Preprint of a paper accepted by Chem.Engine.Sci, © Elsevier 2001

Reactive milling with the Simoloyer®: environmentally benign quantitative reactions without solvents and wastes

G. Kaupp*, J. Schmeyers, M. R. Naimi-Jamal

University of Oldenburg, FB 9 - Organic Chemistry 1, P.O. Box 2503; D-26111 Oldenburg, Germany;

H. Zoz, H. Ren

Zoz GmbH, Maltoz-Strasse, D-57482 Wenden, Germany; ren@zoz.de

Abstract: The pilot set-up of a horizontal high energy ball mill (Simoloyer®) that has been successful in reactive milling, mechanical alloying, and formation of ductile metal-flakes is used for semi-continuous processing of organic solid-solid reactions which proceed rapidly and without wastes in the absence of solvents. 200 g batches were run in the exploratory investigation. Stoichiometric 1:1-mixtures of alpha-(D)-glucose and urea formed the 1:1-complex with quantitative yield after 5 min at a rotor frequency of 900 cm-1. Similarly the condensation reaction of solid p-hydroxybenzaldehyde with solid p-aminobenzoic acid was complete after 15 min yielding the pure imine derivative in quantitative yield without waste producing work-up procedures. Further scale-up appears possible also with other solid-solid reactions.

Keywords: Chemical Reactors, Condensation, Environment, High Energy Milling, Reactive Milling, Powder Technology, Scale-up

*Corresponding author. Tel.: + 49 441 798 3842; fax: + 49 441 798 3409
E-mail address: kaupp@kaupp.chemie.uni-oldenburg.de (G. Kaupp)

1. Introduction

Chemical production continues to depend on environmentally benign techniques. The important goal of avoiding wastes is only obtained in 100% yield reactions that do not require workup for removal of catalysts, supports, solvents or other auxiliaries. Thus, the up-scaling of quantitative solid-state reactions in ball-mills (Kaupp, Schmeyers, and Boy, 2001) is of technical importance. We report here on the first semi-technical milling for the quantitative and waste-free production of an organic 1:1-complex 3 and a condensation product 6 in batches of 200 g that can be easily up-scaled further. The starting materials are commercially available and used as obtained.

2. Processing route

Since almost half a century scientists are using the mechanical alloying technique (MA) which can be used to produce powder-materials with new and insofar interesting properties (Benjamin and Volin, 1974; U. Koester, 1997). The literature describes MA as repeated deformation, fracture and welding of powder particles by highly energetic ball collisions. This definition may be amplified in that MA stands out by the transfer of high level energy into powder and so Mechanical Alloying often leads to material transformation of the crystalline structure by solid state reactions. Atomic dislocations, a high defect structure of the lattice, the immense magnification of the boundary surface and a high diffusion rate leads to low activation energies for those reactions.
If the same technique is applied for particle size reduction and/or particle deformation of single-systems e.g. to receive a special particle geometry, this route is to be described as High Energy (ball) Milling (HEM) and is in particular very suitable for rapid particle size reduction of enamels, glass fluxes and glaze frits. On the contrary HEM has already been found to be very effective for the production of ductile metal-flakes for paint-pigments, electrical conductive pastes, anti-corrosives and many others (Zoz, et al, 1999a,b).

The definition of Reactive Milling (RM) is suitable if during milling a chemical reaction is wanted and observed. The advantage here can be an ultra-fine dispersion of particles in a matrix: e.g. 7Ag + SnO2 where the starting powder is 2Ag2O + Ag3Sn (Zoz, Ren, and Spaeth, 1999a).
A major goal for high kinetic processing (HKP) is the possibility of producing large quantities in a short time at low cost. Therefore a project has been founded to explore the potential of a continuous route of HKP where the material is injected in a carrier gas flow and separated again after processing. The scheme of the continuous pilot plant is shown in Figure 1. A further use of the same set-up is in organic solid-state reactions where the necessity for repeated contacts between the micronized reacting crystals prevails for obtaining quantitative reactions without producing wastes. No significant activation energies have to be introduced in the exoenergetic reactions of molecular crystals. Unlike metals or oxides (where metal- or covalent-bonds have to be broken) only van der Waals interactions or hydrogen bridging bonds are to be broken.

3. Experimental

200 g quantities of stoichiometric 1:1-mixtures of the loosely premixed crystalline reaction partners (all starting materials were of 99% purity and purchased from Merck KGaA, Darmstadt) were fed to a stainless steel 2 l Simoloyer® horizontal rotary-ball-mill equipped with a hard metal rotor of standard geometry (Zoz, Ren, and Spaeth, 1999a), 2 kg of steel balls (100Cr6) with 5 mm diameter, and water cooling. The temperature was 15°C at the wall with a maximum of 19°C in the center of the Simoloyer®. The rotor was run at 900 cm-1 (the power was 610 W) for 5 min in the case of 3 and 15 min in the case of 6 for quantitative reaction (m.p., IR spectra, chemical analyses and DSC experiments indicated 100% conversion) and the products were milled out for 10 min at the same rotor frequency. The yields of pure products 3 and 6 were quantitative from the second batch milled. For quantitative recovery of the material in one-batch runs and in the last batch, the Simoloyer® ball-mill can be connected to the air cycle for deposition through a cyclone (see Fig. 1). A Gallenkamp melting point apparatus was used. All FT-IR spectra were taken in KBr at a Perkin Elmer 1720 instrument.

4. Results and Discussion

We chose formation of the 1:1-complex of alpha-(D)-glucose (1) (m.p. 153 - 156°C) with urea (2) (m.p. 132.5 - 134.5°C) and a condensation reaction of p- hydroxybenzaldehyde (4) (m.p. 114 - 117°C) with p-aminobenzoic acid (5) (m.p. 186 - 189°C) for this exploratory investigation, in order to demonstrate the application potential in very different fields of interest.
Short milling of compounds 1 and 2 quantitatively produced complex 3 (m.p. 115 -120°C). This is easily followed by IR spectroscopy (most pronounced changes: loss of the absorptions of glucose at 3306, 996, 839, 777, 623 and of urea at 1683 and

IDunitIDunit
01Simoloyer CM01-s115transparent pipe module GR-DN25x75
02side-channel-turbine SKV180a-DN4016transparent pipe module GR-DN40x200
03Vacuum pump DUO 1017pipe switch RW40-16-A
04gas-bottle 10 liter18KF-space-switch RW40-B
05electronic cabinet19pipe bends RBA-DN40 & DN25
06laboratory cyclone ZK70-L (a)20adapter KF-A
07laboratory cyclone ZK70-L (b)21valve adapter DN*G*DN*
08rotary vane feeder ZS40m (a)22KF-glass-container DN40-G1-500 cc
09rotary vane feeder ZS25m23KF-valve-container DN40-G1-2l
10rotary vane feeder ZS40m (b)24vacuum screen unit VSK28
11butterfly valve KV-DN40 (a*)25KF-calming pipes
12butterfly valve KV-DN40 (b*)26KF-tubes, straight
13butterfly valve KV-DN40 (c*)27KF-junction-tubes
14transparent pipe module GR-DN40x10028pressure-gauge DMD16

Figure1: Scheme of the pilot-set-up of a High Energy Ball Mill (Simoloyer®) with air/inert carrier gas cycle and separation/classification system.

1625 cm-1; new bands of the complex at 3457, 3357, 1666, 1650, 1503, 1035, 1017, 865 cm-1). The alpha-configuration of the glucose part in 3 was shown by an initial rotation angle of [alpha]D? > 82° (1% in water) when the usual mutarotation occurred. The same compound had been prepared in low yield from an aqueous solution by inoculation with seed that was obtained after six months of crystallization (Hatt and Triffett, 1963) or from melts (Quehl, 1938) and characterized by an X-ray structural analysis (Snyder and Rosenstein, 1971). Interestingly, the chemical transformation proceeded rapidly without intervening liquid phases, despite the numerous hydrogen bonds in the crystal lattices of both starting materials.Quantitative solid-state synthesis of compound 6 (m.p. 238 - 240°C, decomposition) from 4 and 5 was initially tested at the gram scale in a 10 ml Retsch ball-mill MM 2000. It can be easily followed by IR spectroscopy. The carbonyl vibrations of 4 at 1667 and 5 at 1668 cm-1 disappear and the frequency of 6 appears at 1687 cm-1. 6 forms a crystalline hydrate. Thus, the water of reaction from the condensation reaction does not disturb the procedure by a liquid phase both in the small or large scale runs, as it is incorporated in the product lattice. The water of reaction can be removed from 6 . H2O in a vacuum at 80°C, if required. Importantly, the condensation reaction of 4 and 5 proceeds smoothly and efficiently at near ambient conditions without intervening liquid phase, and further up-scaling in larger mills appears possible. This favorable waste-free technical procedure is remarkable, as the previous synthesis of 6 required 12 h boiling in ethanol and no yield was reported (Cevasco and Thea, 1999).
The semi-technical stoichiometric solid-state reactions run to completion. This fact indicates a deltaG-value of < -6 kcal mol-1 for the overall process. The solid-state reactions proceed rapidly due to the ingenious kinetics involved (Kaupp, 1996). Thus, the contact areas at the different touching crystals experience the well-established phase rebuilding, followed by phase transformation and disintegration with formation of fresh surface, and HKP produces large quantities in a short time.

5. Concluding Remarks

Precise stoichiometry of the reaction partners is crucial for 100% reactions yielding pure products without any necessity for purification. There is no trouble in this respect with (semi continuous) batch runs in ball-milling. Continuous techniques will be more involved, because the stoichiometric ratio has to be maintained throughout. However, the number of batches can be kept low with larger ball-mills that are commercially available at sizes of up to 400 l and should be equipped with an internal air cycle cyclone (Fig. 1) for quantitative recovery at the last batch. The new technique has potential for resource saving environmentally benign up-scaling of numerous further solid-state syntheses.

References

Benjamin, J.S., Volin, T.E. 1974. Metall. Trans. Vol. 5.
Cevasco, G., Thea, S. 1999. Mechanism of Alkaline Hydrolysis of Some HO-p-COOAr Acyl Derivatives. The Journal of Organic Chemistry 64, 5422-5426.
Hatt, H. H., Triffett, A. C. K. 1963. A molecular addition compound of alpha-D-glucopyranose. Journal of the Chemical Society 2079-2080.
Kaupp, G. 1996. Davies` Comprehensive Supramolecular Chemistry Volume 8, pp. 381-423. Elsevier, Oxford.
Kaupp, G., Schmeyers, J., Boy, J. 2001. Waste-free solid-state syntheses with quantitative yield. Chemosphere 43, 55-61.
Koester, U. 1997. Phase separation, crystallization, and grain growth. Micromechanism in the design and stability of nanocrystalline alloys. Mater. Sci. Forum 235-238, 377-388.
Snyder, R. L., Rosenstein, R. D. 1971. The Crystal and Molecular Structure of the 1:1 Hydrogen Bond Complex Between a-D-glucose and Urea. Acta Crystallographica B27, 1969-1975.
Quehl, K. 1938. Molecular compounds of carbamide and saccharoses. U.S. Patent 2116 640; Chemical Abstracts 32, 5001(3).
Zoz, H., Ren, H., Spaeth, N. 1999a. Improved Ag-SnO2 electrical contact material produced by mechanical alloying. Metall 53, 7-8, 423-428.
Zoz, H., Ernst, D., Reichardt, R., Ren, H., Mizutani, T. 1999b.
Okouchi, Simoloyer CM100s: Semi-Continuous Mechanical Alloying on a Production Scale Using Cycle Operation - Part II. Materials and Manufacturing Processes, Vol. 14, 6.