Chemical processes from saccharides into lactate were also reported in alkaline aqueous solutions [27,28] or in hot water with or without metal salts [21,22,29].
These chemical processes are simpler than the biotechnological processes, flexible as far as the limitation of reaction conditions, and can be carried out by using conventional chemical plants, is concerned, which indicates the possibility to increase the lactate supply. The reaction mechanism of alkaline degradation of glucose into lactate includes a reverse aldol condensation from C6compounds to C3 compounds, as shown in Fig.4.3 [21, 30]. Our group focused on activated hydrotalcite, which has been reported to have Brứnsted-base sites and to show catalytic activity for aldol condensations [31–33], which are inverse reactions of the reverse aldol condensation in Fig.4.3. Consequently, our group first demon- strated the transformation of D-glucose to produce lactic acid over activated hydrotalcite catalysts and the determination of accessible Brứnsted-base sites on the catalyst.
4.5.1.1 Lactic Acid Production from Glucose Over Activated Hydrotalcite Solid Base Catalysts
Our group has undertaken the present work with the aim of studying the trans- formation of D-glucose to produce lactic acid over activated hydrotalcite catalysts and the determination of accessible Brứnsted-base sites on the catalyst [34].
Solid Brứnsted Base Hydrotalcite catalysts
Hydrotalcite-like compounds, also known as Layered double hydroxides of clay- like minerals, have recently received much attention [35] in view of their potential usefulness as adsorbents, anion exchangers, and basic catalysts [32]. These com- pounds are represented by the formula [M(1-x)2+ Mx3+(OH)2]x+(Ax/mm-)nH2O, typically [Mg(1-x)2+ Alx3+(OH)2]x+(CO3x/22-)nH2O (x=0.2-0.4). Hydrotalcite (HT) was prepared by the method described previously [33]. A mixed aqueous solution of Mg(NO3)26H2O and Al(NO3)39H2O was added slowly to a mixed aqueous solu- tion of NaOH and Na2CO3. The obtained as-synthesized hydrotalcite sample (HTas) had the Mg/Al atomic ratio equal to 2.1 (Table4.1), which was close to the ratio in the mother solution, and had a HT-platelet structure according to X-ray diffraction (Fig.4.4a). It was heated in an inert gas flow to 573 and 723 K to prepare the heated samples (HT573 and HT723). The calcination at 723 K resulted in marked increases of the surface area and the total pore volume, to 237 m2 g-1 and 0.52 cc g-1 (Table4.1), and the disappearance of the layered structure. The high surface area and large pore volume of HT723 might be attributed to the removal of interlayer water molecules and carbon dioxide from the carbonate anion present in the brucite layer [33,36–38]. The heated samples were rehydrated by distilled deionized water at room temperature and dried at room temperature under vacuum, which resulted in the rehydrated samples, HT573rehydrand HT723rehydr, generating the reconstruction
of the layered HT structure (Fig.4.4b). The layered structure is recovered to a large extent with hydroxyl ions (Brứnsted-base sites) incorporated in the interlayer [31,32,39]. The images of as-synthesized hydrotalcite, heated hydrotalcite, and rehydrated hydrotalcite are shown in Fig.4.5.
The accessible Brứnsted-base sites on these hydrotalcite catalysts were deter- mined by the ion-exchange method and the conventional CO2adsorption method.
An aqueous solution of sodium gluconate or sodium chloride (0.20 mol L-1) was added to slurry of the HT sample with a small amount of water. The OH-ions at Brứnsted-base sites of the HT are ion-exchanged with gluconate or chloride ions and the corresponding number of sodium hydroxide ions is formed in the solution.
After filtration, the produced sodium hydroxide was determined by titration with an aqueous HCl solution. These basic sites determined by the ion-exchange method with sodium gluconate might be accessible for glucose. As shown in Table4.1, the molar ratio of exchangeable OH-ions to aluminum atoms in the rehydrated HTas(HTas-rehydr) was estimated to be about 2 %. The number of OH- sites of HT573rehydr, whose OH/Al molar ratio was 3 %, is almost double those of HTas. In contrast, the number of OH- sites of HT723rehydr, whose OH/Al molar Table 4.1 The relationships between basic sites and lactic acid production on solid catalysts Catalyst Basic sites Lactic acid yielda Sugar conversiona
Titration CO2adsorption
OH /Al OH /Al C-% %
Control – – 0.4 4
HTas 0.017 0.009 0.9 5
HT623 0.034 0.035 1.4 10
HT723 (0.3 g) 0.36 0.038 11 28
HT723 (0.6 g) – – 20 57
MgO – – 1.7 8
c-Al2O3 – – 0.4 5
aReaction conditions: glucose 25 mmol L-1, NaOH 50 mmol L-1, flow rate 0.1 mL min-1, 323 K, time on stream 8 h, catalyst 300 mg. Reprinted from Ref. [34], Copyright 2008, with permission from Elsevier
Fig. 4.4 aMg6Al2(OH)16CO34H2O: Mg-Al hydrotalcite with carbonate anions, andbactivated Mg-Al hydrotalcite with Brứnsted-basic hydroxyl ions
ratio was 36 %, increased more than ten times. The determined OH/Al molar ratios were less than 40 %, which would indicate that gluconate ions might not com- pletely intercalate into the interlayers of the HT platelets. The X-ray diffraction peaks of HT723NaCl (Fig.4.4d) and HT723Na_gluconate (Fig.4.4e) were broad.
During the ion-exchange, both HT samples were white powders but not or little dissolve in the aqueous solutions. The ion-exchange with sodium salts was reversibly at least six times using NaOH aqueous solution.
The rehydrated HT samples (HTas, HT573rehydr, and HT723rehydr) were dried at 393 K under vacuum, and then the base sites were determined by the conventional CO2adsorption method in a glass vacuum system [33]. The both dried HT573rehydr and dried HT723rehydr samples have almost the same number of accessible basic sites for CO2as shown in Table4.1. The number of accessible basic sites of dried HT573rehydr for CO2was almost the same as that of HT573 for anions in water phase. It is considered that a certain number of interlayer Brứnsted-base sites of the HT platelets are accessible in water, although only Brứnsted-base sites near extra- surface of the HT platelets are accessible in gas phase.
Catalytic conversions of glucose with or without NaOH
The reactions of D-glucose were done in the aqueous solution with D-glucose (25 mmol L-1) over pretreated HT catalysts at 323 K. The HTassamples (0.30 g) were in situ pretreated at 323 (HTas), 573 (HT573), or 723 K (HT723) under flowing nitrogen gas in a fixed-bed flow reactor. An aqueous solution of D-glucose (25 mmol L-1) with or without NaOH (50 mmol L-1) was introduced to the reactor.
In the case of reactions without NaOH, the HT723 catalyst showed significantly catalytic activity for the lactate production. However, magnesium ions were simultaneously eluted and the catalytic activity decreased immediately, which might be due to the acidic property of produced lactic acid and the coking of products, respectively. In the case of fermentations, a stoichiometric amount of Fig. 4.5 XRD patterns of
hydrotalcite samples.
Reprinted from Ref. [34], Copyright 2008, with permission from Elsevier
alkali is added for pH-regulation [10]. Therefore, low concentration of alkali was introduced into the reactant solution for the inhibition of leaching with produced organic acids and for the removal of products from the HT surface. The concen- tration of alkali depended on the concentration of sugar as substrate, because one D-glucose molecule can be converted into two lactic acid molecules to be neu- tralized by alkali.
Effects of NaOH concentration on the glucose conversion over solid basic catalysts
Figure4.6shows effects of NaOH concentration on the lactic acid formation with or without HT723 catalyst. In the case of reactions without solid catalyst, the lactic acid formation rate was quite low even in 0.2 mol L-1NaOH aqueous solution. In the case of reactions with HT723 catalyst, the lactic acid formed even in neutral water, and its formation rate was markedly higher than those of reactions without solid catalyst in aqueous solutions with more than 0.02 mol L-1NaOH concentration. In addition, no leaching of magnesium and aluminum ions in the filtrate after reactions with NaOH was detected by ICP measurement. The NaOH would play the role to neutralize produced lactic acids and to remove products from the surface, and also it might be like a catalytic promoter as supplying hydroxyl group and its concentrated adsorption at the active sites of the HT723 catalysts during the reaction.
Reaction conditions: glucose 25 mmol L-1, flow rate 0.1 mL min-1, 323 K, time on stream 8 h. catalyst 300 mg.
In the case of reactions with NaOH (0.05 mol L-1) aqueous solution as shown in Table4.1, the blank reaction show that the conversions of D-glucose were less than 5 % and the yield of lactic acid was less than 0.5 C-% for the reaction of D-glucose. The HTas and HT573 catalysts showed the small catalytic activity as well as MgO and Al2O3catalysts. In contrast, the HT723 catalyst showed sig- nificant activity for the lactic acid production from glucose. The lactate yield was 11 C-%. A linear dependence of lactic acid yield on the catalyst content was also obtained. The other products were unknown but water soluble organic compounds, Fig. 4.6 Lactic acid
production from glucose without (a) or with activated hydrotalcite HT723 catalyst (b)
because after reactions the TOC values of all solutions were more than 90 C-% of the reactant D-glucose solution.
At 30 h of time-on-stream over the HT723 catalyst (300 mg), the lactate yield was 7.5 C-% and the catalyst changed the color from white to brown. Although the catalyst was deactivated probably by the coking which might be due to by-products via the aldol condensation reactions, the TON was more than three. Consequently, it was clarified that the activated hydrotalcite has catalytic activity for the lactic acid production from D-glucose as a solid base catalyst.
4.5.1.2 Glucose to Lactic Acid and Other Useful Compounds Under Alkaline Solutions with Solid Catalyst
The significant disadvantage of the chemical processes described above is the large amount of by-products whose total yields is about 40 C-% or more. Many kinds of compounds were reported as the by-products, such as carboxylic acids, C6isomers, acidic aldol condensation products, dehydration products, and unidentified com- pounds. Although it is difficult to increase the yield of lactic acid to almost 100 C-%
in the alkaline degradation of monosaccharides, it is concerned to be able to pro- duce lactic acid and a useful chemical compound simultaneously, instead of complex by-products.
So, our group demonstrated a one-pot reaction for production of lactic acid and gluconic acid from D-glucose using metal supported catalysts with an alkaline aqueous solution, which is expected to be a new highly atomic-effective chemical process from renewable saccharides into useful chemicals, that is, the aim of this study is to reduce complex by-products in the alkaline degradations of D-glucose (Fig.4.7) [40]. Gluconic acid is used in the food and pharmaceutical industries and a biodegradable chelating agent in admixing concrete and in removing calcareous and rust deposits from metal surfaces. Various works dealt with the study of Fig. 4.7 Reaction of D-glucose in the conventional alkaline degradation (a) and in the proposed new chemical process (b)
glucose oxidation into gluconic acid over platinum, palladium, gold, and these bimetallic catalysts [41–46].
Reaction procedures
The metal-supported catalysts were prepared by an impregnation method. The metal contents were 5 wt%. Activated carbon, silica-gel, magnesia, andcsilica-alumina were used as supports. After drying overnight at 333 K, the supported catalysts were treated under flowing hydrogen at 573 K. Catalytic reactions were performed in a polypropylene copolymer (PPCO) batch reactor (40 ml) equipped with two tubes for gas inlet and gas outlet of reflux condenser. The typical reaction procedure was as follows: 5.0 mL of NaOH (1.0 mol L-1) aqueous solution was introduced in the reactor. The solution was stirred, and it was bubbled by flowing air at 353 K. And then, 45 mg of D-glucose and 50 mg of a solid catalyst were introduced simulta- neously, and the reaction was started. After reaction, the aqueous solution was separated from the solid catalyst by the filtration and cooled in an ice bath.
Effects of catalytic supports
Table4.2shows the product selectivity for the reactions of D-glucose over plati- num-supported catalysts in the 1.0 M NaOH aqueous solution at 353 K for 2 h. The reaction without metal catalysts (control) resulted in 55 C-% yield of lactic acid, which was the alkaline degradation catalyzed by sodium hydroxide. The total yields of lower organic acids, as acetic, glycolic and formic acids, was about 7 C-%, and the yields of others was about 38 C-% which were unidentified by HPLC and GC-MS analysis. The TOC value of the resultant aqueous solution was almost completely equivalent to the introduced glucose, which indicated that most of by-products were water soluble organic compounds.
To increase the atomic-efficiency for the reaction of glucose into useful chemicals, with remaining the high yield of lactic acid, platinum-supported cat- alysts were added as solid oxidation catalysts in the alkaline solution under air Table 4.2 The effects of support materials on the reaction of D-glucose over 5 wt% platinum catalystsa
Catalysts Surface area
Conversion Yield /C-%
m2/g /% GluA LA GlyA AceA ForA Others
Control – 100 – 55 1.7 2.7 2.5 38
Pt/Al2O3 136 100 13 57 1.3 2.4 1.3 25
Pt/SiO2 404 100 17 50 1.9 2.6 2.3 25
Pt/MgO 4.2 100 23 57 1.6 2.5 1.1 15
Pt/C 1200 100 45 43 2.9 2.9 0.8 6
Reprinted from Ref. [40], Copyright 2008, with permission from Elsevier
GluA (gluconic acid), LA (lactic acid), GlyA (glycolic acid), AceA (acetic acid), ForA(formic acid)
aGlucose (0.25 mmol) were added to 5.0 mL of NaOH (1.0 mol L-1) aqueous solution with 50 mg of catalyst in a batch reactor bubbled by air flow (20 mLmin-1). Reaction time and temperature were 2 h and 353 K, respectively
bubbling. The simultaneous formations of D-gluconic acid and D, L-lactic acid were expected. The platinum contents on various supports were about 5 wt%.
Table4.2shows the yield of lactic acid was between 43 and 57 C-%, which little depended on the kinds of supports. In contrast, although the yield of gluconic acid increased with the addition of any platinum catalysts, it significantly depended on the kinds of supports and increased as follows:
Pt/Al2O3 \Pt/SiO2 \Pt/MgO\Pt/C:
The yield of unknown by-products (others) decreased in the above order. The selectivity might be due to the oxidation activity of platinum catalysts. According to the XRD data in Fig.4.8, the dispersion of platinum on supports probably increased with the following order;
Pt/MgO\Pt/Al2O3; Pt/SiO2 \Pt/C:
It was considered that the high activity of Pt/C might be due to relatively high dispersion of platinum on activated carbon with high surface area. However, the dispersion of platinum could not account for that Pt/MgO catalyst showed the relatively high yield of gluconic acid among the metal-oxide support we tested.
The basicity of magnesium oxide and/or the interactions between platinum and supports might affect the catalytic activity. Consequently, the addition of Pt/C catalyst into the alkaline degradation under air resulted in the highest atomic- efficiency for the conversion of D-glucose into 43 C-% yield of lactic acid and 45 C-% yield of gluconic acid.
Fig. 4.8 Changes in product distribution during the reaction of glucose using Pt/C catalyst with 1.0 M NaOH at 353 K. Reaction conditions: D-glucose 0.25 mmol, Catalyst 50 mg, NaOH aqueous solution 5.0 mL, Air flow 20 mL min-1. (Filled square) lactic acid; (Filled triangle) gluconic acid; (9) C2carboxylic acids; (+) formic acid; (Open circle) others. Reprinted from Ref. [40], Copyright 2008, with permission from Elsevier
Changes in the product distribution as a function of reaction time
Figure4.8 shows the product distribution as a function of reaction time in the reaction of D-glucose over Pt/C catalyst in a 1.0 mol L-1 of NaOH aqueous solution at 353 K. At 10 min of reaction time, D-glucose disappeared completely and the yield of lactic acid, gluconic acid, and others were 39, 24, and 33 C-%, respectively. The TOC value of the resultant aqueous solution was equivalent to the introduced glucose, which indicated that most of others were water soluble.
Between the reaction time of 20 and 40 min, the product distribution was almost constant. After the reaction time of 40 min, the yield of gluconic acid increased from 24 C-% (40 min) to 40 C-% (60 min) and to 49 C-% (180 min). As shown in Fig.4.9, the yield of others decreased with increase of the yield of gluconic acid, which suggested the conversion of the others into gluconic acid. The yield of lactic acid was almost constant to be 43 C-% after the reaction time of 20 min. At 180 min of reaction time, the yield of lactic acid and gluconic acid was 42 and 49 C-%, respectively, and the yields of lower carboxylic acids (glycolic, acetic, and formic acids) and others were about 6 C-% and about 3 C-%, respectively.
Most of the others were water soluble organic compounds (WSOCs), and they seemed to form with the formation of lactic acid and were further converted into gluconic acid after an induction period between 10 and 40 min. Probably, during the induction period, WSOCs (A) formed with lactic acid from D-glucose might be converted into WSOCs (B) easily oxidized into gluconic acid. Both WSOCs were not identified by HPLC and GC-MS, which were not oligosaccharides and sugar alcohols. In the previous reports, the alkaline degradation of monosaccharides results in the formation of unidentified products, besides of the well-characterized carboxylic acids equal and lower than C6, which can be assigned as carboxylic Fig. 4.9 Effect of reaction temperature on product distribution for the reaction of glucose using the Pt/C catalyst with 1.0 M NaOH for 2 h. Reaction conditions: D-glucose 0.25 mmol, Catalyst 50 mg, NaOH aqueous solution 5 ml, Air flow 20 mL min-1. (j) lactic acid; (m) gluconic acid;
(9) C2carboxylic acids; (+) formic acid; (s) others. Reprinted from Ref. [40], Copyright 2008, with permission from Elsevier
acids higher than C6[27]. In this study, the others were highly selectively oxidized into gluconic acid, which suggested that the others might have similar composition to glucose and gluconic acid, such as these dimmers and/or isomers.
Effects of reaction temperature on the product distribution
Figure4.9shows the product distribution as a function of the reaction temperature in the reaction of D-glucose using Pt/C catalyst in 1.0 mol L-1of NaOH aqueous solution for 2 h. The TOC values of the resultant aqueous solutions after the reaction at 303–363 K were equivalent to the introduced glucose. The reaction below 323 K favored the production of gluconic acid, which corresponded with previous reports about the oxidation of glucose over noble metal catalysts [42].
The yield of gluconic acid decreased with increasing the yields of lactic acid and others over 323 K, however, the yield of others decreased over 353 K. The yields of lactic acid, gluconic acid, and others were 44 C-%, 46 C-%, and about 3 C-%, respectively, at 363 K. Consequently, the reaction over 353 K resulted in the highly atomic-effective conversion of D-glucose into lactic acid and gluconic acid, whose yields were about 44 and 46 C-%, respectively.
Catalytic conversion of several sugars
Table4.3 shows the product yields for the reactions of monosaccharides and gluconic acid using Pt/C catalysts in the NaOH aqueous solution at 353 K for 2 h.
The reactions of fructose and mannose resulted in about 48 C-% yield of lactic acid as high as that in the reaction of glucose, but lower yield of C6aldonic acids and higher yield of others than those of glucose. In contrast, the reaction of galactose resulted in 43 C-% yield of C6aldonic acid as high as that of glucose, but lower yield of lactic acid and higher yields of glycolic acid (5 C-%) and others than those of glucose. The reaction of xylose, i.e., an aldopentose, resulted in about 20 and 27 C-% yields of C5aldonic acid and lactic acid, respectively. The both yields were lower and the yield of others was higher than those of glucose. Although it is considered that glycolic acid is produced as a pair product of lactic acid in the xylose alkaline degradation, the yield of glycolic acid was 6 C-% which was much Table 4.3 The reactions of various monosaccharides and gluconic acid using Pt/C with a NaOH aqueous solutiona
Reactants Conversion Yield /C-%
/% C5-C6acids LA GlyA AceA For A Others
Xylose 100 20 27 6.2 1.8 1.5 43
Glucose 100 45 43 2.9 2.9 0.8 6
Fructose 100 29 48 3.7 2.2 1.1 16
Mannose 100 30 48 3.3 1.7 1.2 15
Galactose 100 43 16 5.1 1.9 1.9 32
Reprinted from Ref. [40], Copyright 2008, with permission from Elsevier LA(lactic acid),GlyA(glycolic acid),AceA(acetic acid),ForA(formic acid)
aMonosaccharide (0.25 mmol) were added to 5.0 mL of NaOH (1.0 mol L-1) aqueous solution with 50 mg of catalyst in a batch reactor bubbled by air flow (20 mLmin-1). Reaction time and temperature were 2 h and 353 K, respectively