemical species share electrons to make up outer shell elec- tron deficiencies. Coordination complexes, in the context of this work, are sub- stances whereby a central metallic ion has multiple organic ligands bonding to it covalently. The coordination is that of multiple simultaneous bonds around a cen- tral point. Examples of coordination complexes are shown in Figure 2.2. 9 Hydrophobic interactions are not a bonds, but are phase and energy related phenomena, whereby a hydrophobic substance has a high Gibb’s free energy in water phase, and a lower Gibb’s free energy in organic or on a solid phase. The change in energy of the mixture drives the hydrophobic solute from one phase to another. 2.2.2 Solubility The most promising aspect of mixing inorganic minerals with naphthenic acids as a means of OSPW treatment is that the expected resultant coordination complexes are only sparingly soluble in water. Havre [2002] used naphthenic acid surrogates with carbon number ranging from C11 to C16 and zero to two rings. Upon mixing these surrogate compounds with calcium cations at pH 11.2, solid calcium carboxylates formed. The re- searchers were able to estimate the solubility products of the resultant solids, and found pKsp’s ranging from 7 - 12 and a linear relationship between solubility and the molecular weight of the carboxylic acid. Mauchauffee and Meux [2007] used sodium decanoate to precipitate divalent metal cations from aqueous solution at pH 4.7. They found pKsp’s of 10.16, 13.05 and 14.65 for manganese, zinc and copper decanoates, respectively. While none of these research findings break any records for insolubility of a solid compounds, they indicate that, given the correct conditions, naphthenic acids could potentially be removed to very low concentrations. Furthermore, this implies that organic removal from the effluent stream, after mineral uptake, can be accomplished by conventional solid-liquid separation. 2.2.3 Infrared spectroscopy Infrared spectroscopy is a technique of associating spectral adsorption bands with surface compounds and chemical bonds. For the present experiments, the prin- ciple spectral bands, their assigned chemical bonds and motion and the literature references are outlined in Table 2.1 (organic compounds) and Table 2.2 (inorganic compounds). There is broad agreement in the literature on the key bands associated with carboxylic acids, their metallic soaps and some inorganic species appearing 10 in this research. Although there is some variability in the reported wavenumbers (per Table 2.1), the key points are: 1. Bands centred around 3600cm−1 are assigned to the hydroxide groups on hydroxide minerals 2. Bands centred around 3400cm−1 are assigned to the water of hydration on sulphide minerals 3. The broad band from 3300cm−1 to 2500cm−1 is assigned to O-H stretching on carboxylic acids; this band vanishes when the carboxyl group proton is substituted with a metallic ion 4. The sharp double bands at 2930cm−1 and 2840cm−1 are assigned to the C- H stretching on the carboxylic acid’s hydrocarbon backbone - these groups play little part in the reactions 5. The narrow, sharp band at 1700cm−1 is assigned to the C=O stretch on free carboxylic acids; this band vanishes when the carboxyl-group proton is sub- stituted with a metallic cation 6. The narrow, sharp band between approximately 1600cm−1 and 1530cm−1 is assigned to the antisymmetric C-O stretch on the cation-substituted carboxyl group 7. The narrow, sharp band between approximately 1470cm−1 and 1420cm−1 is assigned to the symmetric C-O stretch on the cation-substituted carboxyl group In addition to the bands as listed, there are other indicators as to the surface species being formed. Mehrotra and Bohra [1983] indicate that, when the dif- ference in wavenumber between the antisymmetric (approx. 1600cm−1) and the symmetric (approx. 1450cm−1) stretching of the carbon - oxygen bond is less than 100cm−1, the resultant complexation structure is chelating (Figure 2.2a). When that difference in wavenumber is in the range 160 - 180cm−1, the resultant com- plex is bridging (Figure 2.2b). Note that the stoichiometry of organic to metal is the 11 Table 2.1: Referenced infrared spectral bands associated with carboxylic acids and their metallic soaps Wavenumber (cm−1) Chemical Group and Motion Reference 3300 to 2500 (s, b) O-H stretching of carboxyl group Brandal et al. [2006] 2700 - 2500 (b) O-H stretching of carboxyl group Mehrotra and Bohra [1983] 2930 - 3020 (b) stretching of -CH3 group Andres et al. [2010] 2928 (s) symmetric C-H stretching Craciun and Kamdem [1997] 2870 (sh) asymmetric C-H stretching Craciun and Kamdem [1997] 2920 (s) asymmetric C-H stretching Ngah and Hanafiah [2008] 2851 (sh) symmetric C-H stretching Ngah and Hanafiah [2008] 1734 (s) C=O stretching Ngah and Hanafiah [2008] 1708 (s) C=O stretching Zyskowski and Kamdem [1999] 1706 (s) dimer carboxyl group stretch Rogers et al. [2002a] 1700 (s) C=O stretching Mehrotra and Bohra [1983] 1700 (s) C=O stretching Brandal et al. [2006] 1636 (s) COO− anion stretch Ngah and Hanafiah [2008] 1603 (s) antisym. C-O stretch (bridging) Mehrotra and Bohra [1983] 1602 (vs) symmetric COO− anion stretch Silva et al. [2006] 1550 (s) antisym. C-O stretch (chelating) Mehrotra and Bohra [1983] 1530 (s) antisym. COO− anion stretch Palacios et al. [2004] 1470 (s) symmetric COO− anion stretch Palacios et al. [2004] 1456 (s) symmetric C-O stretch (chelating) Mehrotra and Bohra [1983] 1441 (vs) antisym. COO− anion stretch Silva et al. [2006] 1418 (s) symmetric C-O stretch (bridging) Mehrotra and Bohra [1983] Abbreviations: vs = very strong, s = strong, w = weak, vw = very weak, b = broad, sp = sharp, sh = shoulder same in both cases, as is the coordination number of the central metallic atom(s). The only difference is the decreased bond angle of the O-C-O bonds in the case of the chelating orientation. This results in a higher energy compound, and is thus the less likely form of the coordination complex. 12 Table 2.2: Referenced infrared spectral bands associated with various inor- ganic compounds Wavenumber (cm−1) Chemical Group and Motion Reference 3637 O-H stretch in NaOH Phillips and Busing [1957] 3580 (s) O-H stretch on Cu(OH)2 Craciun and Kamdem [1997] 3550 - 3200 (b) O-H stretch in water Simonescu et al. [2007] 3424 (s) O-H stretch in water Craciun and Kamdem [1997] 1200 - 950 (b) O-H stretch in water Simonescu et al. [2007] 1144 (s) CuSO4 anhydrous Dunn and Muzenda [2001] 1109 (w) CuSO4*H2O Dunn and Muzenda [2001] 1106 (s) SO4 stretch Silva et al. [2006] Abbreviations: vs = very strong, s = strong, w = weak, vw = very weak, b = broad, sp = sharp, sh = shoulder Cu C C O O R R H2OH2O OO (a) Chelating complex Cu O C R Cu C C C O O O O O O O R R R H2O H2O (b) Bridging complex Figure 2.2: Three-dimensional rendering of copper carboxylate complexes. Images are adapted from Mehrotra and Bohra [1983] . 13 2.3 Other Literature In addition to this core of naphthenic acid literature, the fields of surface chemistry and mineral flotation have yielded some information regarding the behaviour of inorganic surfaces and the adsorption of slightly polar organic molecules. 2.3.1 Surface chemistry Models of surface adsorption There are several models of liquid to solid surface adsorption that may be applied to the phenomena. The simplest is a linear adsorption model whereby the quantity of the sorbate/solute adhering to the surface (q) is a linear function of the concen- tration of the residual solute in the solution (Ce) and the slope of the sorbed vs. dissolved quantities is a single value Kd . The final expression is: q= KdCe (2.1) While linear models are practical for engineering purposes, they typically only represent physical phenomena in sufficiently limited ranges. Alternative models include the Langmuir isotherm, shown by: q= kKLCe (1+ kCe) (2.2) In this equation, q is the quantity of solute adsorbed onto the solid surface, Ce is the equilibrium concentration of solute in the solution and the parameters k and KL are fitting constants that incorporate the activity coefficients of the reagents and products and are generally temperature dependent. Equation 2.2 assumes a monolayer adsorption of the solute to a homogeneous surface and no other inter- actions [Calvet, 1989]. In this model, the adsorption reaches an upper limit and plateaus, in spite of increasing solute concentrations in solution, thus representing the phenomenon of maximal loading. While the Langmuir isotherm is still an ideal, empirically-fit model, it can extend the description of adsorption beyond a narrow linear range. A third adsorption model for this type of interaction is the Freundlich isotherm: 14 q= K fC n f e (2.3) In this equation q is again the quantity of solute adsorbed onto the solid surface, Ce is the equilibrium concentration of solute in the solution and the parameter K f is a fitting constant that incorporate the activity coefficients of the reagents and products. n f relates to the shape of the adsorption curve. The Freundlich isotherm (Equation 2.3) is applicable in situations where the solid surface and the solute to solid interactions are heterogeneous. This is reflected in the shapes of the isotherms. For example, n f = 1 yields a linear relationship indicative of constant partition between the liquid and solid phases, regardless of the solute concentration. This trend implies that the solute can adsorb as readily to the solid surface as to the layers of sorbed solute. n f < 1 yields a plateau in sorbtion after a signifiant initial solid uptake; while the shape may appear similar to the Langmuir isotherm, the Langmuir monolayer cannot be taken for granted. n f > 1 yields a q vs. Ce curve of increasing slope with Ce, indicating that the solute sorbs more readily to itself than to the solid surface. The Langmuir and Freundlich isotherms (with Freundlich n f = 1 doing double duty demonstrating the linear adsorption model) are shown in Figure 2.3 to illustrate the descriptions. Note that the fitting parameters are fabricated to create pleasing and demonstrative axis scales. Mineral surface phenomena The presence of oxy-hydroxides on the surface of metal sulphide minerals is an established phenomenon. Zhu et al. [1994] performed cyclic voltammetric stud- ies to show that pyrite will hydroxylate under almost all applied potentials. The cyclic potential showed current peaks that corresponded to particular pyrite oxida- tion and reduction reactions. Every reaction featured ferric or ferrous hydroxide as a participating species. Bebie et al. [1998] showed a similar result by electrophoretic measurements. If not prepared in anoxic conditions, the iron sulphide oxidizes readily into a pre- dictable hydroxide surface. The ”clean” sulphide surfaces show more heterogene- ity in the various sulphur, metal and hydroxide groups on the wetted surface. These results indicate that true metal sulphides surfaces are hard to come by. 15 0 2 4 6 8 10 12 0 0.2 0.4 0.6 0.8 1 1.2 q (Langmuir) q (nf = 1) q (nf < 1) q (nf >1) Ce (equilibrium concentration) q (s or be d qu an tit y) Figure 2.3: Typical linear, Langmuir and Freundlich adsorption isotherms 2.3.2 Flotation chemistry There are many studies involving adsorption of organic compounds onto inorganic mineral surfaces in the flotation chemistry literature. Of particular relevance is the study of polysaccharides as depressants for the mineral processing industry. Polysaccharides are relevant to the proposed research as they have a alcoholic hy- droxide groups around a single or multiple-ring structure, much like naphthenic acids have a carboxyl-associated hydroxide group on a ring structure. The solid surfaces involved in the adsorption can show variation in properties. Depending on solution pH, applied potential and oxidation conditions, the metal sulphide surfaces may behave more like their corresponding metal hydroxides [Liu et al., 2000, Bogusz et al., 1997, Raju et al., 1997]. This research has found that the hydroxylated surfaces behave like Brønsted bases and donate a hydroxyl group during reaction with polysaccharides [Liu et al., 2000]. The polysaccharides, in 16 turn, behave like Brøtnsted acids and donate protons from their hydroxyl groups. This has been demonstrated experimentally by observing peak adsorption of the organic onto the mineral surface at conditions of maximum surface hydroxylation [Raju et al., 1997]. Bogusz et al. [1997] showed that oxide-free sulphide-mineral surfaces did not adsorb any organics at all, but that upon oxidation, the surface reactivity was restored. Liu et al. [2000] showed an interaction that releases free acidity as dextrin interacts with a metal mono-hydroxide surface. This acid/base reaction complexes the organic with the inorganic surface and release a hydronium ion in the process: R− (OH)2 +MeOH↔ R−O2Me+H3O+ (2.4) Experimentally, this mechanism is supported by observed pH drops during adsorp- tion. Raju et al. [1998] extended the description of the reaction, using IR spectra of dextrin and metal(II) hydroxide products to demonstrate the stoichiometry of the adsorption reaction. Dextrin was shown to complex neutrally with the metal cation per Equation 2.5. R− (OH)2 +Me(OH)2↔ R−O2Me+2H2O (2.5) 2.3.3 Wood preservers Naphthenic acids extracted from crude or refined oils are used in several markets. Copper and zinc naphthenates are primary used as wood preservatives [Brient et al., 2000]. Other metal naphthenate salts are also used in the paint industry for drying catalysts - however, this use is in decline. In wood preserving, interest in defining the preservative compound has resulted in insightful research using similar methods as presented here. Craciun and Kam- dem [1997] used XPS analysis to observe the oxidation state of the copper in a copper naphthenate wood preserver, concluding that the cation is Cu(II). The same study also performed FTIR spectral analysis to observe changes in the con- stituent compounds of the wood preserver (Cu(OH)2 and naphthenic acid). The 17 IR spectrographs showed that the Cu(OH)2 lost its hydroxyl groups and the naph- thenic acids lost the proton on the carboxyl group upon mixing together. Likewise, the characteristic band associated with the carboxyl group on the naphthenic acid (wavenumber at 1700cm−1) shifted to a band associated with metal-carboxylates (1600cm−1). These observations confirmed that the copper interacts with the naph- thenate as a ligand, forming a co-ordination complex. Similar conclusions have been made in other studies [Zyskowski and Kamdem, 1999], although the Cu- naphthenate observed was extracted from a stained wood, and may have other metallo-organic compounds appearing in the spectra. 18 Chapter 3 Research Hypotheses The research presented here attempts to elucidate the reaction mechanisms be- tween naphthenic acids and inorganic minerals. From the literature review, one can anticipate several outcomes of the experiments (described in Chapter 4). The experimental matrix is shown in Table 3.1 to clarify the hypotheses. 3.1 Low pH Reactions Owing to the pKa of the surrogate naphthenic acids, low pH (pH ∼ 3) solutions of these materials is unlikely to result in any coordination complexation between the acids and the minerals. This is due to the re-association of the acid into a non-polar hydrophobic substance of low aqueous solubility, along with the non- availability of the carboxylate anion or the unhydroxylated copper sulphide surface. This may result in some hydrophobic adhesion to the mineral surface, but little or no formation of metal-ligand complexes; a more likely outcome is that the organics will simply form a third phase in the mixing vessel. 3.2 High pH Reactions High pH (pH∼ 8.5) aqueous solutions will yield high mineral uptake of the organic acids. This is due to the aqueous solubility of the carboxylic acids at pHs above the pKa, and to the availability of the carboxylate anion. 19 Table 3.1: Experimental matrix, short form Organic Acid Mineral pH Method Molar Ratios Decanoic Acid (DA), Cyclohexane Pentanoic Acid (CHPA) Copper Sulphide (CuS) 3 Titration 0 (blank), 0.01:1, 0.05:1, 0.1:1, 0.25:1 Decanoic Acid (DA), Cyclohexane Pentanoic Acid (CHPA) Copper Sulphide (CuS) 8.5 Titration 0 (blank), 0.01:1, 0.05:1, 0.1:1, 0.25:1 Decanoic Acid (DA), Cyclohexane Pentanoic Acid (CHPA) Copper Hydroxide (Cu(OH)2) 8.5 Titration 0 (blank), 0.01:1, 0.05:1, 0.1:1, 0.25:1 Decanoic Acid (DA), Cyclohexane Pentanoic Acid (CHPA) Copper Hydroxide (Cu(OH)2) 8.5 Co-Precipitation 0 (blank), 0.01:1, 0.05:1, 0.1:1, 0.25:1 3.2.1 Copper sulphide at high pH From the literature review, the point of maximum hydroxylation of the copper sul- phide mineral is approximately pH8.5. At this pH, the mineral behaves like a sul- phide/hydroxide hybrid and the surface has the most potentially active adsorption sites for the carboxylate anions in solution. Therefore, coordination complexes between the organic acid and the metal contained in the mineral are expected to form. However, this uptake will be limited by the number of hydroxide sites on the surface. 3.2.2 Copper hydroxide titration In these experiments, the copper hydroxide is formed from a copper sulphate so- lution with caustic soda additions to pH 8.5. The naphthenic acid surrogates are added after the mineral formation. Copper hydroxide has maximum formation at approximately pH 8.5. Under these conditions, the copper hydroxide is anticipated to take up more naphthenic acid surrogate than copper sulphide, due to the fact that the mineral is not a sulphide-hydroxide hybrid; the only anions competing for the metal centres are the hydroxides and naphthenates. While the extent of this competition is beyond the scope of this thesis, one can be certain that: 20 1. The hydroxides in a metal hydroxide mineral are less tightly bound to the metal than the sulphides in a metal sulphide mineral 2. There are more hydroxides in/on a metal hydroxide mineral than there are hydroxides on the surface of a metal sulphide mineral Therefore, it is anticipated that the copper hydroxide mixed with the surrogate naphthenic acids will uptake more organic than the copper sulphide in equivalent aqueous conditions. 3.2.3 Copper hydroxide co-precipitation Co-precipitation of the copper hydroxide/naphthenate species is performed by adding the naphthenic acids to the copper sulphate solution at native pH (typically pH 4.4), followed by pH adjustment up to pH 8.5. Of the four experiment types listed here, the co-precipitation is expected to remove the most organic from solution. This is due to the increased extent of copper cation and carboxylate anion mixing prior to pH adjustment. This means that the reaction is no longer a surface limited compe- tition for metal centres in a solid structure; rather, the reaction is purely aqueous precipitation. 3.2.4 Coordination complexes The use of infrared spectroscopy is used to determine the final structure of the re- sultant metallorganic complexes in all these experiments. While the low pH mixing is expected to result in little or no complexation, the high pH experiments should all result in some copper-carboxylate formation. While there are two possible co- ordination complex structures (chelating and bridging, as noted in Figure 2.2), ac- cording to Mehrotra and Bohra [1983], the chelating is a less likely outcome due to the higher energy of the pinched O-C-O bond angle. Therefore, bridging com- plexes are the expected outcome from the IR spectroscopy. These can be observed by calculating the differences in the symmetric and antisymmetric C-O stretching bands, again outlined in Chapter 2 21 3.2.5 Stoichiometry of reaction The anticipated stoichiometry of reaction is based on the oxidation states of the two reacting compounds. The carboxylate anion has a valence of negative one while the copper cation used is in the positive two state. The expected reaction is: 2(R−COO−)+Cu2+↔Cu(R−COO)2 (3.1) Note that the coordination complexes shown in Figure 2.2 confirm this sto- ichiometry for either bridging or chelating structure. hTe reaction is effectively universal for each of the experiment types - in the case of sulphide mineral, the copper cation will become surface available upon release of the surface hydroxide anions. Likewise, copper hydroxide will expose the bare cation upon exchanging the hydroxide for the carboxylate. 22 Chapter 4 Materials and Methods The experimental portion of the research has two main objectives: 1. the quantitative adsorption equilibria of surrogate naphthenic acids onto in- organic minerals; and 2. the semi-quantitative observations of surface group chemistry in pre- and post-reacted organic and mineral mixtures. 4.1 Adsorptive Equilibria The adsorptive equilibria experiments involve mixing the naphthenic acid surro- gates with the mineral solids in a fixed total solution volume. The pH is adjusted after the addition of the materials to the flask, and the slurry is allowed to mix for 24 - 36 hours to equilibrate. 4.1.1 Minerals The metal sulphide minerals were synthesized in-house prior to use in the ad- sorption experiments. Copper sulphate, anhydrous (98.4% purity CuSO4, from Fisher Scientific Fairlawn, NJ) is dissolved in deionized water to a concentration of 4774mg/L. Sodium hydrosulphide (NaHS-xH2O, Acros Organics, Geel, Belgium) is likewise dissolved in deionized water to a concentration of 10g/L. Portions of the two solutions (no fixed quantity) are mixed together in a beaker in a fumehood, due 23 to evolution of hydrogen sulphide gas during the reaction. The resultant precipitate is copper sulphide (CuS). This solid mineral is transferred to a qualitative filter pa- per (Whatman International Ltd. Maidstone, England), and rinsed multiple times during filtration. The rinsing is to remove residual NaHS or CuSO4 salts from the slurry. The final rinsed, filtered product is placed in an oven (SHEL LAB Model Number 1327F, Cornelius, OR) at 60◦C for 24 hours. Once the mineral is dried, it is transferred to an anaerobic box (Coy automatic airlock and sealed chamber, Model AALO, Coy Laboratory Products, Grass Lake, MI) for longer term storage. The anaerobic box is filled with nitrogen (85% N2), carbon dioxide (10% CO2) and hydrogen (5% H2) gasses (Praxair Canada, Mississauga, ON). The use of the anaerobic storage minimizes the oxidation of the CuS back to CuSO4, although some oxidation of this type is inevitable and noticeable during mixing of CuS with water and during IR spectroscopy. 4.1.2 Surrogate naphthenic acids As NAs are a mixture of many compounds, creating a molecular model of inter- action by experimenting with native or commercial samples is unwieldy. Instead, this research uses model NA compounds. Two surrogates were selected; decanoic acid (DA, 98% purity, Sigma Aldrich, St. Louis, MO) and cyclohexanepentanoic acid (CHPA, 98% purity, Sigma Aldrich, St. Louis, MO). These were chosen due to their aliphatic structure, similar masses, but difference in ring number (zero and one, respectively). They are also readily available from chemical vendors. Syn- thesis of other surrogates, although possible [Smith et al., 2008], is beyond the scope of this work. Both organic acids were used as-received from the chemical suppliers. Decanoic acid solution is prepared to a 3390mg/L standard. To achieve this, a 1000ml volumetric flask is half-filled with deionized water and 3390mg of DA are added to this water. The flask is then placed in a water bath (VWR Scientific Model number 1202, Sheldon Manufacturing Inc, Cornelius, OR) at 30◦C. The elevated temperature helps to melt the DA. Once the DA has melted, approximately 25ml of 40g/L sodium hydroxide (NaOH, Acros Organics, Geel, Belgium) solution are added to the volumetric flask. This caustic is sufficient to neutralize the acid groups 24 Table 4.1: Properties of the selected naphthenic acid surrogates (from manu- facturer’s MSDS) Property Decanoic Acid Cyclohexane Pentanoic Acid Formula C10H20O2 C11H20O2 Structure O OH O OH Molecular Mass (g/mol) 172.27 184.28 Specific Gravity 0.890 0.960 Melting Point (◦C) 30 - 32 16 - 17 pKa* 4.90 N/A Solubility** (g/L) 0.15 N/A *While there is no readily available pKa for CHPA, the value is expected to be the same as for DA **In water. Note that water solubility is a function of pH. The native associated acids are effectively water insoluble on the DA and raise the pH to between 10 and 12. At this elevated pH, the DA dis- sociates, becomes ionic and can dissolve in water. Upon final dissolution of the DA into the high-pH water, the solution become clear with shaking, instead of cloudy due to a DA-water emulsion. At this point, the volumetric flask is removed from the hot-water bath and the solution is allowed to return to room temperature prior to topping up to the score mark. The final solution is effectively sodium decanoate, although the term DA is used throughout this work to describe the material. Cyclohexane pentanoic acid solution is prepared to a 3624mg/L standard. To achieve this, a 1000ml volumetric flask is half-filled with deionized water and 3624mg of CHPA are added to this water. CHPA is a liquid at room temperature, so no heating is required. Approximately 25ml of 40g/L sodium hydroxide (NaOH, Acros Organics, Geel, Belgium) solution are added to the volumetric flask. The caustic is sufficient to neutralize the acid groups and raise the solution pH above neutral. The final solution pH is between 10 and 12. Upon dissolution of the CHPA into the high pH water, the solution becomes clear and the oily emulsion is no longer visible. The volumetric flask is then topped up to the mark. The so- lution is effectively sodium cyclohexane pentanoate, but the term CHPA is used 25 throughout this work to describe the material. 4.1.3 Adsorption equilibria The adsorption experiments are performed over several days. There are three types of experiments, and their respective steps are listed here: 1. Metal sulphide adsorption - 50ml of deionized water is added to a clean 250ml Erlenmeyer flask. To this water, 0.1g of copper contained in the CuS mineral is added. The DA or CHPA solution is then added to meet the desired acid-to-metal molar ratio. 2. Metal hydroxide adsorption - 50ml of deionized water is added to a clean 250ml Erlenmeyer flask. To this water, 0.1g of copper contained in a 44g/L copper sulphate solution is added. Then caustic soda solution is added un- til pH 8.5 is reached. This precipitates the copper as copper hydroxide (Cu(OH)2). Then the DA or CHPA solution is added to the target molar ratio. 3. Metal hydroxide co-precipitation - 50ml of deionized water is added to a clean 250ml Erlenmeyer flask. To this water, 0.1g of copper contained in a 44g/L copper sulphate solution is added. Then the DA or CHPA solution is added to the target molar ratio. Due to the high pH of the DA and CHPA solutions, the sample pH rises slightly from the native pH of the copper sul- phate solution (approximately pH 4 - 4.5). Then caustic soda solution is added until a pH 8.5 is reached. This precipitates the copper as copper hy- droxide (Cu(OH)2) while the metal cations and surrogate naphthenate anions are intimately mixed. For these experiments, the range of acid to copper molar ratios start at 0.01 and reach a high ratio of 0.25. This ensures that copper is always in excess given the expected 2:1 stoichiometric ratio. After the addition of the metal and organic acid, more deionized water is added to the 250ml flask to bring the total solution volume up to approximately 120ml. At this point, the pH is adjusted to within 0.5pH units the experimental target using 26 NaOH and H2SO4. After pH adjustment, the flask is topped up to the 150ml mark. The pH is not re-checked or adjusted after the final water addition. Mineral blank samples are prepared concurrently with the experimental sam- ples. These are identical to the aforementioned experiments, but without any min- eral or CuSO4 solution added. The purpose of the mineral blank is to evaluate the DA and CHPA content of the solution in the absence of the mineral. This accounts for DA and CHPA losses to the flask surfaces, volatilization, light and bio-degradation. The mineral blank is used to compute the partition from water phase to the mineral surface. Acid blanks are also performed. These are set up with only the mineral and CuSO4 solution; no organic is added. These are used to evaluate the background organic carbon content and to provide baseline minerals for the IR spectroscopy. The sample flasks are then covered with Parafilm (Pechiney Plastic Packaging, Menasha, WI) and placed on a incubator shaker (Innova 4000, New Brunswick Scientific, Edison, NJ) for 24 - 36 hours. The shake table is operated at 200RPM and has an internal heater controlled to 25◦C. After the shaking, the sample pHs are measured to observe any drift during the experiment. The samples are removed from the shake table and prepared for organic carbon analysis. 4.1.4 Total organic carbon (TOC) The organic acid content of the experimental samples and mineral blanks must be evaluated in order to determine the partition from the water phase to the min- eral surface. Measuring the concentrations of these model compounds was not performed using any literature methods described in Chapter 2. GC-MS is unsat- isfactory due to its inability to measure absolute concentrations [Holowenko et al., 2002]. Because these experiments use only specific model compounds, the speci- ation of a mixed sample is not relevant. While HPLC can give absolute values of concentration, it is known to have variations in the measurements [Clemente et al., 2003]. Even following the improved methods developed by Yen et al. [2004], this inaccuracy is present. The industry standard FTIR method offers higher accuracy in samples that have 27 no competing organic acids. However the preparatory steps are tedious and numer- ous, leading to compounded errors. Although not mentioned in the literature, Total Organic Carbon analysis (TOC) offers a simple, rapid and accurate evaluation of the selected surrogate NAs in solu- tions. Because there are no competing organics, the TOC exclusively accounts for the surrogate NA in a given solution. As well, the carbon fraction of both DA and CHPA are known, allowing the calculation of total DA and CHPA concentration from the TOC value. After the experiments are complete, the samples must be prepared for TOC analysis. The samples are first diluted to meet the analytical upper limit of 20mg/L carbon content. Depending on the dilution requirement for each experiment, either a 1ml hand-held auto-pipette or a 10ml glass pipette is used to transfer solutions up to 1ml or up to 10ml, respectively. These sample volumes are transferred to a 100ml volumetric flask, which is topped up with deionized water. Following di- lution, a 100ml beaker is rinsed with a small volume of the diluted solution, and the rinsate is discarded. The remaining diluted solution is placed in the beaker. From these beakers, 30ml of solution is drawn into a 30ml capacity polypropy- lene syringe (BD 30ml Luer-Lok tip syringe, Franklin Lakes, NJ). 15ml of diluted solution is then passed through a syringe filter (VWR International, 25mm filter diameter with 0.22µm pore size nylon membrane) into a brown glass TOC vial. The vial is rinsed with the first solution and the rinsate is discarded. The remaining 15ml is filtered into the vial. The vial is then covered with Parafilm and placed in the refrigerator for storage prior to TOC analysis. The TOC analyzer used in these experiments is a Shimadzu TOC-V CPH ana- lyzer with an ASI-V autosampler, operated by Shimadzu TOC Control-V software Version 1.05.01 (Shimadzu, Kyoto, Japan). This instrument uses a combustion catalytic oxidation method with Ultra-Zero compressed air (Praxair Canada, Mis- sissauga ON) as carrier gas. The carrier gas flowrate is 150ml/min. Samples are acidified using 7% HCl. The CO2 detection is by non-dispersive infrared gas an- alyzer. The instrument performs an internal calibration after every 5 samples; this calibration is checked periodically (either separately from the sample runs, or as an inserted standard sample) using a standard solution of potassium hydrogen phtha- late (KHP). Deionized water blanks are introduced into the series every 4 samples 28 Table 4.2: Shimadzu TOC-V CPH instrument settings Setting Name Option Oxidation Combustion Syringe Type 5000 µl Catalyst Type Normal Tube Diameter Regular Syringe Wash Volume Standard Cell Length Long Furnace Temp 680◦C UV Lamp Off Sparge Kit No IC Unit Yes Auto IC Regeneration Yes ESU No ASI Yes Tray Type 40 ml Vials Needle Type Sample Number of Needle Washes 2 Number of Flow Line Washes 2 Rinse Yes Rinse after Acid Addition Yes Stirrer No Action on Standby Shut Down Instrument Acid Addition External to check the instrument zero. The instrument control option settings are listed in Table 4.2. 4.2 Solid Surfaces The second portion of the experimental work is examination of the mineral surfaces before and after adsorption of the organic acids. Infrared spectroscopy is used to observe these surfaces. For the purposes of this research, Diffuse Reflectance Fourier-Transform Infrared (DRIFT) spectroscopy was used, instead of traditional Fourier-Transform IR (FTIR) spectroscopy. The DRIFT spectroscopy relies on the reflection of IR light off of the sample surface into the detector, instead of 29 the FITR in which the IR light passes through a compressed sample pellet. The DRIFT method has the advantages that it is quicker and has less sensitive sample preparation. After the experiments have completed mixing and the TOC samples are pre- pared and loaded, the solids are filtered by a qualitative filter paper (Whatman In- ternational Ltd. Maidstone, England). The resultant solids are rinsed several times with deionized water in order to completely remove any residual salts or ions, such at sulphate and sodium. These washed and filtered solids are then labelled and placed in an oven (SHEL LAB Model Number 1327F, Cornelius, OR) at 60◦C for 24 hours. Once the mineral is dried, it is transferred to an anaerobic box (Coy automatic airlock and sealed chamber, Model AALO, Coy Laboratory Products, Grass Lake, MI) for longer term storage. The anaerobic box is filled with nitrogen (85% N2), carbon dioxide (10% CO2) and hydrogen (5% H2) gasses. The use of the anaerobic storage minimizes sample oxidation, as some observations during this work revealed product oxidation. Due to the optical qualities of the minerals used, they must be diluted with an IR transparent material in order to provide any light reflection in the DRIFT analysis. The samples in this research were all diluted with potassium bromide (KBr, Fisher Scientific, Fairlawn, NJ) prior to analysis. The KBr is dried in an oven at 60◦C at all times to reduce the moisture content in the salt. The dilution ratio is approximately 1 - 5mg of mineral sample in 200 - 230mg KBr. This mix is then ground and blended in a quartz mortar and pestle to a homogeneous dry powder - no particle size distribution is performed, as it is not required for this analysis. The blend is transferred to the DRIFT unit sample cup (a holder approximately 5mm diameter by 3mm deep), tamped down with a stainless steel spatula and the final surface of the sample cup is smoothed off. The infrared spectroscopy instrument is a PerkinElmer Spectrum 100 FTIR Spectrometer, using the Diffuse Reflectance Sampling Accessory and operated by Spectrum software, Ver.10.00.00 (PerkinElmer Life and Analytical Sciences, Bea- consfield, UK). The resolution is set to 4cm−1 and each sample runs 100 scans. The wavenumber range is set from 4000cm−1 to 400cm−1, although most of the range below 1400cm−1 is not used. The DRIFT accessory requires calibration against the KBr blank prior to sample analysis. The plain KBr is ground in the quartz 30 mortar and pestle, to the same uniform powder as the diluted samples. This ground KBr is also tamped into the sample cup, smoothed and inserted into the DRIFT unit. Before running the background scans, the background energy is monitored and maximized by the ”Nudge Down” operation - usually, nudging the sample platform to the bottom of the range results in a KBr background energy of between 95 - 105 units. Maximizing this background energy increases the resolution of the spectra. Following the background energy monitor, the control software calibrates to the pure KBr. Following the calibration, the KBr pellet is run as a sample to confirm the background. The expected transmission of the IR light is 100% ± 1% across the wavenumber range. If this KBr-as-sample flat transmission is not achieved (usually due to high humidity causing a large water band in the sample), the calibration is run again, or the pure KBr sample is discarded and replaced with drier material. Once satisfactory calibration and background scan are completed, the diluted mineral samples are scanned. As with the KBr blank, the sample background ener- gies are monitored and adjusted via platform elevation. Diluting the samples with KBr raises this background energy from nil (for pure black CuS mineral) to 35 - 50 units. The platform nudge up and down are used to trim the background energy to 40 units, giving each DRIFT spectrograph similar resolution. The diluted samples are then scanned, using the same platform elevation and scan settings as outlined above. Between each sample dilution, grinding and homogenization with KBr, the weigh boats, mortar and pestle, spatula and sample holder cups are rinsed with wa- ter and wiped completely dry with Kimwipes. This mitigates cross contamination of samples between scans. 31 Chapter 5 Results and Discussion The results of this research demonstrate that the surface phenomenon between inor- ganic minerals and model naphthenic acids are mixed between some hydrophobic attachment (primarily in the case of low pH experiments) and metallic coordination complexes (in the case of high pH experiments). 5.1 Adsorption Results The adsorption results are shown in Figure 5.1. The figures show the results of the adsorptive experiments between copper minerals and DA and CHPA, respectively. The abscissa shows the final concentration of DA or CHPA after the adsorption equilibria is complete. The ordinate axis shows the computed loading of DA or CHPA onto the mineral. This value is computed by taking the difference between the mineral blank acid concentration less the final equilibrium acid concentration, then normalizing the mass of acid per mass of metal contained in the mineral; this is shown in Equation 5.1. q= (Bm−Ce)V MCu (5.1) Here, q is the mineral uptake of organic compound (in mg organic/g Cu), Bm is the mineral blank final acid concentration of the corresponding experiment (mg organic/L), Ce is the equilibrium acid concentration of the experimental run (mg organic/L), V is the total experimental volume (150ml in all cases) and MCu is the 32 20 40 60 80 100 120 140 -50 50 100 150 200 250 300 CuS – low pH, titration CuS – high pH, titration Cu(OH)2 – titration Cu(OH)2 – co-precipitation Equilibrium Acid Concentration (mg/L) A ci d up ta ke  (m g ac id /g  C u) (a) DA Isotherm 50 100 150 200 250 300 -50 50 100 150 200 250 300 CuS – low pH, titration CuS – high pH, titration Cu(OH)2 – titration Cu(OH)2 – co-precipitation Equilibrium Acid Concentration (mg/L) A ci d up ta ke  (m g Ac id /g  C u) (b) CHPA Isotherm Figure 5.1: Naphthenic acid surrogate mineral loading as equilibrium acid concentration 33 mass of copper contained in the mineral sample (0.1g Cu in all cases). 5.1.1 Low pH adsorption Mixing DA and CHPA with CuS at pH 3 resulted in little remaining organic in the resultant solution. This is due primarily to the re-association of the carboxylic acid group below the acid pKa of 4.9. Once the acids re-associate, they are no longer soluble or miscible in water, and the solutions becomes oily. This was verified by observations during the experiments, in which the water surfaces showed an oily film and flecks of solid DA or globules of immiscible CHPA. Thus the surrogate naphthenic acids simply leave the water phase and do not react chemically with the minerals. This is confirmed in the IR spectral results, which show little com- plexation between the metal and the organic. Although there is some hydrophobic sorption to the mineral, these results cannot quantify the three-way partition be- tween water, mineral surface and re-associated organic phases. 5.1.2 High pH adsorption and coprecipitation Mixing DA and CHPA with CuS at pH 8.5 resulted in increased partition of the organic to the mineral surface phase. This is evidenced by the data shift from the water-phase axis to the mineral-phase axis. This, along with the IR spectroscopy data presented later, suggests that at this pH the copper at the CuS surface can be accessed by the carboxyl group on the DA and CHPA, leading to the complexa- tion of the organic with the copper. This is due to the CuS surface being highly hydroxylated at this pH [Liu and Laskowski, 1989]. These hydroxide groups can be replaced by the dissociated carboxylate anion, which complexes with the naked copper under the displaced hydroxide. This complexation is demonstrated in the IR spectrographs, to be discussed later. Mixing DA and CHPA with Cu(OH)2 shows a decrease in adsorption of the organic to the mineral phase. This is demonstrated by the move of the organic away from the mineral-phase axis and more towards the water-phase axis (relative to the pH 8.5 CuS isotherm). This decrease in adsorption is unexpected - per the hypotheses in Chapter 3, the expectation was that more acid would be able to contact and react with the copper contained in the Cu(OH)2 mineral. 34 Table 5.1: Summary of Langmuir isotherm fitting parameters for DA and CHPA at high pH Experiments KL (mg organic/g Cu) k (L/mg organic) R2 coefficient Decanoic Acid CuS, titration -72.46 -0.03507 0.9081 Cu(OH)2, titration -19.92 -0.00979 0.997 Cu(OH)2, co-precip -4.59 -0.01257 0.961 Cyclohexane Pentanoic Acid CuS, titration -131.58 -0.00419 0.999 Cu(OH)2, titration 4.05 -0.01319 0.952 Cu(OH)2, co-precip 4.34 -0.01693 0.948 Likewise, DA and CHPA co-precipitated with copper to make copper hydrox- ide/copper carboxylate solids show the lowest organic removal from the water. This is again unexpected and contrary to the hypotheses in Chapter 3. Regardless of expectations, surface adsorption models can be applied to these figures to determine the fitting constants of linear, Langmuir or Freundlich isotherms, as described in Chapter 2. Only the high pH experiments were fit, as the low pH experiments yielded little actual adsorption. Of the three potential models, the Langmuir isotherm fit the data best and the constants and R2 values are shown in Table 5.1. Note that while the R2 values appear to be high, indicating a good fit of data to model, due to certain experimental limitations described below, these values should not be taken as a universal description of the behaviour of these interactions. Curiously, when the decanoic acid to copper ratios are greatly increased, the trend in increasing organic acid adsorption is reversed, per Figure 5.2 (these high range experiments were only performed with DA, not with CHPA). This matches the hypothesis that pre-mixing the carboxylate anion and the metallic cations be- fore precipitation improves the ability of the two materials to form stable com- plexes. However, this leads to the conclusion that the nature of the interactivity between the carboxylates and metals change with increasing acid to metal ratios. This is corroborated by the fact that none of the linear, Langmuir or Freundlich isotherm models fit these data with any quality (R2 < 0.75 in all cases, data not 35 500 1000 1500 2000 2500 -500 500 1000 1500 2000 2500 3000 3500 CuS – low pH, titration CuS – high pH, titration Cu(OH)2 – titration Cu(OH)2 – co-precipitation Equilibrium acid concentration (mg/L) A ci d up ta ke  (m g ac id /g  C u) Figure 5.2: Decanoic acid mineral loading as equilibrium acid concentration - high acid:metal ratios shown). The explanation for this phenomenon is the impact of pH on the metal and organic species’ behaviour. This is described in the following section. One observation during the experiments is that the organic acids are not limited to partition between water and mineral surfaces; instead, there is a three or four- way partition of the organics between the water phase, the mineral surface, the flask surface and, at the higher acid concentrations, re-associated organic phase. Experimental observations confirm that, in some situations, the mixing flasks had visible oily slicks on the water surface and up the flask walls, even in the case of the high pH experiments. Note that these particles and slicks are true organic phase, not micelles; Shinoda [1954] showed that the critical micelle concentration (CMC) for potassium decanoate in water is approximately 0.1mole/L. The maximum de- canoate concentration for all experiments shown is at least an order of magnitude 36 500 1000 1500 2000 2500 3000 -500 500 1000 1500 2000 2500 3000 3500 CuS – low pH, titration CuS – high pH, titration Cu(OH)2 – titration Cu(OH)2 – co-precipitation Initial Acid Concentration (mg/L) Ac id  u pt ak e (m g ac id /g  C u) Figure 5.3: Decanoic acid loading as function of initial acid concentrations lower than this value. Because of this multiple partition effect, the above Figure 5.1 may not accu- rately represent the mineral adsorption effect with respect to the acid equilibrium concentrations. While the acid loading is computed using the difference between experimental equilibrium acid concentration and the mineral blank acid concen- tration (ie: to account for the organic losses to the alternative sinks), the equilib- rium concentration as independent variable is unsuitable as it does not represent the intended single equilibria between water phase and mineral surface. However, plotting the acid loading with respect to the mineral blank acid concentrations (ie: the initial acid concentration after accounting for acid losses to alternative sinks) confirms the hypothesis that the acid adsorption will increase from CuS through to copper hydroxide co-precipitation. This is shown in Figure 5.3. In this figure, all DA:Cu ratios are shown. While the low DA:Cu ratios still demonstrate that acid 37 Table 5.2: Summary of Langmuir isotherm fitting parameters for DA uptake with respect to initial acid concentration Experiments KL (mg organic/g Cu) k (L/mg organic) R2 coefficient Decanoic Acid CuS, titration -28782.8 -0.000034 0.994 Cu(OH)2, titration -67.11 -0.002772 0.993 Cu(OH)2, co-precip -13.83 -0.004445 0.967 adsorption to CuS at high pH is greater than the copper hydroxide adsorptions, the differences are less pronounced and quickly corrected after the CuS saturates and plateaus. Likewise, the two copper hydroxide isotherms show identical ad- sorption until the Cu(OH)2 mineral saturates and plateaus. The co-precipitation of decanoate with copper does not show this plateau as the data stops at the expected inflection point of 2:1 stoichiometric acid to metal ratio. Strictly speaking, the use of initial acid concentration as the independent vari- able in these isotherms means that the linear, Langmuir and Freundlich surface ad- sorption models shown in Chapter 2 should longer apply. In spite of this, determi- nation of the fitting parameters for each isotherm model revealed that the Langmuir model again found the best fit; the parameters are summarized in Table 5.2. Again, while these regressed parameters are encouraging with apparent high-quality fits, they are not truly representative of the intended adsorption phenomena. This is due to limitations both theoretical (due to use of initial concentration instead of equi- librium concentration) and practical (as described above, due to multiple partition avenues). 5.1.3 Adsorption effects In addition to the adsorption data generated and shown above, there were several unexpected findings during this research. Effect of pH changes during reaction During the adsorption experiments, the pH of the solution is adjusted initially to within 0.5 pH units of target and the value is recorded. After the mixing is com- 38 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 -50 0 50 100 150 200 250 300 350 400 450 0.01:1 DA:Cu 0.05:1 DA:Cu 0.1:1 DA:Cu 0.25:1 DA:Cu pH change during reaction Ac id  p ar tit io n (m g/ L) (a) DA Partition -5 -4 -3 -2 -1 0 1 -20 0 20 40 60 80 100 120 140 160 180 0.01:1 CHPA:Cu 0.05:1 CHPA:Cu 0.1:1 CHPA:Cu 0.25:1 CHPA:Cu pH change during reaction Ac id  p ar tit io n (m g/ L) (b) CHPA Partition Figure 5.4: Naphthenic acid surrogate partitions from water phase as a func- tion of experimental pH change 39 pleted after 24 - 36 hours, the final pH is measured. Invariably the pH shifts down (in the case of the high pH experiments) or very slightly up (in the case of the low pH experiments). Plotting the acid partition (the difference between initial and fi- nal acid concentration) as a function of the pH change during the mixing reveals an almost linear relationship between the two parameters, as shown in Figure 5.4. Note that the partition shown is exclusively a function of the pH change; there is no particular ordering in the data of the experiment type, although the high pH CuS ad- sorption experiments tended to have the highest pH changes, and the origin where the lines intersect is the low pH CuS experimental data (which had categorically low partition over the course of the experiments, and low pH changes). These data show that, in these acid:copper ratios, the pH of the solution has an overwhelming effect on the water/solid partition of the organic materials. Curiously, at higher acid:copper ratios (from 0.5:1 - 2:1) such a trend is absent. Operating mineral and acid blanks, respectively, reveals that both substances, absent the other, when shaken under the same conditions as the experiments, re- alize pH drops spontaneously. In the blank mineral experiment, the high pH CuS dropped over 3 pH units, from 8.79 to 5.31; in the case of the Cu(OH)2 blanks, the pH dropped up to 2 pH units. The exact cause of this drop is unclear, although it may be due to the lack of buffering in the mineral solution and the tendency of soluble copper cations to have a native pH that is slightly acidic - this may be ex- acerbated in the case of the CuS experiments as the re-solublized Cu2+ ions will consume hydroxide groups, thus realizing larger pH changes than the copper hy- droxide experiments. With the organic acids, the buffering capacity of the acid/conjugate sodium salt system will tend to push the pH towards the pKa of the weak acid, which is approximately 4.9. In general, the pH drops of both DA and CHPA mineral blank solutions (starting at pH 8.5) were between 1 and 2 pH units, and, in general the pH drops were higher with the lower acid concentrations. This suggests that the extra NaOH or HCl used to adjust the pH in the higher concentration acid solutions lend some counter-buffering effect. An alternative perspective on the effect of pH is shown in Figure 5.5. Here, one can observe that the percentage uptake of the organic to the mineral (relative to initial organic concentration) is again a function of pH. However, in this figure, 40 2 3 4 5 6 7 8 9 -40% -20% % 20% 40% 60% 80% 100% 120% 0.01:1 DA:Cu 0.05:1 DA:Cu 0.1:1 DA:Cu 0.25:1 DA:Cu Final pH Fr ac tio na l u pt ak e (% ) (a) DA Uptake 2 3 4 5 6 7 8 9 10 -60% -40% -20% % 20% 40% 60% 80% 100% 0.01:1 CHPA:Cu 0.05:1 CHPA:Cu 0.1:1 CHPA:Cu 0.25:1 CHPA:Cu Final pH Fr ac tio na l u pt ak e (% ) (b) CHPA Uptake Figure 5.5: Naphthenic acid surrogate fractional uptake to mineral as a func- tion of final solution pH 41 2 3 4 5 6 7 8 9 10 11 000.0E-2 500.0E-5 100.0E-4 150.0E-4 200.0E-4 250.0E-4 0.1:1 DA:Cu 0.5:1 DA:Cu 1:1 DA:Cu 2:1 DA:Cu [Cu2+] (mol/L) pH C on ce nt ra tio n (m ol /L ) Figure 5.6: Solubility of copper cation and decanoate anion species with pH it is evident that the point of maximum uptake is at some intermediate pH between to the two selected experimental conditions. This can be explained with respect to both reagents. In the case of the organic compounds, the drops in pH push the carboxylic acids towards re-association, in which state the acids are insoluble in water. This insolubility will then tend to either force the acid onto the mineral surface, or the surface of the flask - the exact partition cannot be determined; how- ever the end result is that the mineral loading appears higher in these cases. In the case of the copper mineral, depression of the pH will result in more solubliza- tion or availability of copper cations, which can then react more easily with the decanoate anions. Thus the intermediate pH may actually represent the optimal reaction conditions between the organic acids and the copper minerals. This the- oretical proposition is shown in Figure 5.6. Note that the calculation of copper cation solubility is greatly simplified, using only a single Ksp value for Cu(OH)2 42 and neglecting the formation of ionic copper hydroxide species. The purpose of Figure 5.6 is to show that, while the hypothesis of adsorption is predicated on an- ionic carboxylate groups replacing hydroxides on the mineral surfaces, the mineral uptake may be improved in the presence of naked copper cations reaction with the carboxylates. Verily, Mauchauffee and Meux [2007] used a solution pH of 4.7 dur- ing their experiments for recovery of soluble copper using sodium decanoate. This pH was selected to optimize the formation of the copper decanoate species instead of a mixed precipitation of copper decanoate and copper hydroxide. Effect of molecular structure An important observation from Figure 5.1 is that, in all experimental cases, the CHPA has a higher equilibrium acid concentration than the DA. Likewise, from Figure 5.4 , one can see that CHPA always shows lower partition out of the water phase than the DA. This implies that the CHPA and its sodium salt are more hy- drophilic than the DA and its sodium salt. This is likely related to the molecular structure of the two compounds; the cyclic nature of the CHPA results in lower molecular surface area of the non-polar moiety than the straight chain equivalent portion of the DA. The correlation between molecular surface area and hydropho- bic Gibb’s energies has been studied [SanchezRuiz, 1995, Tunon et al., 1992], al- though the quantification of this effect is beyond the scope of this thesis. In this sit- uation, molecular structural differences appear to be sufficient to result in doubling of the partition away from the water phase. Considering that in situ naphthenic acids are a mixture of compounds that are both larger than the present surrogates (more hydrophobic) and typically multi-cyclic (lower molecular surface area per compound mass, thus more hydrophilic), this observation may not be helpful for modelling of real NA processes. 5.2 Infrared Spectroscopy Infrared spectroscopy (IR spec) was used to determine the nature of the resultant mineral-organic coordination complexes. 43 Table 5.3: Summary of key spectral bands for CuS mixed with DA and CHPA at low pH Acid:Cu Ratio Wavenumber (cm−1) Chemical Group and Motion 0:1 (Pure CuS) 3450 O-H stretching on water of hydration 0:1 (Pure CuS) 1610 (up) unknown, likely related to water 1:0 (Pure DA) 3400 - 2500 O-H stretching of carboxyl group 1:0 (Pure DA) 2940 and 2860 C-H hydrocarbon stretching 1:0 (Pure DA) 1710 C=O stretching on free carboxyl 1:0 (Pure DA) 1470 and 1410 C-O stretching 0.25:1 (DA) 2940 and 2860 C-H stretching on hydrocarbon 0.25:1 (DA) 1710 C=O stretching on free carboxyl 0.25:1 (DA) 1470 and 1410 C-O stretching 1:0 (Pure CHPA) 3400 - 2500 O-H stretching of carboxyl group 1:0 (Pure CHPA) 2925 and 2850 C-H hydrocarbon stretching 1:0 (Pure CHPA) 1710 C=O stretching on free carboxyl 1:0 (Pure CHPA) 1450 and 1410 C-O stretching 0.25:1 (CHPA) 2925 and 2850 C-H stretching on hydrocarbon 0.25:1 (CHPA) 1710 C=O stretching on free carboxyl 0.25:1 (CHPA) 1580 antisymmetric C-O stretch 0.25:1 (CHPA) 1440 and 1400 symmetric C-O stretch 5.2.1 Copper sulphide at low pH The DRIFT spectra of CuS mixed with DA and CHPA at pH 3 are shown in Fig- ure 5.7. The different spectra represent the different molar ratios of copper metal to acid, as well as pure DA and CHPA for reference. The key bands are tabu- lated in Table 5.3 - for simplicity, only the reference and highest acid:copper ratio peaks are included and the reader is invited to follow the trend of the intermediate acid:copper ratio. The presence of water of hydration in the CuS indicates that water is persistent in CuS in spite of drying at modest temperature. The indication that little copper carboxylate complex is forming is the band at 1700cm−1 in both the DA and CHPA samples. When copper carboxylate forms, this band should shift from 1700 to 1600 cm−1 (or slightly less). This failure to 44 950 1100 1250 1400 1550 1700 1850 2000 2150 2300 2450 2600 2750 2900 3050 3200 3350 3500 3650 3800 3950 Pure CuS 0.05:1 Acid:Cu 0.25:1 Acid:Cu Pure Decanoic Acid Wavenumber (cm-1) % T (a) Decanoic acid 950 1100 1250 1400 1550 1700 1850 2000 2150 2300 2450 2600 2750 2900 3050 3200 3350 3500 3650 3800 3950 Pure CuS 0.05:1 Acid:Cu 0.25:1 Acid:Cu Pure CHPA Wavenumber (cm-1) % T (b) Cyclohexane pentanoic acid Figure 5.7: IR spectrograph of naphthenic acid surrogates mixed with CuS at low pH 45 Table 5.4: Summary of key spectral bands for CuS mixed with DA and CHPA at high pH Acid:Cu Ratio Wavenumber (cm−1) Chemical Group and Motion 1:0 (Na-DEC) 2920 and 2850 C-H hydrocarbon stretching 1:0 (Na-DEC) 1560 antisym C-O stretching 1:0 (Na-DEC) 1450 and 1440 symmetric C-O stretching 0.25:1 (DA) 2915 and 2845 C-H stretching on hydrocarbon 0.25:1 (DA) 1600 antisym. C-O stretch (bridging compound) 0.25:1 (DA) 1400 symmetric C-O stretching (bridging compound) 0.25:1 (DA) 1500 antisym. C-O stretch (chelating compound) 0.25:1 (DA) 1440 symmetric C-O stretching (chelating compound) 1:0 (Na-CHP) 2930 and 2840 C-H hydrocarbon stretching 1:0 (Na-CHP) 1570 antisym C-O stretching 1:0 (Na-CHP) 1440 symmetric C-O stretching 0.25:1 (CHPA) 2920 and 2830 C-H stretching on hydrocarbon 0.25:1 (CHPA) 1590 antisym. C-O stretch (bridging compound) 0.25:1 (CHPA) 1410 symmetric C-O stretching (bridging compound) shift indicates that undissociated, uncomplexed DA and CHPA exist on the surface of the mineral, likely sorbed to the CuS mineral hydrophobically. In the case of CHPA, one can see a band at 1580−1 as well. This suggests that some copper car- boxylate complex is forming. However, the band at 1700cm−1 is not completely displaced, indicating that both associated and complexed organic acid exist simul- taneously on the CuS. Because there is excess copper even at this high ratio, one must conclude that the complexation reaction is limited to that copper which is sur- face available and to carboxylic acid that is dissociated at this low pH. Both species are limited. 5.2.2 Copper sulphide at high pH The DRIFT spectra of CuS mixed with DA and CHPA at pH 8.5 are shown in Figure 5.8. The different spectra show the indicated acid to metal molar ratios. Pure sodium decanoate (Na-DEC) and sodium cyclohexane pentanoate (Na-CHP) is shown for reference. The key bands are tabulated in Table 5.4. 46 950 1100 1250 1400 1550 1700 1850 2000 2150 2300 2450 2600 2750 2900 3050 3200 3350 3500 3650 3800 3950 Pure CuS 0.05:1 Acid:Cu 0.25:1 Acid:Cu Pure Na-Decanoate Wavenumber (cm-1) % T (a) Decanoic acid 950 1100 1250 1400 1550 1700 1850 2000 2150 2300 2450 2600 2750 2900 3050 3200 3350 3500 3650 3800 3950 Pure CuS 0.05:1 Acid:Cu 0.25:1 Acid:Cu Pure Na-CHP Wavenumber (cm-1) % T (b) Cyclohexane pentanoic acid Figure 5.8: IR spectrograph of naphthenic acid surrogates mixed with CuS at high pH 47 Although CuS is omitted from Table 5.4, one again can observe the water of hydration on the CuS mineral species at approximately 3425cm−1. In this series of experiments, the 1700cm−1 band due the associated carboxylic acid group R-COOH is absent as expected - this occurs as the sodium or cop- per cations replace the proton at the carboxyl functional group. Instead of the 1700cm−1 band, the Na-DEC and Na-CHP show two peaks, one at 1570cm−1 and at 1440cm−1. These two bands agree with those assigned to the antisymmetric and symmetric vibrations, respectively, of the C-O stretching in sodium acetate Mehrotra and Bohra [1983]. In the case of DA mixed with CuS, the shifted 1700cm−1 band re-appears at 1600 and 1500cm−1, as the antisymmetric portion of the C-O stretch. The sym- metric complements appear at 1400 and 1440cm−1, respectively. The two anti- symmetric and symmetric peaks have the respective wavenumber differences of 200 and 60cm−1. This indicates that there are both bridging and chelating com- pounds formed under these conditions. In the case of CHPA mixed with the CuS, the shifted 1700cm−1 band re- appears at 1590cm−1. This, along with the1410cm−1 band, represent the anti- symmetric and symmetric C-O vibrations, respectively. As discussed before, the difference between the antisymmetric and symmetric can illuminate the resultant mineral complex. In this case, the difference is approximately 180cm−1, indicative of a bridging complex. 5.2.3 Copper hydroxide at high pH The DRIFT spectra of Cu(OH)2 mixed with DA and CHPA at pH 8.5 are shown in Figure 5.9 The different spectra show the indicated acid to metal molar ratios. Pure sodium salts of DA and CHPA are shown for reference. The key bands are tabulated in Table 5.5. Similarly to the CuS at pH 8.5 reaction, the 1700cm−1 band is absent as there is no true carboxyl group on the organic. Instead, both the reference sodium car- boxylate salts and the resultant copper-organic complexes (at 0.25:1 acid to copper ratio) show the shifted band at under 1600cm−1. Both the DA and CHPA resultant minerals have antisymmetric and symmetric C-O stretching bands spaced closely 48 950 1100 1250 1400 1550 1700 1850 2000 2150 2300 2450 2600 2750 2900 3050 3200 3350 3500 3650 3800 3950 Pure Cu(OH)2 0.05:1 Acid:Cu 0.25:1 Acid:Cu 2:1 Acid:Cu Pure Na-Decanoate Wavenumber (cm-1) % T (a) Decanoic acid 950 1100 1250 1400 1550 1700 1850 2000 2150 2300 2450 2600 2750 2900 3050 3200 3350 3500 3650 3800 3950 Pure Cu(OH)2 0.05:1 Acid:Cu 0.25:1 Acid:Cu Pure Na-CHP Wavenumber (cm-1) % T (b) Cyclohexane pentanoic acid Figure 5.9: IR spectrograph of naphthenic acid surrogates mixed with Cu(OH)2 at high pH 49 Table 5.5: Summary of key spectral bands for Cu(OH)2 mixed with DA and CHPA at high pH Acid:Cu Ratio Wavenumber (cm−1) Chemical Group and Motion 0:1 (Pure Cu(OH)2) 3580 O-H stretching on Cu(OH)2 0:1 (Pure Cu(OH)2) 3400 and 3270 unknown, likely O-H on Cu 0:1 (Pure Cu(OH)2) 1110 SO4 stretching 0.25:1 (DA) 3580 O-H stretching on Cu(OH)2 0.25:1 (DA) 3400 and 3200 unknown, likely O-H on Cu 0.25:1 (DA) 2920 and 2840 C-H stretching on hydrocarbon 0.25:1 (DA) 1510 antisym. C-O stretch (chelating compound) 0.25:1 (DA) 1450 symmetric C-O stretching (chelating compound) 2:1 (DA) 2920 and 2840 C-H stretching on hydrocarbon 2:1 (DA) 1550 antisym. C-O stretch (bridging compound) 2:1 (DA) 1390 symmetric C-O stretching (bridging compound) 0.25:1 (CHPA) 3410 and 3320 unknown, likely O-H on Cu 0.25:1 (CHPA) 2930 and 2860 C-H stretching on hydrocarbon 0.25:1 (CHPA) 1500 antisym. C-O stretch (chelating compound) 0.25:1 (CHPA) 1410 symmetric C-O stretching (chelating compound) together - the differences between the bands are 60 and 90cm−1, respectively. This indicates that the principle complexation mechanism is chelating at the low acid to copper ratio. Once the decanoic acid to copper ratio is increased to the expected stoichiom- etry (2:1 DA:Cu), the difference between the antisymmetric and symmetric C-O stretching bands increases to 160cm−1, indicating a bridging compound at the higher reagent ratios. As well, at this highest DA:Cu ratio, the O-H bands for the mineral hydroxide groups (around 3600/3400/3200cm−1 vanish, indicating that all the hydroxide groups are consumed by the carboxylate bonding. Note as well that the 0.05:1 decanoic acid to copper ratio spectrograph shows little activity at the 2900cm−1 range (C-H stretching) or the 1600/1400cm−1 bands. However, the 0.05:1 CHPA to copper ratio shows much more activity in these band regions. This suggests that straight chain carboxylic acids are less reactive at low ratios than cyclic materials. This may be due to the extra hydrophobicity of the 50 straight chain compounds (with increased molecular surface area relative to a cyclic structure) driving a hydrophobic partition of the organic to the mineral surface. In the case of the cyclic material, with less hydrophobic impetus, reaction with the mineral surface assists the partition out of the water phase. Taken together, these shifts in the spectral bands demonstrate that: 1. Low DA to copper ratios (0.05:1 or less) shows little complexation. 2. Low CHPA to copper ratios (less than 0.25:1) show some complexation. 3. Intermediate acid to copper ratios (approximately 0.25:1) show chelating complexes. 4. High DA to copper ratios ( 2:1, stoichiometric) has bridging complexation and complete consumption of the surface hydroxide groups. The significance of these findings is that the naphthenic acid/mineral interac- tions are not of a uniform type, but vary with the acid to metal ratio and the structure of the organic compound in question. 5.2.4 Copper hydroxide co-precipitation The DRIFT spectra of Cu(OH)2 co-precipitated with DA and CHPA at pH 8.5 are shown in Figure 5.10 The different spectra show the indicated acid to metal molar ratios. Pure sodium salts of DA and CHPA are shown for reference. The key bands are tabulated in Table 5.6. As with the previous high pH experiments, the 1700cm−1 band representing true carboxylic acid is absent. Instead, the reference sodium carboxylate salts and the resultant copper-organic complexes (at 0.25:1 acid to copper ratio) show the shifted band to approximately 1600cm−1. For both the DA and CHPA samples, the resultant material has antisymmetric and symmetric C-O stretching bands spaced widely - the differences between the bands are both 190cm−1. This indicates that the principle complexation mechanism is bridging at this acid to copper ratio. As above, the 0.05:1 DA:Cu ratio spectrograph shows no band at the 2900cm−1 or the 1600/1400cm−1 wavenumbers. However, the CHPA does show bands at 51 950 1100 1250 1400 1550 1700 1850 2000 2150 2300 2450 2600 2750 2900 3050 3200 3350 3500 3650 3800 3950 Pure Cu(OH)2 0.05:1 Acid:Cu 0.25:1 Acid:Cu 2:1 Acid:Cu Pure Na-Decanoate Wavenumber (cm-1) % T (a) Decanoic acid 950 1100 1250 1400 1550 1700 1850 2000 2150 2300 2450 2600 2750 2900 3050 3200 3350 3500 3650 3800 3950 Pure Cu(OH)2 0.05:1 Acid:Cu 0.25:1 Acid:Cu Pure Na-CHP Wavenumber (cm-1) % T (b) Cyclohexane pentanoic acid Figure 5.10: IR spectrograph of naphthenic acid surrogates co-precipitated with Cu(OH)2 52 Table 5.6: Summary of key spectral bands for Cu(OH)2 co-precipitated with DA and CHPA at high pH Acid:Cu Ratio Wavenumber (cm−1) Chemical Group and Motion 0.25:1 (DA) 3580 O-H stretching on Cu(OH)2 0.25:1 (DA) 3400 and 3260 unknown, likely O-H on Cu 0.25:1 (DA) 2920 and 2850 C-H stretching on hydrocarbon 0.25:1 (DA) 1600 antisym. C-O stretch (bridging compound) 0.25:1 (DA) 1410 symmetric C-O stretching (bridging compound) 0.25:1 (DA) 1110 SO4 stretching 2:1 (DA) 2920 and 2840 C-H stretching on hydrocarbon 2:1 (DA) 1590 antisym. C-O stretch (bridging compound) 2:1 (DA) 1420 symmetric C-O stretching (bridging compound) 0.05:1 (CHPA) 3600 O-H stretching on Cu(OH)2 0.05:1 (CHPA) 3400 and 3270 unknown, likely O-H on Cu 0.05:1 (CHPA) 2930 and 2850 C-H stretching on hydrocarbon 0.05:1 (CHPA) 1510 antisym. C-O stretch (chelating compound) 0.05:1 (CHPA) 1420 symmetric C-O stretching (chelating compound) 0.25:1 (CHPA) 3600 O-H stretching on Cu(OH)2 0.25:1 (CHPA) 3400 and 3270 unknown, likely O-H on Cu 0.25:1 (CHPA) 2930 and 2850 C-H stretching on hydrocarbon 0.25:1 (CHPA) 1600 antisym. C-O stretch (bridging compound) 0.25:1 (CHPA) 1410 symmetric C-O stretching (bridging compound) 0.25:1 (CHPA) 1510 antisym. C-O stretch (chelating compound) 0.25:1 (CHPA) 1440 symmetric C-O stretching (chelating compound) these wavenumbers, and with the wavenumber difference between the antisym- metric and symmetric C-O stretching peaks at 90cm−1, the final structure appears to be a chelating complex. As explained above, this may be caused by the extra hydrophobicity of the straight chain compounds driving a hydrophobic partition of the organic out of the water phase, whereas the cyclic structure of CHPA may encourage complexation as a motive force for water escape. As with the Cu(OH)2 mixing experiments, at the expected stoichiometry of 2:1DA:Cu ratio, the difference between the antisymmetric and symmetric C-O stretching bands remains high, at 180cm−1, indicating a bridging compound. Like- 53 wise, at this highest acid:Cu ratio, the O-H bands for the mineral hydroxide groups are absent, indicating that all the hydroxide groups are consumed by the carboxy- late bonding. These changes to the spectral bands with changing acid:Cu ratios indicate that: 1. Low DA:Cu ratios have little complexation occurring, although low CHPA:Cu ratios may result in some chelating complexes; this difference can be ex- plained by the difference in the molecular structures. 2. Intermediate DA:Cu ratios show bridging complexes, while intermediate CHPA:Cu ratios result in mixed chelating and bridging complexes 3. High DA:Cu ratio (stoichiometric) demonstrates a bridging complexation and complete consumption of the surface hydroxide groups. As described above, the conclusion is that these naphthenic acid surrogates do not follow a particular interaction, but rather the reaction mechanism depends on the acid to metal ratio and on the molecular structure. 5.2.5 Summary From the adsorption isotherms, pH measurements during reaction and IR spec- troscopy performed on the resultant organic/mineral mixtures, it appears that there is no singular reaction or complexation mechanism to describe the interactions be- tween these surrogate naphthenic acids and copper minerals. The exact nature of the water-solid partition varies strongly with pH, with acid to metal ratio and with the molecular structure of the organic sorbate . From the data presented here, it ap- pears as though pH plays the dominant role for water-solid partition of the organic compounds at relatively low acid:metal ratios, whereas when the acid:metal ratio is increased closer to the true chemical stoichiometry of the complexation com- pounds, the acid to metal ratio plays the dominant role. The hydrophobicity of the organic compound appears to complement these two effects. 5.3 Process Implications These experiments generally do not hold great promise for process development. The best recovery of DA and CHPA from the aqueous phase was on the order of 54 80% - 85% and it is conceivable that, pending further research as outlined later, this order of NA removal from effluent may be sufficient as a process keystone. One potential avenue for treatment could be: 1. Mix OSPW with the inorganic mineral in a vessel, either continuous stirred tank reactor (CSTR) or batch reactor. 2. The outflow from the initial reaction vessel reports to a solid-liquid separa- tion stage 3. The separated water flows to polishing treatment (ie: pH adjustment or con- ventional softening) or to process reuse 4. The organic-bearing mineral reports for further treatment, either; a) stockpiling the wet solids in smaller sludge ponds and left for biodegra- dation, or; b) passing the sludge through an ozone column and oxidizing the refrac- tory compounds prior to stockpiling for bioremediation in sludge ponds. While the copper minerals may be recovered after the organics have decom- posed, the time required for this natural process is unknown. Ultimately the amount of copper required for these recoveries or concentrating steps is prohibitive. To wit, the amount of copper required (as copper in sulphide mineral form) in the best re- covery scenario is 0.1g in a liquid volume of 150ml. This scales up to 666g Cu in one cubic meter of contaminated water. Assuming copper sulphide can be had for approximately one-half the market value of cathode copper ($9000/DMT USD using London Metal Exchange spot price for June 17, 2011), the copper value of semi-treated water would be approximately $3.00/m3. This does not include the cost of other peripheral treatments, such as ozonation, holding ponds for long-term biodegradation of the adsorbed compounds, the cost of acids, caustics or buffers for finer process control, or disposal of the other contaminant materials. However, these experiments represent only the first step in understanding how naphthenic acids could be adsorbed onto inorganic minerals. Future research into these reac- tions could yield improved process potential. 55 5.4 Future Research While these experiments resulted in ambiguous results that are slippery to interpret, much can be suggested for future research to clarify the described effects. Clearly the dominant effect on the mineral uptake of the organic acids is the solution pH. In particular, the experiments described herein treat the effect of pH as a binary value. Low pH was assumed to result in re-associated acids and non-hydroxylated mineral surfaces, giving minimal organic:mineral reaction. High pH was assumed to result in highly dissociated acids and hydroxylated mineral surfaces, resulting in maximal organic:mineral reaction. However, from the findings shown here, this is not the case, as the shifting solution pH clearly changed the mineral uptake of acid. Any future work on this topic must treat pH as an independent variable in the experimental design. Likewise, the experiments as performed fixed the copper quantity in the mix- ing flasks and varied the concentration of the naphthenic acid surrogates to adjust the acid:Cu molar ratio. Considering that high concentrations of organic acids tended to result in heterogenous phase separations, and that any industrial process would assume the native organic acid concentration (typically 100ppm in oil sands surface mining tailings ponds) and add copper to optimal process conditions, any future research into these phenomena with the goal of developing a water treatment method must use fixed acid concentrations and adjust the copper content to meet the acid:metal molar ratio targets. Other considerations for further investigation include the use of cheaper and more readily available metallic minerals. Copper, while having predictable oxi- dation states and low carboxylate solubility, is uneconomical for the purpose of partially treating waste water. Alternative metals could include zinc or iron, both much cheaper than copper and both known to form acetate complexes. Any final process development must accurately describe the in situ conditions of the target effluent. These experiments, operating in de-ionized water at 25◦C, are not representative of a true oil sands surface mining pond. Final process exper- iments must include the ionic strength and make-up of the OSPW, the temperature of an outdoors pond in northern Canada and demonstrate usefulness on a complete native mixture of all naphthenic acids, not just two surrogates of relatively low 56 molecular weight and low ring number. In addition, one must determine the kinetics of reaction. This is required to ensure that, while all other aspects of the reactions may be feasible, the whole treatment can take place in a reasonable period that does not require outlandish reactor residence times. 57 Chapter 6 Conclusions The study of naphthenic acids is relatively new. While these compounds have much industrial consequence, little is known for certain about their reactivity in multi- phase mineralogical systems. With this research, I have demonstrated that the reactions between inorganic minerals and the selected naphthenic acid surrogates, decanoic acid and cyclohexane pentanoic acid, varies with solution pH, organic to metal molar ratio and molecular structure. Low solution pH clearly hinders the complexation of the organic ligand to the metal in the mineral. High pH solutions tend to retain the organic acids in solu- tions. pH changes during the acid:metal reaction have a significant impact on the solubility of the organic compounds in the water phase. The net result is that in the organic/mineral system, there is not a clean two-way partition between water and mineral surface. In spite of the ambiguous adsorption isotherms, IR spectroscopy demonstrates that, at elevated solution pH (resulting in dissociation of the acid and hydroxylation of the mineral surfaces), sorbtion of the carboxylate anions to the surface available copper in the mineral results in species that are copper-carboxylate complexes. The nature of these complexes changes with the type of mineral (whether sulphide or hydroxide) and the acid to metal ratio. The structure of the organic compounds impacted the adsorption experiments and the identification of the resultant acid-metal complexes. The cyclic nature of the CHPA results in a more hydrophilic molecule that tends to remain in the wa- 58 ter phase more so than the straight-chain DA. Concurrently, the CHPA tended to form organic-metal complexes at lower acid:metal ratios than the DA. These obser- vations imply that the decreased hydrophobicity of molecules with lower surface area per mass show have a significant impact on how the organics behave in these systems. These results confirm the hypothesis that the principle mechanism for naph- thenic acid adsorption to inorganic minerals is the carboxyl ligand function with respect to the metals contained in the mineral. However, there are several other confounding factors in these experiments that are not quantified. While this is a step towards the understanding of this family of toxic substances, more research is required to describe these systems with more accuracy and clarity. 59 Bibliography E. Allen. Process water treatment in canadas oil sands industry: I. target pollutants and treatment objectives. J Environ Eng Sci, 7:123 – 138, 2008. → pages 1 M. S. Andres, J. M. de la Roja, V. G. Baonza, and N. Sancho. Verdigris pigment: a mixture of compounds. input from raman spectroscopy. J Raman Spectrosc, 41(11):1178–1186, 2010. → pages 12 P. Atkins. Physical Chemistry. W.H. Freeman and Company, 6th edition, 1997. → pages 9 K. Banerjee, P. Cheremisinoff, and S. Cheng. Adsorption kinetics of o-xylene by flyash. Water Res, 31(2):249–261, 1997. → pages 8 R. Bayard, L. Barna, B. Mahjoub, and R. Gourdon. Investigation of naphthalene sorption in soils and soil fractions using batch and column assays. Environ Toxicol Chem, 17(12):2383–2390, 1998. → pages 8 J. Bebie, M. Schoonen, M. Fuhrmann, and D. Strongin. Surface charge development on transition metal sulfides: An electrokinetic study. Geochim Cosmochim Ac, 62(4):633–642, 1998. → pages 15 L. Bijan and M. Mohseni. Integrated ozone and biotreatment of pulp mill effluent and changes in biodegradability and molecular weight distribution of organic compounds. Water Res, 39(16):3763–3772, 2005. → pages 5 E. Bogusz, S. Brienne, I. Butler, S. Rao, and J. Finch. Metal ions and dextrin adsorption on pyrite. Minerals Engineering, 10(4):441–445, 1997. → pages 16, 17 O. Brandal, A. Hanneseth, P. Hemmingsen, J. Sjoblom, S. Kim, R. Rodgers, and A. Marshall. Isolation and characterization of naphthenic acids from a metal naphthenate deposit: Molecular properties at oil-water and air-water interfaces. J Disper Sci Technol, 27(3):295–305, 2006. → pages 12 60 J. A. Brient, P. J. Wessner, and M. N. Doyle. ”Naphthenic Acids”. In: Kirk-Othmer Encyclopedia of Chemical Technology, pages 1–10. John Wiley and Sons, Inc., Dec 2000. → pages 1, 17 R. Calvet. Adsorption of organic chemicals in soils. Environ Health Persp, 83: 145–177, 1989. → pages 14 J. Clemente and P. Fedorak. A review of the occurrence, analyses, toxicity, and biodegradation of naphthenic acids. Chemosphere, 60(5):585–600, 2005. → pages 6 J. Clemente, T. Yen, and P. Fedorak. Development of a high performance liquid chromatography method to monitor the biodegradation of naphthenic acids. J Environ Eng Sci, 2(3):177–186, 2003. → pages 7, 27 D. Connaughton, J. Stedinger, L. Lion, and M. Shuler. Description of time-varying desorption-kinetics - release of naphthalene from contaminated soils. Environ Sci Technol, 27(12):2397–2403, 1993. → pages 8 R. Craciun and P. Kamdem. Xps and ftir applied to the study of waterborne copper naphthenate wood preservatives. Holzforschung, 51(3):207–213, 1997. → pages 12, 13, 17 J. Dunn and C. Muzenda. Thermal oxidation of covellite (cus). Thermochim Acta, 369(1-2):117–123, 2001. → pages 13 R. A. Frank, R. Kavanagh, B. K. Burnison, G. Arsenault, J. V. Headley, K. M. Peru, G. V. der Kraak, and K. R. Solomon. Toxicity assessment of collected fractions from an extracted naphthenic acid mixture. Chemosphere, 72(9): 1309–1314, 2008. → pages 5, 6 R. A. Frank, K. Fischer, R. Kavanagh, B. K. Burnison, G. Arsenault, J. V. Headley, K. M. Peru, G. V. der Kraak, and K. R. Solomon. Effect of carboxylic acid content on the acute toxicity of oil sands naphthenic acids. Environ Sci Technol, 43(2):266–271, 2009. → pages 6 P. Gosselin, S. Hrudey, M. A. Naeth, A. Plourde, R. Therrien, G. VanDerKraak, and Z. Xu. Environmental and health impacts of canada’s oil sands industry. The Royal Society of Canada, December 2010. → pages 2 X. Han, M. D. Mackinnon, and J. W. Martin. Estimating the in situ biodegradation of naphthenic acids in oil sands process waters by hplc/hrms. Chemosphere, 76(1):63–70, 2009. → pages 2 61 T. E. Havre. Formation of Calcium Naphthenate in water:oil systems, naphthenic acid chemistry and emulsion stability. PhD thesis, Norwegian University of Science and Technology, Oct 2002. → pages 10 J. Headley and D. McMartin. A review of the occurrence and fate of naphthenic acids in aquatic environments. J Environ Sci Heal A, 39(8):1989–2010, 2004. → pages 1, 2 J. V. Headley, K. M. Peru, and M. P. Barrow. Mass spectrometric characterization of naphthenic acids in environmental samples: A review. Mass Spectrom. Rev., 28(1):121–134, 2009. → pages 7 F. Holowenko, M. MacKinnon, and P. Fedorak. Characterization of naphthenic acids in oil sands wastewaters by gas chromatography-mass spectrometry. Water Res, 36(11):2843–2855, 2002. → pages 5, 7, 27 A. Janfada, J. Headley, K. Peru, and S. Barbour. A laboratory evaluation of the sorption of oil sands naphthenic acids on organic rich soils. J Environ Sci Heal A, 41(6):985–997, 2006. → pages 8 R. Kunzig. Scraping bottom. National Geographic, Mar 2009. → pages 2 Q. Liu and J. Laskowski. The interactions between dextrin and metal-hydroxides in aqueous-solutions. J Colloid Interf Sci, 130(1):101–111, 1989. → pages 34 Q. Liu, Y. Zhang, and J. Laskowski. The adsorption of polysaccharides onto mineral surfaces: an acid/base interaction. Int J Miner Process, 60(3-4): 229–245, 2000. → pages 16, 17 M. MacKinnon and H. Boerger. Description of two treatment methods for detoxifying oil sands tailings pond water. Water quality research journal of Canada, 21(4):496–512, 1986. → pages 4, 5 J. W. Martin, T. Barri, X. Han, P. M. Fedorak, M. G. El-Din, L. Perez, A. C. Scott, and J. T. Jiang. Ozonation of oil sands process-affected water accelerates microbial bioremediation. Environ Sci Technol, 44(21):8350–8356, 2010. → pages 5 S. Mauchauffee and E. Meux. Use of sodium decanoate for selective precipitation of metals contained in industrial wastewater. Chemosphere, 69(5):763–768, 2007. → pages 10, 43 R. C. Mehrotra and R. Bohra. Metal carboxylates. Academic Press, 1983. → pages vi, 11, 12, 13, 21, 48 62 W. S. W. Ngah and M. A. K. M. Hanafiah. Biosorption of copper ions from dilute aqueous solutions on base treated rubber (hevea brasiliensis) leaves powder: kinetics, isotherm, and biosorption mechanisms. J Environ Sci-China, 20(10): 1168–1176, 2008. → pages 12 E. Palacios, G. Juarez-Lopez, and A. Monhemius. Infrared spectroscopy of metal carboxylates - ii. analysis of fe(iii), ni and zn carboxylate solutions. Hydrometallurgy, 72(1-2):139–148, 2004. → pages 12 J. Peng, J. Headley, and S. Barbour. Adsorption of single-ring model naphthenic acids on soils. Can Geotech J, 39(6):1419–1426, 2002. → pages 8 B. Phillips and W. Busing. Comparison of the infrared and raman spectra of some crystalline hydroxides. J Phys Chem-Us, 61(4):502–502, 1957. → pages 13 M. Piwoni and P. Banerjee. Sorption of volatile organic solvents from aqueous solution onto subsurface solids. Journal of Contaminant Hydrology, 4:163 – 179, 1989. → pages 8 M. Protic and A. Sabljic. Quantitative structure activity relationships of acute toxicity of commercial chemicals on fathead minnows - effect of molecular-size. Aquat Toxicol, 14(1):47–64, 1989. → pages 6 E. Quagraine, H. Peterson, and J. Headley. In situ bioremediation of naphthenic acids contaminated tailing pond waters in the athabasca oil sands region-demonstrated field studies and plausible options: A review. J Environ Sci Heal A, 40(3):685–722, 2005. → pages 5 G. Raju, A. Holmgren, and W. Forsling. Adsorption of dextrin at mineral/water interface. J Colloid Interf Sci, 193(2):215–222, 1997. → pages 16, 17 G. Raju, A. Holmgren, and W. Forsling. Complexation mechanism of dextrin with metal hydroxides. J Colloid Interf Sci, 200(1):1–6, 1998. → pages 17 D. Roberts. Qsar issues in aquatic toxicity of surfactants. Sci Total Environ, 109: 557–568, 1991. → pages 6 V. Rogers, K. Liber, and M. MacKinnon. Isolation and characterization of naphthenic acids from athabasca oil sands tailings pond water. Chemosphere, 48(5):519–527, 2002a. → pages 1, 12 V. Rogers, M. Wickstrom, K. Liber, and M. MacKinnon. Acute and subchronic mammalian toxicity of naphthenic acids from oil sands tailings. Toxicol Sci, 66 (2):347–355, 2002b. → pages 6 63 R. A. Ronconi. Predicting bird oiling events at oil sands tailings ponds and assessing the importance of alternate waterbodies for waterfowl: A preliminary assessment. Can Field Nat, 120(1):1–9, 2006. → pages 2 J. SanchezRuiz. An estimate of shape-related contributions to hydrophobic gibbs energies. J Phys Chem-Us, 99(32):12076–12080, 1995. → pages 43 A. C. Scott, R. F. Young, and P. M. Fedorak. Comparison of gc-ms and ftir methods for quantifying naphthenic acids in water samples. Chemosphere, 73 (8):1258–1264, 2008a. → pages 7 A. C. Scott, W. Zubot, M. D. Mackinnon, D. W. Smith, and P. M. Fedorak. Ozonation of oil sands process water removes naphthenic acids and toxicity. Chemosphere, 71(1):156–160, 2008b. → pages 5 K. Shinoda. The critical micelle concentration of soap mixtures (2-component mixture). J Phys Chem-Us, 58(7):541–544, 1954. → pages 36 C. E. Silva, L. P. Silva, H. G. M. Edwards, and L. F. C. de Oliveira. Diffuse reflection ftir spectral database of dyes and pigments. Anal Bioanal Chem, 386 (7-8):2183–2191, 2006. → pages 12, 13 C. M. Simonescu, V. S. Teodorescu, O. Carp, L. Patron, and C. Capatina. Thermal behaviour of cus (covellite) obtained from copper-thiosulfate system. J Therm Anal Calorim, 88(1):71–76, 2007. → pages 13 B. E. Smith, C. A. Lewis, S. T. Belt, C. Whitby, and S. J. Rowland. Effects of alkyl chain branching on the biotransformation of naphthenic acids. Environ Sci Technol, 42(24):9323–9328, 2008. → pages 24 I. Tunon, E. Silla, and J. Pascualahuir. Molecular-surface area and hydrophobic effect. Protein Eng, 5(8):715–716, 1992. → pages 43 T. Yen, W. Marsh, M. MacKinnon, and P. Fedorak. Measuring naphthenic acids concentrations in aqueous environmental samples by liquid chromatography. J Chromatogr A, 1033(1):83–90, 2004. → pages 7, 8, 27 L. Zhou, B. Han, H. Yan, K. Kasperski, U. Xu, and L. Hepler. Enthalpy of adsorption and isotherms for adsorption of naphthenic acid onto clays. J Colloid Interf Sci, 190(2):472–475, 1997. → pages 8 X. Zhu, J. Li, and M. Wadsworth. Characterization of surface-layers formed during pyrite oxidation. Colloid Surface A, 93:201–210, 1994. → pages 15 64 J. Zyskowski and D. Kamdem. Ultraviolet spectrophotometry and fourier transform infrared spectroscopy characterization of copper naphthenate. Wood Fiber Sci, 31(4):441–446, 1999. → pages 12, 18 65

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0059078/manifest

Comment

Related Items