Understanding Hydrogen in Bayer Process Emissions. 1. Hydrogen

ARTICLE pubs.acs.org/IECR

Understanding Hydrogen in Bayer Process Emissions. 1. Hydrogen Production during the Degradation of Hydroxycarboxylic Acids in Sodium Hydroxide Solutions Allan Costine,*,† Joanne S.C. Loh,† Greg Power,† Mark Schibeci,† and Robbie G. McDonald‡ CSIRO Process Science and Engineering (Parker CRC), P.O. Box 7229, Karawara WA 6152, Australia ABSTRACT: The formation of potentially explosive gas mixtures during Bayer process digestion and the wet oxidation of Bayer process liquors underscores the need for an improved understanding of the degradation reactions of organic compounds that produce flammable gases. This study is the first of a series investigating the production of hydrogen from different classes of organic compounds in sodium hydroxide solutions. The alkaline degradation of a range of aliphatic and aromatic carboxylates and hydroxycarboxylates was investigated under anaerobic conditions in an autoclave. It was found that aliphatic C4 carboxylates possessing a single β-hydroxy substituent are particularly reactive under these conditions, generating hydrogen gas and a range of low molecular weight (LMW) carboxylates. The effect of temperature (175
275 °C) and NaOH concentration (0
6 M) on the degradation of 3-hydroxybutanoate and hydroxybutanedioate (malate) was investigated in detail for reaction times up to 120 min. Under conditions that promote the total degradation of the organic compound, the β-hydroxycarboxylates have similar hydrogen production capacities, each generating about 1 mol of hydrogen gas per mole of organic compound consumed. The results provide direct evidence for an ionic degradation mechanism involving base-catalyzed oxidation by water, consistent with the stoichiometry of the formation of hydrogen gas and the other main reaction products (LMW carboxylates). These findings have important implications for the production of hydrogen in Bayer process digestion and the safe application of wet oxidation technologies for the treatment of organic compounds in alkaline liquors, such as those found in the Bayer and Kraft processes. The results also shed new light on the mechanism of hydrogen production during the base-catalyzed gasification of biomass at low temperature.

1. INTRODUCTION Virtually all of the world’s alumina is produced from bauxite by the Bayer process.1 The recirculating Bayer process liquor generally contains organic impurities, which originate primarily from the bauxite.2
5 These organic compounds have numerous deleterious effects on practically all aspects of alumina production.2,6
8 Wet oxidation9 is one strategy for the selective destruction of organic compounds in Bayer liquors that has been applied commercially.10,11 However, the alkaline degradation of complex mixtures of organic compounds under oxygen-deficient and oxygen-rich conditions is not yet fully understood. In particular, the degradation pathways that generate hydrogen gas have not been established. The production of hydrogen in the Bayer process is known to occur under the oxygen-deficient conditions associated with bauxite digestion12,13 and also under the oxidizing conditions employed during wet oxidation.10,14
17 A study of the composition of the digestion vent gases at the Queensland Alumina refinery in Australia showed that the emissions typically contained 2% noncondensable gases in steam.13 These noncondensable gases consisted mainly of hydrogen (89.7%), with smaller amounts of nitrogen (8.6%), methane and other hydrocarbons (1.2%), the carbon oxides (0.3%) and oxygen (0.2%). If wet oxidation is used to remove organic compounds from liquor, the potential exists for the production of hydrogen and other flammable gases to result in the formation of explosive gas mixtures.10 The minimum ignition energy of a hydrogen-air mixture is below 0.02 mJ, whereas that of other flammable gases such as methane and benzene, both of which have been identified Published 2011 by the American Chemical Society

in alumina refinery emissions,18 is usually higher than 0.2 mJ.19 In addition, the flammability limits for hydrogen are exceptionally wide.20 Little work has been reported to date on the production of hydrogen from specific classes of organic compounds in sodium hydroxide solutions. One class of organic compounds
the hydroxycarboxylic acids
is of particular interest because they are produced in the alkaline degradation of humates, lignins, and all carbohydrates (from mono- to polysaccharides). Indeed, a remarkable variety of linear and branched, C2
C6 hydroxycarboxylates have been detected following the alkaline degradation of glucose, cellobiose, and cellulose at temperatures ranging from 20 to 320 °C.21
23 On this basis, it seems likely that these types of compounds are also produced in Bayer digestion. Consistent with this, Niemel€a and Grocott24 have reported the identification of more than 200 compounds in actual plant liquors, including numerous hydroxycarboxylates and a variety of phenolic compounds.25 However, the concentrations reported are relatively low, and in general the literature on the identification of organic compounds in Bayer liquor shows a predominance of unsubstituted aliphatic and aromatic carboxylates. This has led to the suggestion that hydroxycarboxylates may be subject to continuous removal by ongoing reaction in the recirculating liquor and/or by adsorption processes, particularly to the surfaces of the aluminum hydroxide precipitation product,3 Received: June 28, 2011 Accepted: October 1, 2011 Revised: September 26, 2011 Published: October 01, 2011 12324

dx.doi.org/10.1021/ie201387x | Ind. Eng. Chem. Res. 2011, 50, 12324–12333

Industrial & Engineering Chemistry Research consistent with their known effects as yield inhibitors.26
29 It may also be that these compounds degrade relatively rapidly to produce low molecular weight (LMW) compounds and possibly hydrogen. Two different degradation mechanisms have been proposed to explain the reactivity of the hydroxycarboxylates, such as sodium tartrate, in alkaline solutions. In the presence of oxygen, free radical intermediates are likely to be involved,30
32 consistent with the observation that tartrate has a tendency to facilitate the degradation of succinate, which was unreactive under the same conditions in the absence of tartrate. In the absence of oxygen, an ionic degradation mechanism involving base-catalyzed oxidation by water has been proposed.27,33 The significance of this mechanism in the current context is that it not only explains the occurrence of oxidative reactions under anaerobic conditions2,34
36 but also shows how and why hydrogen gas can be produced even in the presence of oxygen. The aims of this work are to identify the specific structural features and define the reaction conditions that promote hydrogen production during the alkaline degradation of a range of organic compounds, with particular attention to the hydroxycarboxylates, and to provide quantitative information on the products of reaction. It will also serve as a baseline for future wet oxidation studies of organic compounds, particularly in light of reports which indicate that the production of hydrogen may either decrease15 or increase16 with increasing oxygen concentrations during wet oxidation of Bayer liquors. Knowledge of the thermal degradation reactions of organic compounds that produce hydrogen is therefore of significant interest to the development of improved wet oxidation processes.

2. EXPERIMENTAL SECTION 2.1. Autoclave Facility. The thermal degradation experiments were carried out in a 2 L Inconel 600 autoclave (Parr Instruments Co., IL, U.S.A.). The autoclave was controlled to within (2 °C of the specified temperature using a Parr 4842 temperature controller. Agitation of the solutions was provided by magnetically driven, twin, six-pitched-blade impellers (260 rpm during N2 purging and 60 rpm during solution heat-up). The autoclave pressure was recorded using an online pressure transducer. Two duplicate experiments were also performed in a 2 L Parr zirconium (Zr 705) autoclave to enable detection of any possible effect on product yields of the materials of construction of the Inconel autoclave (Ni, Fe, Cr). The production of hydrogen and LMW carboxylate products in both autoclaves were the same within experimental error. The oxygen and hydrogen concentrations in the headspace were measured using online electrochemical sensors. Each sensor features a gas permeable membrane that regulates the gas flow into the sensor. In principle, gases other than hydrogen could also give signals if they could react to produce or consume protons at the electrode, but the permeability of hydrocarbons through the membrane is expected to be much lower than for hydrogen. To test this, the hydrogen sensor was subjected to nitrogen gas flows containing known concentrations of propylene, which has been identified in alumina refinery emissions,18 and no response was found. The detection limit of the sensor is 0.01 mol % hydrogen. The analytical performance of electrochemical hydrogen sensors was recently reviewed.37 A detailed description of the autoclave facility can be found elsewhere.38 2.2. Organic Compounds. The range of organic compounds tested is shown in Table 1. All reagents were AR grade and used

ARTICLE

as received. Most of the compounds listed are carboxylic anions, though three phenols
catechol, resorcinol, and pyrogallol
were also tested. Three forms of tartrate (D, L, and meso) were tested to determine whether the stereochemistry of hydroxy groups has any effect on product yields. The presence of monoaromatic compounds in Bayer liquors is an expected result of the degradation of humates and lignins in sodium hydroxide solutions.39,40 Of the 40 compounds investigated, 29 have been specifically identified in actual Bayer liquors.3 The initial concentration was 40 mM for all the compounds except oxalate, terephthalate and trimesate, for which 5 mM was used because of their limited solubility. The effects of various reaction parameters on the degradation of two structurally similar hydroxycarboxylates, DL-3-hydroxybutanoate (CH3 3 CH(OH) 3 CH2 3 CO2
) and DL-malate (CO2
3 CH(OH) 3 CH2 3 CO2
), were then further studied in detail. 2.3. Solution Preparation. Sodium hydroxide solutions were prepared by dissolving the appropriate mass of NaOH pellets (>97%, 1% Na2CO3, Rowe Scientific) in deionized water. Solutions of each organic compound were made up by adding the required mass of organic compound to the sodium hydroxide solution. A solution volume of 1 L was used for each degradation experiment (i.e., headspace volume fraction = 50%), and care was taken to restrict contact of the solutions with air. The autoclave was then sealed and inserted into the heating cavity. 2.4. System Preparation. After confirmatory pressure testing, the autoclave and gas sampling system were purged with nitrogen. When the concentration of residual oxygen in the headspace was below 0.01 mol %, the excess nitrogen pressure was released. This procedure ensured that reproducible oxygen-deficient conditions were created for each degradation experiment prior to heatup, to minimize the possibility of oxygen-induced degradation reactions. The solutions were heated at a rate of approximately 5 °C min
1, with stirring, to the designated temperature. The time at which the required temperature was reached was taken as t = 0 min, the nominal starting point of the reaction. Accordingly, “reaction time” is defined as the period for which the solution is held at a specified temperature. For each analysis, it was necessary to stop the reaction by cooling the autoclave to enable the hydrogen content of the headspace gas to be measured as described in section 2.5. Each data point in Figures 2
7 therefore represents an individual autoclave run carried out for the nominated reaction time in the range 0
120 min. 2.5. Gas Analysis. In order to measure the hydrogen content at the end of each autoclave run, the internal cooling coil circuit of the autoclave was activated to rapidly reduce the temperature. For example, this would cool the contents from 250 to 175 °C within 2 min and to 25 °C within 30 min. Once at 25 °C the autoclave was pressurized with a known amount of nitrogen to establish a controlled flow of the headspace gas to the hydrogen sensor. The internal pressure and temperature of the cooled autoclave were noted just before gas sampling. The measurements from the hydrogen analyzer were obtained in mole percent and were converted to moles of hydrogen using the general gas equation. Duplicate experiments were within (0.05 mol % hydrogen. The accuracy and precision of the gas analysis method were quantified by measurements of a range of hydrogen calibration gas mixtures and also by direct measurement of hydrogen production from the reaction of aluminum wire in sodium hydroxide solution as previously described.12 2.6. Solution Analysis. Two complementary techniques were used for the determination of the following LMW carboxylate 12325

dx.doi.org/10.1021/ie201387x |Ind. Eng. Chem. Res. 2011, 50, 12324–12333

Industrial & Engineering Chemistry Research

ARTICLE

Table 1. Aliphatic and Aromatic Compounds Used in the Current Work compound

structure H 3 CO2


formate

CH3 3 CO2


acetate

CH3 3 CH2 3 CO2


propanoate

CH3 3 (CH2)2 3 CO2


butanoate

CH3 3 (CH2)3 3 CO2


valerate

CH3 3 (CH2)4 3 CO2


caproate

CH3 3 (CH2)5 3 CO2


heptanoate

(CH3)2 3 CH 3 CO2
CO2
3 CO2


isobutyrate oxalate

CO2
3 CH2 3 CO2


malonate

CO2
3 (CH2)2 3 CO2


succinate

CO2
3 (CH2)3 3 CO2


glutarate

CO2
3 CH2 3 CH(CO2
) 3 CH2 3 CO2
C6H5 3 CO2


tricarballylate benzoate

C6H4 3 (CO2
)2

phthalate (1,2-benzenedicarboxylate)

C6H4 3 (CO2
)2 C6H4 3 (CO2
)2

isophthalate (1,3-benzenedicarboxylate) terephthalate (1,4-benzenedicarboxylate)

C6H3 3 (CO2
)3

trimellitate (1,2,4-benzenetricarboxylate)

C6H3 3 (CO2
)3

trimesate (1,3,5-benzenetricarboxylate)

C6H2 3 (CO2
)4 CH2(OH) 3 CO2


pyromellitate (1,2,4,5-benzenetetracarboxylate) glycolate

CH3 3 CH(OH) 3 CO2


lactate

CH3 3 CH2 3 CH(OH) 3 CO2


2-hydroxybutanoate

CH3 3 CH(OH) 3 CH2 3 CO2
CO2
3 CH(OH) 3 CH2 3 CO2


3-hydroxybutanoate malate

CO2
3 (CH(OH))2 3 CO2


D-tartrate

CO2
3 (CH(OH))2 3 CO2


L-tartrate

CO2
3 (CH(OH))2 3 CO2


meso-tartrate

CH2(OH) 3 (CH(OH))3 3 CO2


ribonate

CO2
3 (CH(OH))4 3 CO2


galactarate

CO2
3 CH2 3 C(CO2
)(OH) 3 CH2 3 CO2


citrate gluconate 16-hydroxyhexadecanoate

CH2(OH) 3 (CH(OH))4 3 CO2
CH2(OH) 3 (CH2)14 3 CO2


catechol (1,2-dihydroxybenzene)

C6H4 3 (OH)2

resorcinol (1,3-dihydroxybenzene)

C6H4 3 (OH)2

CO2
3 C6H4 3 OH

salicylate (2-hydroxybenzoate)

CO2
3 C6H4 3 OH

m-salicylate (3-hydroxybenzoate) 3,4-dihydroxybenzoate

CO2
3 C6H3 3 (OH)2

pyrogallol (1,2,3-trihydroxybenzene)

C6H3 3 (OH)3

CO2
3 C6H2 3 (OH)3

gallate (3,4,5-trihydroxybenzoate)

products: formate, acetate, oxalate, malonate, succinate, and lactate.41 Capillary electrophoresis was performed on a BeckmanCoulter MDQ instrument using a diethanolamine-molybdate buffer and a fused-silica capillary.42 Ion chromatography was performed on a Dionex dual ICS 3000 instrument, equipped with eluent generators, electrolytic suppressors, and matrix elimination by ion-trapping. Molar yields were calculated as the number of moles of product formed per mole of organic compound used.

3. RESULTS AND DISCUSSION 3.1. Screening Study. Figure 1 shows the production of hydrogen and LMW carboxylates during the anaerobic degradation of 40 organic compounds at 250 °C. The results are grouped

according to the structural classes of the compounds. The C1
C7 aliphatic compounds without hydroxy groups are relatively stable under these conditions. They produce little or no hydrogen gas, and the observed LMW carboxylate products account for 94% of the original carbon in all cases. 12327

dx.doi.org/10.1021/ie201387x |Ind. Eng. Chem. Res. 2011, 50, 12324–12333

Industrial & Engineering Chemistry Research

ARTICLE

Figure 2. Number of moles of H2 produced per mole of the organic compound and the corresponding H2 headspace concentrations. nH2/norganic compound is constant for concentrations up to 80 mM. Conditions: temperature = 250 °C, time at temperature = 15 min, [NaOH] = 3.77 M.

3.2.1. Effect of Initial Concentration. The effect of the initial concentration of the parent organic compound on the production of hydrogen from 3-hydroxybutanoate and malate is shown in Figure 2. The hydrogen concentration in the headspace increases linearly with the initial concentration of the organic compound as expected, and the number of moles of hydrogen produced per mole of the parent organic compound is constant. This indicates that the degradation mechanism remains unchanged over the range of concentrations studied and that the hydrogen production capacity of these compounds is independent of the initial concentration of the parent compound. It should be noted that increasing the solution volume in the autoclave (and hence decreasing the headspace fraction) would increase the hydrogen concentrations in the headspace correspondingly. This has important implications for the safe operation of wet oxidation reactors which contain oxygen because of the potential for oxygen and hydrogen to form explosive mixtures. 3.2.2. Reaction Kinetics. As noted above, the number of moles of hydrogen produced per mole of parent compound is constant throughout the reaction for both 3-hydroxybutanoate and malate. In addition, after complete degradation the amount of hydrogen produced is almost exactly equal to the initial number of moles of parent compound in both cases. Therefore, the concentration of the parent compounds at any time during the reaction may be calculated from hydrogen measurements, as follows: ½At ¼ ½Ao
nt =V

ð1Þ

where: [A]t = concentration of parent compound, A, in mol L
1, at time t; [A]o = concentration of parent compound, A, in mol L
1, at time t = 0; nt = number of moles of H2 produced in time t; V = volume of solution (L) The dependence of [A] on time follows a first-order law for both compounds, as demonstrated by Figure 3, which shows the results of experiments in which the rate of degradation of each compound was calculated in this way as a function of time at 250 °C. The molar ratio of compound A at time t to its initial concentration is defined as r ¼ ½At =½Ao

Figure 3. First-order plots calculated using eq 1 for the degradation of 3-hydroxybutanoate and malate. Conditions: temperature = 250 °C, time at temperature = 0
60 min, [NaOH] = 3.77 M.

Figure 4. Number of moles of H2 and LMW carboxylate products produced per mole of (A) 3-hydroxybutanoate (3-Hb) and (B) malate. Conditions: temperature = 250 °C, time at temperature = 0
120 min, [NaOH] = 3.77 M.

From these results it follows that the reaction rates may be described by a first order rate law of the form: d½A=dt ¼
k½A where k is the first-order rate constant. The results show that 3-hydroxybutanoate degrades almost three times faster than malate (Figure 3). 3.2.3. Evolution of Products. The evolution of the principal degradation products of 3-hydroxybutanoate and malate at 250 °C is shown in Figure 4. The results are presented in terms of the ratio of the number of moles of each degradation product to the initial number of moles of parent organic compound in each case. 12328

dx.doi.org/10.1021/ie201387x |Ind. Eng. Chem. Res. 2011, 50, 12324–12333

Industrial & Engineering Chemistry Research

ARTICLE

Figure 6. Arrhenius plot for the degradation of 3-hydroxybutanoate and malate. Conditions: temperature = 200
250 °C, time at temperature = 15 min, [NaOH] = 3.77 M.

Figure 5. Number of moles of H2 and LMW carboxylate products produced per mole of (A) 3-hydroxybutanoate (3-Hb) and (B) malate. Conditions: temperature = 175
275 °C, time at temperature = 15 min, [NaOH] = 3.77 M.

The solid lines represent the number of moles of hydrogen produced according to a first-order law with the rate constants derived from Figure 3. For both organic compounds, these lines show an excellent fit to the experimental data provided that the starting time is adjusted to allow for a degree of reaction occurring during the heat-up period. The best fit was found by offsetting the measurements by 4
5 min, indicating that the heat-up period was equivalent to 4
5 min at temperature for both compounds. The dotted lines representing the concentrations of the parent organic compounds as a function of time were calculated from the amount of hydrogen produced assuming the 1:1 molar ratio discussed above. For 3-hydroxybutanoate (Figure 4A), the only major product other than hydrogen is acetate, accompanied by lesser amounts of formate and oxalate. The molar ratio of the sum of the organic products to the hydrogen produced (and 3-hydroxybutanoate consumed) is slightly above 2:1. Acetate accounts for about 90% of the organic products, with about 8% formate and 2% oxalate. These ratios are essentially constant throughout the reaction period. In the degradation of malate (Figure 4B), the molar sum of organic reaction products is almost exactly equal to twice the amount of hydrogen produced throughout the reaction period. In this case, only 1 mol of acetate is produced per mole of parent compound degraded, with slightly less than 1 mol of oxalate, which appears to be compensated for by a corresponding production of formate. 3.2.4. Effect of Temperature. The dependencies of the product yields on temperature are shown in Figure 5. For 3-hydroxybutanoate degradation, little reaction occurs below 200 °C (Figure 5A).

After 15 min at 250 °C, the observed LMW products accounted for 94% of the parent carbon, indicating that the reaction is essentially complete. Acetate and hydrogen are the major products of 3-hydroxybutanoate degradation. The ratio of the molar yields of acetate to hydrogen is close to 2:1 over the entire temperature range, indicating that these products are formed through a common degradation pathway. Only minor amounts of formate and oxalate were produced. Similarly, for malate degradation, little reaction occurs below 200 °C (Figure 5B). After 15 min at 250 and 275 °C, the observed LMW products accounted for, respectively, 46% and 94% of the initial carbon. Acetate, hydrogen, and oxalate are the dominant products of malate degradation, with the ratio of the molar yields of these products being close to 1:1:1. Minor amounts of formate were produced, particularly at 275 °C. The appearance of lactate at 200 °C is consistent with previous work at similar temperatures,49 although the formation of this product in only trace quantities implies that the direct decarboxylation of malate is relatively insignificant under the current conditions. 3-Hydroxybutanoate reacts more rapidly than the dicarboxylate malate and, after 15 min, produces higher yields of hydrogen and LMW products at lower temperatures. A range of β-hydroxycarboxylates will need to be investigated to understand the reasons for this reactivity difference. Both of the hydroxycarboxylates studied here show the same hydrogen production capacity, generating approximately 1 mol of hydrogen gas per mole of the organic compound consumed. The rate constants at different temperatures can be estimated from measurements of the extent of reaction at a fixed time provided that the rate constant has been measured for at least one temperature. This was achieved using Polymath version 6.10 software to create a model calibrated with the rate constants measured at 250 °C. The results are shown in Figure 6 in the form of an Arrhenius plot, from which the activation energies have been estimated. The activation energies for the reactions of 3-hydroxybutanoate and malate were found to be 140 and 144 kJ/mol, respectively, indicating that the rate-determining steps for alkaline degradation are similar for each compound. 3.2.5. Effect of NaOH Concentration. NaOH concentration was found to have a significant influence on the production of hydrogen from 3-hydroxybutanoate and malate, suggesting that the degradation reactions are base-catalyzed. A base-catalyzed mechanism has previously been shown to apply in the case of malonate.30 In that case, [OH
] appears as a factor in the first-order rate constant 12329

dx.doi.org/10.1021/ie201387x |Ind. Eng. Chem. Res. 2011, 50, 12324–12333

Industrial & Engineering Chemistry Research as follows: k ¼ ko ½OH
 Using the modeling approach described in section 3.2.4, the rate constant determined under one set of conditions can be used to calibrate a model which is used to predict the extent of reaction at any given time under any chosen set of conditions, assuming that the molar ratios of the organic products formed to the hydrogen produced are constant. This method has been used to compare the first-order model with measurements of the extent of reaction after 15 min at 275 °C for 3-hydroxybutanoate and malate as a function of [OH
]. This is shown in Figure 7, in which the curves were calculated using the pseudo-first-order model and the points are the experimental results. These results suggest a common rate-determining step involving a unimolecular reaction with OH
in each case. By analogy with the case of malonate,30 this probably involves the abstraction of a hydrogen atom and the formation of an equilibrium with the deprotonated species, R
, which is the reactive species that participates in the rate-determining step, as follows: RH þ OH h R
þ H2 O R
f products 3.3. Proposed Degradation Mechanisms. In this section a number of mechanisms are proposed that are consistent with the

Figure 7. Number of moles of H2 produced per mole of the organic compound. Conditions: temperature = 275 °C, time at temperature = 15 min, [NaOH] = 0
6 M.

ARTICLE

evidence so far, and which provide a basis for ongoing investigation which will be reported in due course. An ionic degradation mechanism, consistent with the formation and distribution of the observed dominant reaction products of 3-hydroxybutanoate (1) and malate (2), is illustrated in Scheme 1. These reactions are analogous to the degradation mechanisms proposed for other hydroxycarboxylates, such as tartrate and threonate, in alkaline solution.27,33 Abstraction of the hydrogen of the β-hydroxy group can result in cleavage of the carbon backbone of the parent compounds. In pathway A, which is common to both of the hydroxycarboxylates, an alkene species (3) reacts with water to produce acetate (4) and regenerates a hydroxyl ion. The initial cleavage reaction also produces acetaldehyde (5) and glyoxylate (6) from 3-hydroxybutanoate (1) and malate (2), respectively (pathway B). Subsequent nucleophilic attack by a hydroxyl ion on the carbonyl carbon results in an intermediate species (7, 8). These intermediate species undergo base-catalyzed oxidation by water, involving a hydride transfer to a water molecule, to produce acetate (4) or oxalate (9), hydrogen gas, and hydroxyl ion. On the basis of the mechanism proposed for the degradation of 3-hydroxybutanoate (1), the expected molar ratio of acetate to hydrogen is 2:1 and the formation of oxalate as a major product is not expected. On the basis of the reaction mechanism proposed for malate (2) degradation, the expected molar ratio of the products
acetate, hydrogen, and oxalate
is 1:1:1. For both of these compounds, 1 mol of hydrogen is produced per mole of the parent compound consumed. These mechanistic interpretations are consistent with the nature and amounts of the products observed as shown in Figures 4 and 5. The production of minor amounts of formate from 3-hydroxybutanoate (1) (Figures 4A and 5A) may occur via the formation of glyoxylate from the alkene intermediate ((3) in Scheme 1) as shown in Scheme 2. Decarboxylation of the glyoxylate (6) thus formed produces formaldehyde (12) which subsequently undergoes base-catalyzed oxidation by water to give formate (14) (Scheme 3). This mechanism is consistent with earlier studies in which hydrogen was observed as a product of the alkaline degradation of formaldehyde.50
53 The trace amounts of oxalate (9) produced in the degradation of 3-hydroxybutanoate could result from the base-catalyzed oxidation of glyoxylate (6) as shown in pathway B of Scheme 1 (where R = COO
). Formate (14) was also observed from the degradation of malate (2) at 275 °C (Figure 5B). The production of formate

Scheme 1. Proposed Degradation Mechanism for 3-Hydroxybutanoate (1) and Malate (2),49 in the Explanation of the Results Presented in Figures 4 and 5

12330

dx.doi.org/10.1021/ie201387x |Ind. Eng. Chem. Res. 2011, 50, 12324–12333

Industrial & Engineering Chemistry Research

ARTICLE

Scheme 2. Proposed Degradation Mechanism for the Production of Glyoxylate (6) from Alkene Intermediate (3) in the Degradation of 3-Hydroxybutanoate (1)

Scheme 3. Proposed Degradation Mechanism for the Production of Formate (14) from Glyoxylate (6)

(14) and hydrogen could be due to decarboxylation of glyoxylate (6) as proposed in Scheme 3. The formation of formate (14) from glyoxylate (6), a cleavage product of malate (Scheme 1, pathway B), may account for the slightly lower oxalate yield observed at 275 °C (Figure 5B) as glyoxylate is an intermediate species common to both products. Additional tests (not shown) on the alkaline degradation of glyoxylate under similar conditions confirmed the formation of hydrogen with oxalate as a major product and smaller amounts of formate. This result is consistent with the mechanisms proposed in Schemes 1 and 3. The mechanism proposed here for the degradation of the hydroxycarboxylates involves the transfer of a hydride ion, in which respect it is similar to the Cannizzaro reaction.54,55 However Cannizzaro reactions do not produce hydrogen gas, but are base-induced disproportionation reactions of aldehydes, for which the oxidation products are carboxylic acids and the reduction products are alcohols (reaction 2, Scheme 1 in the work of Loh et al.27). In the current work, the measured oxidation products accounted for all of the carbon in the parent compounds and this was balanced by an equivalent number of moles of hydrogen as the sole reduction product. Therefore no significant amounts of alcohols were produced as would be the case if a Cannizzaro reaction were occurring. The observed formation of hydrogen rather than alcohols as the reduction product is consistent with base-catalyzed oxidation by water as the reaction mechanism. According to the degradation mechanisms outlined in Schemes 1
3, half of the hydrogen gas is contributed by the reduction of water and the other half comes from the oxidation of the organic compound. These mechanisms suggest that it is the proton at the β-carbon position of 3-hydroxybutanoate and malate that produces the hydrogen through a hydride transfer to a water molecule. This observation is consistent with the hydrogen measurements from the alkaline degradation of citrate (a β-hydroxycarboxylate) shown in Figure 1. Acetate, formate, and oxalate were the major products, and the hydrogen yield was small. The absence of a proton at the β-carbon position of citrate accounts for the lower hydrogen yield measured in comparison with the other β-hydroxycarboxylates. Ashby and co-workers50

studied the mechanism for the production of hydrogen during the alkaline degradation of various aldehydes, including formaldehyde, glyoxylate, pivalaldehyde, and benzaldehyde. They suggested that for the degradation of formaldehyde, one hydrogen atom originates in the water and the other in the formaldehyde. Isotopic labeling studies have confirmed that while this is the exclusive route to H2 from glyoxylate, in the case of formaldehyde, it is possible for both atoms in H2 to originate from the organic moiety.53 In alkaline solutions, β-hydroxycarboxylates exist in equilibrium with their corresponding α,β-unsaturated carboxylates.56 The nature and amounts of products observed are the same when either the β-hydroxycarboxylates or their corresponding unsaturated carboxylates are used as the starting materials, consistent with the regioselective hydration of the CdC double bond. For example, 3-hydroxybutanoate (1) and 2-butenoate degrade to produce similar products; malate (2) and maleate degrade to produce similar products.57 3.4. Application of Base-Catalyzed Oxidation by Water to Previous Studies. The current work confirms that the production of hydrogen from the hydroxycarboxylates studied can occur via an ionic degradation mechanism involving base-catalyzed oxidation by water. These findings are also applicable to low-temperature alkali-promoted gasification of biomass58
61 in which hydroxycarboxylates21
23 (and other classes of compounds) are formed and which can subsequently react to generate hydrogen. For example, Ishida et al.58 have demonstrated the synthesis of COx-free hydrogen from the alkalipromoted reaction of steam with cellulose, glucose, sucrose, and starch below 300 °C, and Su et al.61 have shown that the product gases resulting from the base-catalyzed steam gasification of cellulose between 130 and 270 °C consist mainly of hydrogen. The absence of COx from the product gases in these studies confirms that the hydrogen produced is not due to the water-gas shift reaction (CO + H2O f CO2 + H2), which is the case at higher temperatures.61
63 The current work offers basecatalyzed oxidation by water as an explanation for the production of COx-free hydrogen during the alkali-promoted gasification of biomass at low temperature. 12331

dx.doi.org/10.1021/ie201387x |Ind. Eng. Chem. Res. 2011, 50, 12324–12333

Industrial & Engineering Chemistry Research

4. CONCLUSIONS This study is the first of a series aimed at developing a fundamental understanding of the reactions that produce hydrogen gas from the degradation of organic compounds under alkaline conditions. This will provide a basis for strategies to prevent or limit the formation of potentially explosive gas mixtures during Bayer digestion and industrial wet oxidation processes, and for enhancing the production of hydrogen in the gasification of biomass. The key findings from this initial study of a range of organic compounds, which are known to be produced during the alkaline degradation of humates, lignins, and carbohydrates and/or are present in Bayer liquors are: • The main hydrogen-producing compounds are straightchain, aliphatic hydroxycarboxylates possessing a single βhydroxy substituent, such as 3-hydroxybutanoate and malate. These compounds decompose to produce 1 mol of H2 per mole of the parent compound consumed in NaOH at elevated temperature (>200 °C). They also produce LMW carboxylate products, the nature of which depends on the parent compound. • LMW aliphatic and aromatic carboxylates without hydroxy groups which are commonly found in Bayer liquors produce little or no hydrogen gas under the oxygen-deficient conditions encountered in alumina refineries during digestion. Similarly, the aromatic hydroxycarboxylates and phenols studied were found to produce only small quantities of hydrogen. • The alkaline degradation of the β-hydroxycarboxylates studied follows pseudo-first-order reaction kinetics. 3-Hydroxybutanoate degrades almost three times faster than malate. • Quantitative analysis of the products of reaction of the β-hydroxycarboxylates studied is consistent with an ionic degradation mechanism involving base-catalyzed oxidation by water, which accounts for the observed stoichiometry of formation of the main LMW carboxylate products and hydrogen gas. • These findings provide an explanation for the production of COx-free hydrogen during the alkali-promoted gasification of biomass at low temperature, via base-catalyzed oxidation by water. ’ AUTHOR INFORMATION Corresponding Author

*E-mail address: [email protected]. Tel.: +61 8 9334 8031. Fax: +61 8 9334 8001. Notes † ‡

CSIRO Light Metals National Research Flagship. CSIRO Minerals Down Under Flagship.

’ ACKNOWLEDGMENT The authors thank Milan Chovancek for conducting the analyses of the LMW degradation products. The financial support of the Parker CRC for Integrated Hydrometallurgy Solutions (established and supported under the Australian Government’s Cooperative Research Centres Program) and CSIRO Light Metals National Research Flagship is gratefully acknowledged. ’ REFERENCES (1) Roskill, The Economics of Bauxite & Alumina, 7th ed.; Roskill Information Services Ltd: London, UK, 2008.

ARTICLE

(2) Armstrong, L.; Healy, S. J. Degradation of liquor organics during Bayer digestion cycles. Travaux 2006, 33 (37), 291–301. (3) Power, G.; Loh, J. Organic compounds in the processing of lateritic bauxites to alumina: Part 1: Origins and chemistry of organics in the Bayer process. Hydrometallurgy 2010, 105 (1
2), 1–29. (4) Power, G. P. The impact and control of organic compounds in the extraction of alumina from bauxite. 5th AusIMM Extractive Metallurgy Conference, Perth, Australia, Oct 2
4, 1991; Australian Institute of Mining and Metallurgy: Perth, Australia, 1991; pp 337
345. (5) Solymar, K.; Gimpel-Kazar, M.; Molnar, E. Determination and evaluation of organic balances of alumina refineries. Light Met. 1996, 29–35. (6) Brown, E. R.; Headley, A.; Greenaway, A. M.; Magnus, K. E. The effect of low molecular weight organics on precipitation efficiency. J. Geol. Soc. Jamaica, Proc. Bauxite Symp. VI 1986, 158–165. (7) Guthrie, J. D.; The, P. J.; Imbrogno, W. D. Characterization of organics in Bayer liquors. Light Met. 1984, 127–146. (8) Lever, G. Identification of organics in Bayer liquor. Light Met. 1978, 71–83. (9) Bhargava, S. K.; Tardio, J.; Prasad, J.; Foger, K.; Akolekar, D. B.; Grocott, S. C. Wet oxidation and catalytic wet oxidation. Ind. Eng. Chem. Res. 2006, 45 (4), 1221–1258. (10) Arnswald, W.; Kaltenberg, H. G.; Guhl, E. Removal of organic carbon from Bayer liquor by wet oxidation in tube digesters. Light Met. 1991, 23–27. (11) Lalla, A.; Arpe, R. 12 years of experience with wet oxidation. Light Met. 2002, 177–180. (12) Costine, A.; Schibeci, M.; Loh, J.; Power, G.; McDonald, R. Hydrogen production in Bayer process digestion. Travaux 2010, 35 (39), 369–375. (13) Graham, G.; Capil, R.; Davies, R. Odour destruction for digestion vent gases. 6th International Alumina Quality Workshop, Brisbane, Australia, Sept 8
13, 2002; AQW Inc: Brisbane, Australia, 2002; pp 316
319. (14) Anderson, J.; Armstrong, L.; Thomas, D. Wet oxidation in QAL [Queensland Aluminum Limited] high temperature digestion for plant productivity improvement. Light Met. 2001, 121–126. (15) Baker, C. L.; Lehmann, R. W.; Maugans, C. B.; Burnet, S.; Wellisch, K. F.; Lewi, C. Method of hydrothermal treatment of organic contaminants of alkaline Bayer process solutions. Patent no. AU2008229997, 2008. (16) Brown, N. Kinetics of copper-catalysed oxidation of Bayer liquor organics. Light Met. 1989, 121–130. (17) The, P. J.; Williams, F. S.; Guthrie, J. D. Mass transfer studies on wet oxidation of Bayer liquor. Pilot plant evaluation. Light Met. 1985, 103–116. (18) Galbally, I. Wagerup air quality review, Report C/0936, http:// www.epa.wa.gov.au/docs/1215/Response/AppC.pdf. (19) Lewis, B.; Elbe, G. V. Combustion, flames and explosions of gases, 3rd ed.; Academic Press: London, 1987. (20) H€aussinger, P.; Lohm€uller, R.; Watson, A. M. Hydrogen; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, 2007. (21) Knill, C. J.; Kennedy, J. F. Degradation of cellulose under alkaline conditions. Carbohyd. Polym. 2002, 51 (3), 281–300. (22) Niemel€a, K. The formation of hydroxy monocarboxylic acids and dicarboxylic acids by alkaline thermochemical degradation of cellulose. J. Chem. Technol. Biotechnol. 1990, 48 (1), 17–28. (23) Niemel€a, K.; Sj€ostr€om, E. The conversion of cellulose into carboxylic-acids by a drastic alkali treatment. Biomass 1986, 11 (3), 215–221. (24) Niemel€a, K.; Grocott, S. C. Determination of carboxylic acids in sodium aluminate liquor (Bayer liquor). International Conference on Biological Mass Spectrometry; Kyoto, Japan, Sept 20
24, 1992; pp 480
181. (25) Power, G.; Loh, J. S. C.; Niemel€a, K. Organic compounds in the processing of lateritic bauxites to alumina: Addendum to Part 1: Origins and chemistry of organics in the Bayer process. Hydrometallurgy 2011, 108 (1
2), 149–151. 12332

dx.doi.org/10.1021/ie201387x |Ind. Eng. Chem. Res. 2011, 50, 12324–12333

Industrial & Engineering Chemistry Research (26) Coyne, J. F.; Wainwright, M. S.; Cant, N. W.; Grocott, S. C. Adsorption of hydroxy organic-compounds on alumina-trihydrate. Light Met. 1994, 39–45. (27) Loh, J. S. C.; Brodie, G. M.; Power, G.; Vernon, C. F. Wet oxidation of precipitation yield inhibitors in sodium aluminate solutions: Effects and proposed degradation mechanisms. Hydrometallurgy 2010, 104 (2), 278–289. (28) Watling, H. Gibbsite crystallization inhibition - 2. Comparative effects of selected alditols and hydroxycarboxylic acids. Hydrometallurgy 2000, 55 (3), 289–309. (29) Watling, H.; Loh, J.; Gatter, H. Gibbsite crystallization inhibition 1. Effects of sodium gluconate on nucleation, agglomeration and growth. Hydrometallurgy 2000, 55 (3), 275–288. (30) Dong, J.; Power, G.; Loh, J.; Tardio, J.; Vernon, C.; Bhargava, S. Fundamentals of wet oxidation of Bayer process Liquor: Reactivity of malonates. Ind. Eng. Chem. Res. 2010, 49 (11), 5347–5352. (31) Dong, J.; Tardio, J.; Loh, J.; Power, G.; Vernon, C.; Bhargava, S. Wet oxidation of Bayer liquor organics: Reaction mechanisms. Light Met. 2010, 209–213. (32) Tardio, J.; Bhargava, S.; Prasad, J.; Akolekar, D. B. Catalytic wet oxidation of the sodium salts of citric, lactic, malic and tartaric acids in highly alkaline, high ionic strength solution. Top. Catal. 2005, 33 (1
4), 193–199. (33) Loh, J.; Brodie, G.; Power, G. P.; Vernon, C. Mechanisms of degradation of hydrate yield inhibitors by wet oxidation. 8th International Alumina Quality Workshop, Darwin, Australia, Sept 7
12, 2008; AQW Inc: Darwin, Australia, 2008; pp 199
205. (34) Loh, J.; Brodie, G.; Power, G.; Vernon, C. The effects of wet oxidation on hydrate yield inhibitors in alkaline aluminate solutions. Light Met. 2008, 181–186. (35) Power, G. P.; Tichbon, W. Sodium oxalate in the Bayer process: its origin and effects. 2nd International Alumina Quality Workshop, Perth, Australia, Oct 14
19, 1990; AQW Inc: Perth, Australia, 1990; pp 99
115. (36) Yamada, K.; Harato, T.; Kato, H. Oxidation of organic substances in the Bayer process. Light Met. 1981, 117–128. (37) Korotcenkov, G.; Do Han, S.; Stetter, J. R. Review of electrochemical hydrogen sensors. Chem. Rev. 2009, 109 (3), 1402–1433. (38) Costine, A.; Loh, J.; McDonald, R.; Power, G. Unique hightemperature facility for studying organic reactions in the Bayer process. Light Met. 2010, 47–52. (39) Ellis, A. V.; Wilson, M. A.; Forster, P. Bayer poisons: Degradation of klason lignin in sodium hydroxide at 145 degrees C. Ind. Eng. Chem. Res. 2002, 41 (25), 6493–6502. (40) H€anninen, K.; Niemel€a, K. Alkaline-degradation of peat humic acids. 2. Identification of hydrophilic products. Acta Chem. Scand. 1992, 46 (5), 459–463. (41) Power, G.; Loh, J. S. C.; Wajon, J. E.; Busetti, F.; Joll, C. A review of the determination of organic compounds in Bayer process liquors. Anal. Chim. Acta 2011, 689 (1), 8–21. (42) Chovancek, M.; Choo, P.; Macka, M. Development of a fully buffered molybdate electrolyte for capillary electrophoresis with indirect detection and its use for analysis of anions in Bayer liquor. Electrophoresis 2004, 25 (3), 437–443. (43) Tardio, J.; Bhargava, S.; Eyer, S.; Sumich, M.; Akolekar, D. B. Interactions between specific organic compounds during catalytic wet oxidation of Bayer liquor. Ind. Eng. Chem. Res. 2004, 43 (4), 847–851. (44) Oakwood, T. S.; Miller, M. R. The decarboxylation of simple fatty acids. J. Am. Chem. Soc. 1950, 72 (4), 1849–1849. (45) Picard, F.; Audet, D.; Boily, H.; Larocque, J. Identification of hydrate active organics (HAO) present in spent Bayer liquors by stateof-the-art analytical methods. 6th International Alumina Quality Workshop, Brisbane, Australia, Sept 8
13, 2002; AQW Inc: Brisbane, Australia, 2002; pp 46
53. (46) Mishra, V. S.; Mahajani, V. V.; Joshi, J. B. Wet air oxidation. Ind. Eng. Chem. Res. 1995, 34 (1), 2–48. (47) Luo, Y.-R. Comprehensive handbook of chemical bond energies; CRC Press: Boca Raton, FL, 2007.

ARTICLE

(48) Machold, T.; Laird, D. W.; Rowen, C. C.; May, P. M.; Hefter, G. T. Decomposition of Bayer process organics: Phenolates, polyalcohols, and additional carboxylates. Hydrometallurgy 2011, 107 (3
4), 68–73. (49) Machold, T.; Macedi, E.; Laird, D. W.; May, P. M.; Hefter, G. T. Decomposition of Bayer process organics: Low-molecular-weight carboxylates. Hydrometallurgy 2009, 99 (1
2), 51–57. (50) Ashby, E. C.; Doctorovich, F.; Liotta, C. L.; Neumann, H. M.; Barefield, E. K.; Konda, A.; Zhang, K.; Hurley, J.; Siemer, D. D. Concerning the formation of hydrogen in nuclear waste - quantitative generation of hydrogen via a Cannizzaro intermediate. J. Am. Chem. Soc. 1993, 115 (3), 1171–1173. (51) Bi, Y.; Hu, H.; Li, Q.; Lu, G. Efficient generation of hydrogen from biomass without carbon monoxide at room temperature - Formaldehyde to hydrogen catalyzed by Ag nanocrystals. Int. J. Hydrogen Energy 2010, 35 (13), 7177–7182. (52) Kapoor, S.; Barnabas, F. A.; Sauer, M. C.; Meisel, D.; Jonah, C. D. Kinetics of hydrogen formation from formaldehyde in basic aqueous-solutions. J. Phys. Chem. 1995, 99 (18), 6857–6863. (53) Kapoor, S.; Naumov, S. On the origin of hydrogen in the formaldehyde reaction in alkaline solution. Chem. Phys. Lett. 2004, 387 (4
6), 322–326. (54) Cannizzaro, S. Ueber den der Benzo€es€aure entsprechenden Alkohol. Ann. Chem. 1853, 88, 129–130. (55) Geissman, T. A. The Cannizzaro Reaction, Organic Reactions; John Wiley and Sons: New York, 1944. (56) Miltenberger, K. Hydroxycarboxylic acids, aliphatic. In Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, 2005. (57) Costine, A.; Loh, J. S. C.; Power, G. Understanding hydrogen in Bayer process emissions: 2. Hydrogen production during the degradation of unsaturated carboxylic acids in sodium hydroxide solutions. Ind. Eng. Chem. Res. 2011 DOI: 10.1021/ie201388c. (58) Ishida, M.; Otsuka, K.; Takenaka, S.; Yamanaka, I. One-step production of CO- and CO2-free hydrogen from biomass. J. Chem. Tech. Biotech. 2005, 80 (3), 281–284. (59) Muangrat, R.; Onwudili, J. A.; Williams, P. T. Influence of alkali catalysts on the production of hydrogen-rich gas from the hydrothermal gasification of food processing waste. Appl. Catal., B 2010, 100 (3
4), 440–449. (60) Onwudili, J. A.; Williams, P. T. Role of sodium hydroxide in the production of hydrogen gas from the hydrothermal gasification of biomass. Int. J. Hydrogen Energy 2009, 34 (14), 5645–5656. (61) Su, S.; Li, W.; Bai, Z.; Xiang, H.; Bai, J. Effects of ionic catalysis on hydrogen production by the steam gasification of cellulose. Int. J. Hydrogen Energy 2010, 35 (10), 4459–4465. (62) Ratnasamy, C.; Wagner, J. P. Water Gas Shift Catalysis. Catal. Rev. 2009, 51 (3), 325–440. (63) Tanksale, A.; Beltramini, J. N.; Lu, G. M. A review of catalytic hydrogen production processes from biomass. Renew. Sust. Energ. Rev. 2010, 14 (1), 166–182.

12333

dx.doi.org/10.1021/ie201387x |Ind. Eng. Chem. Res. 2011, 50, 12324–12333