MICROWAVES FOR SOL-GEL SYNTHESIS OF BORON CARBIDE (B4C)

Manuel Rodríguez, Ubaldo Ortiz, Juan Aguilar, Zarel ValdezFacultad de Ingeniería Mecánica y Eléctrica, Universidad Autónoma de Nuevo LeónApartado Postal 076F, Cd. Universitaria, San Nicolás de los Garza, NL. 66450 México

ABSTRACTIt is proposed in this research the use of microwaves as an energy supply for producing boron

carbide through a sol-gel process and a thermal treatment. Carbon suppliers were carbon black,calcined saccharose and saccharose. Boron suppliers were boron methoxide, boron ethoxide, andboron tri-isopropoxide and bi-glycerol borate. A comparison with conventional heating showed thatin the microwave case reactions were sped up, and that morphologies are more defined andhomogenous in the synthesis of precursors of boron carbide. It was verified that applying themicrowave sol-gel technique produced nanometric particles of boron carbide either as nanofibers oramorphous, while, greater amounts of nanotubes and less nanofibers are produced when theprecursors were prepared by conventional means.

INTRODUCTIONThe term "Advanced Ceramics" designates ceramics with especial characteristics that made

them useful in different areas than traditional ceramics. Ceramics could be divided in two greatgroups: functional and structural ceramics. Ceramics that perform electronic, electromechanical,magnetic and optical functions belong to the first group. The second group emphasizes theapplications where better mechanical properties at high temperatures are required.

Boron carbide (B4C) exhibit high melting point (2447°C), as a semiconductor type-p show avery high Seebeck coefficient, and it changes to type-n depending of boron content. Other importantproperty is its ability for absorbing neutrons and its thermal stability [1].

Boron carbide is obtained commonly from a reaction at high temperature of boric oxide andcoal coke poured into an electrical arc furnace [2, 3, and 4]. Other technique consisted of loading theelectrical arc furnace with mixtures of boron and metallic magnesium inside the center of a graphitebar, which acts as an anode, producing a mixture of nanofibers and icosahedral carbide crystal (B4C).Other techniques for obtaining this product involved radio frequency plasma [5], vaporelectrochemical deposition and deposition with microwave plasma [6, 7]. Materials produced bythese methods are of high purity and very good for thin film applications [5, 8].

The Sol-Gel process is a technique employed for elaborating ceramics of great quality; thepreparation of a material by this method is carried out by mixing the reactives in liquid phase forhaving a homogeneous dissolution, which leads to polymerization and gel formation [9, 3]. Theresulting gel is transformed latter into the desired compound by means of a controlled heat treatmentduring which, the elimination of volatile components takes place, while density of the material isincreased. Some variants are reported also in literature [10, 11].

It is proposed in this research the use of microwaves as a method for supplying energy to asystem for obtaining a boron carbide precursor and then, boron carbide itself. Conventionalprocessing was also considered for comparison purposes.

EXPERIMENTALThe steps for obtaining boron carbide consisted in producing the precursors from carbon and

boron suppliers. Carbon suppliers were carbon black, calcined saccharose and saccharose; while

boron suppliers were boron methoxide, boron ethoxide, boron tri-isopropoxide and bi-glycerolborate. Necessary energy was provided either conventionally or by microwaves. In this work, themeaning of microwave processing is that all of the stages, but drying, were conducted withmicrowaves. Drying was conducted conventionally in all cases.

Boron alcohoxide preparationFor obtaining boron alcohoxide is advisable to start from boron oxide, which can be obtained

directly from boron acid through a dehydration process [12].

Boron oxide preparationPowdered boric acid (99.95% purity) was employed for obtaining boron oxide in a closed

stainless steel container purged with helium. The system was heated slowly up to 115°C in vacuumat 20 mmHg, under constant stirring for 45 minutes. Then, temperature was increased gradually inone hour to 140°C, and maintained there for 15 minutes, then heated to 150°C and kept for 15minutes more. After that, the system was cooled dawn to room temperature. The system was heatedagain to 150°C, then to 210°C at 0.2°C/min. Once that temperature was stable, it was heated up to250°C at 1.3°C/min. The obtained product was white, fibrous, sponge like and bridle.

Blending and compacting of the powdersBoron oxide was blended in a 1 liter, 500 gr, 12.5 mm alumina balls blender. Also a mixture

of boric oxide (50 gr.), and magnesium chloride (50 gr.), was prepared in three hours. The processwas conducted at 150°C in nitrogen; blender was rotating at 75 rpm. The obtained powders werecompacted in tablet shape 1cm diameter and 1cm thickness (cylindrical) and were maintained at100°C until the moment of running the test.

Reactor system and procedure for synthesis of alcohoxidesThe system consists in a glass column that is connected in the lower part to a three-mouth

flask. The upper part is connected to water-cool a condenser. Condenser outlet was connected to achamber with frozen CO2. A thermometer was inserted in one mouth of the flask and the other isused for inlet of nitrogen, or helium in some cases. In the microwave process, heating was achievedby fixing the bottom part of the flask in a multimode cavity (Fig.1).

The tablets were placed inside the glass column. The system was kept at positive pressure. Apump was connected to the extraction tube and at alternative times hot methanol was added to thecolumn that contain either the boric oxide or the mixture with magnesium chloride, wetting thesample and taken into the flask the reaction products. The precipitated compounds were maintainedunder stirring. Microwave processing was about the same, but microwave were applied over thebottom part of the flask and stirring was mechanical.

The same procedure was conducted with methanol, ethanol, isopropylene and glycerol. In theconventional processing case the temperature was kept at 60°C for methanol, ethanol andisopropylene, and 100°C for glycerol. This procedure was conducted for either 42 hours or 30minutes after the load was consumed, which occurs first.

In the microwave processing case, it was necessary to determine which was the optimalexposition time of the alcohol to microwaves under an on/off system for controlling the temperatureas Table 1 shows for each kind of alcohol. In this case the procedure lasted for 4 hours or 1 minuteafter the load was consumed, whichever occurs first.

Fig. 1. Scheme of the experimental arrangement for the synthesis of the boron suppliers.Conventional heating (left) and microwave heating. 1) Solution; 2) Pipe; 3) Alcohol discharge; 4)Condenser; 5) Condenser; 6) Low temperature condenser; 7) Gas inlet; 8) Thermometer; 9) Heater;10) Magnetic stirrer; 11) Mechanical stirrer; 12) Multimode microwave cavity; 13) Pump.

Table 1. Microwave exposition times and temperatures registered for the different employeddissolvent.

Compound Temperature Microwave heating time (sec)On Cycle

Resting time (sec)Off Cycle

Methanol 60°C 6 420Ethanol 60°C 7 600

Isopropylene 60°C 5 360Glycerol 100°C 8 600

Proposed reaction for boric oxide and alcohol is (R is an alquilic group):

B2O3 + 6 ROH 2B(RO)3 + 3H2O (1)

Reaction (1) is running always that an excess of alcohol is maintained. Hydrogen loss during thereaction makes the alcohol amount to decrease, inducing the production of more alcohol. Water wasremoved by distilling [12, 13]. The compound produced is boron alcohoxide in a methanol solution[13, 14, and 15].Reactions proposed for boric oxide mixed with magnesium chloride are as follows [13, 14]:

B(RO)3 + ROH H+[B(OR)4]¯ (2)

MgCl2 + 2ROH Mg(RO)2 + 2HCl (3)

Mg(RO)2 + 2 H+[B(OR)4]¯ Mg [B(OR)4]2 + 2 ROH (4)

Production of solutions and colloids for carbon suppliersThree agents were used as carbon providers, the first one was carbon black with 99.9% of

carbon. The second and third ones were respectively saccharose (reactive degree), and calcinedsaccharose (180°C in nitrogen atmosphere). 50 gr of saccharose were dissolved in 100 ml of alcoholthat corresponded to the test that was taken place, under constant stirring at 60°C in a closedcontainer with a re-circulator and a condenser, while alcohol level was kept constant. Procedure ofcarbon black and calcined saccharose was almost the same:

1. 15 gr of carbon black or 25 gr of calcined saccharose, were mixed with an humectant agent,ethylene glycol in this case, for making them wettable by an organic solvent.

2. 10 ml of HCl was added to either wet carbon or calcined saccharose, while stirred for 3 hoursas it was heated to 60°C with 100 ml. of the respective alcohol.

3. The system was kept for 1 hr at constant temperature. Alcohol level was maintained byalcohol additions.

Carbon black and calcined saccharose formed a colloidal suspension, and for saccharose it wasdissolution. Each one of these compounds was dissolved in 100 ml of methanol, ethanol,isopropylene or glycerol, and was prepared just on time for mixing with boron or boron-magnesium.

Precursor preparationPrecursors were synthesized by two techniques. First one is called sol-gel-conventional

(SGC), and the second is based on the first one, but heating of the sol is conducted with microwaves,therefore was named sol-gel-microwaves (SGM).

Experimental design and designation of the testsExperimental design and designations for the prepared precursors, according to the employed

compounds, including a second run with Magnesium Chloride are shown in Table 2.

Table 2. Experimental design for precursors prepared by sol-gel-conventional and sol-gelmicrowaves, with and without magnesium chloride.

Boron Methoxide Boron Ethoxide Boron Tri-isopropoxide Di- glycerol borateSGC SGM SGC SGM SGC SGM SGC SGM

Carbon Black MC MCW EC ECW IC ICW GC GCWCalcined

SaccharoseMF MFW EF EFW IF IFW GF GFW

Saccharose MS MSW ES ESW IS ISW GS GSWSGC (Sol-Gel-Conventional), SGM (Sol-Gel-Microwave)“M” represents Boron Methoxide; “E”, Boron Ethoxide; “I”, Boron Iso-propoxide; “G”, Di-glycerolborate; “C”, Carbon Black; “F”, Calcined Saccharose and “S”, Saccharose. Microwave processing isdesignate with a W after the code.A second run was conducted with Magnesium Chloride and they are identified with “-M” at the endof the code (i.e. MC and MC-M).

Experimental arrangement and synthesis procedure of the precursorsExperimental arrangement for the sol-gel-conventional processing is shown in Fig. 2, where a

ball flask and the condensation pipe are observed on the middle mouth. The arrangement for the sol-gel-microwave processing is also shown in the same figure, but in this case stirring was mechanicallyperformed.

Fig. 2. Experimental arrangement of Sol-Gel-Conventional (left) and Sol-Gel-Microwave processes.

Mixing of the reagents was as follows:

1. Boron provider was added first and was heated as shown in Tables 3 and 4 for conventionalprocessing and as shown in Tables 5 and 6 for microwave processing.

2. In the meantime, the carbon supplier was maintained stirred and at constant temperature;then, as described in the above point, boron supplier was added in stages during 5 minutes.After the solution was formed, the gelation and then dehydration continued.

Table 3 Temperature and time of each sol-gel-conventional processing test.Precursor MC MF MS EC EF ES IC IF IS GC GF GSTime (hrs.) 12 11 9 15 15 10 15 15 13 15 15 12Temp.(°C) 60 60 60 65 65 65 70 70 70 100 100 100

Table 4 Temperature and time of each sol-gel-conventional processing test(with Mg).Precursorwith Mg.

MC-M

MF-M

MS-M

EC-M

EF-M

ES-M

IC-M

IF-M

IS-M

GC-M

GF-M

GS-M

Time (hrs.) 18 15 10 18 15 12 18 15 11 18 18 15Temp.(°C) 60 60 60 65 65 65 70 70 70 100 100 100

Table 5 Temperature and time of each sol-gel-microwave processing test.

Precursor MCW MFW MSW ECW EFW ESW ICW IFW ISW GCW GFW GSW

ΣTime(hrs.) 2 2 1.5 3 3 2 3.5 3.5 3.25 3 3 2.5

Temp.(°C) 60 60 60 65 65 65 70 70 70 100 100 100

Table 6 Temperature and time of each sol-gel-microwave processing test(with Mg).

Precursorwith Mg.

MCW-M

MFW-M

MSW-M

ECW-M

EFW-M

ESW-M

ICW-M

IFW-M

ISW-M

GCW-M

GFW-M

GSW-M

ΣTime(hrs.) 2 2 2 2.5 2.5 1.75 3.25 3 2.75 2.75 2.5 2

Temp.(°C) 60 60 60 65 65 65 70 70 70 100 100 100

Gels obtained by both processes were dried in an oven at 150°C in nitrogen for 24 hrs. Followed of aheat treatment at 800°C in nitrogen for 4hrs. for removing organic material.Although it is not the purpose of this work to perform kinetic calculations, heating time and effectiveheating time in the microwave processing case was recorded.

Microwave sensitivity of precursors and heat treatmentOne important issue for microwave processing is that precursors must be microwave

absorbent, otherwise they would not warm up. In cases where they were not absorbent enough theywere mixed with up to 10% of amorphous coal. Only those precursors that absorbed microwaveswere chosen for the heat treatment. This stage consisted in having the precursor encapsulated inquartz at high vacuum (10-4 torr) an placed into a multimode cavity for being exposed to 1400 Wattsat 2.45 GHz. Power was applied for 800 minutes in on/off (15 min/10 min) cycles.

RESULTS AND DISCUSSIONCharacterization

Precursors and obtained materials were characterized by X-ray diffraction, Scanning electronmicroscopy (SEM), Transmission Electron Microscopy (TEM), Thermogravimetrical analysis, andInfrared spectroscopy, but this paper is dedicated to SEM and TEM observations of the samplesobtained after thermal treatment.

Chemical AnalysisChemical analysis of the calcined (800°C for 4 hrs.) precursors searching for boron retention

for both cases, conventional and microwave heating, shows that the highest content was for thoseprepared in isopropanol and magnesium in the synthesis (IC-M, IF-M, IS-M, ICW-M, IFW-M, ISW-M).

Scanning electron microscopy (SEM)Obtained images showed the presence of micro and nanoparticles, above or around the

surface of flakes formed in the matrix. The following images correspond to the samples obtainedfrom precursors IC-M (Iso-propoxide with magnesium chloride) and ICW-M (Iso-propoxide withmicrowaves) which were precursors with high boron content and also were able to absorbmicrowaves. General morphology that is presented by these precursors is flakes and zones with greatamounts of micro and nanofibers, and nanoparticles (Fig. 3). This late morphology was morenotorious in the precursor IC-M (Fig. 4), however, presence of spheroid particles over fibers isgreater in precursor ICW-M. Morphology exhibited by precursor ICW-M is comparable to severalones reported in literature, specially with those where boron carbide is obtained by electric arcdischarges, wish produces nanofibers and nanotubes of boron carbide (B4C) [16-19].

Fig. 3. SEM images of a sample obtained from precursor ICW-M, fibers and particles of micro andnanometrical size can be seen (left), and another zone where particles are evident (right). Scalemarks: 1 μm.

Fig. 4. SEM images of a sample obtained from precursor IC-M showing the same microfibers (left).Sample on the right corresponds to precursor IFW-M, similarity with sample ICW-M (Fig. 3 right) isnoticeable. Scale marks: 1 μm and 3 μm.

Transmission electron microscopy (TEM)Analysis showed by TEM coincides with the morphology reported in literature, which are

nanotubes and nanofibers of carbon and boron carbide, and some spheroid particles nested along thenanofibers. This morphology was more frequent in the precursors prepared by microwaves (ICW-M,IFW-M, Fig. 5) than the conventional precursors, where although there were more nanotubes therewere less spheroid nanoparticles and nanofibers. Figure 5 also allows appreciating similarmorphologies to the previous precursors. TEM and SEM morphologies are coincident. According toliterature these morphology correspond to nanotubes and nanofibers of boron carbide (BxCy) and thespheroid particles to Magnesium diborure (MgB2), or boron carbide (B4C) [16,20].

Figure 6 shows several nano elements in the sample IC-M. Presence of nanofibers, withparticles over the surface are rather isolated in contrast with sample ICW-M (Fig. 5), where presenceof nanofibers and spheroids is greater, but they are not on the surface of the nanofiber, instead theyare crossed by a nanofiber or a nanotube.

Fig. 5. TEM images of a sample obtained from precursor ICW-M (left) and IFW-M showingnanotubes and nanofibers. Scale marks: 200 nm.

Fig. 6 TEM images of a sample obtained from precursor IC-M showing fibers and a nanotube (right).Scale marks: 200 nm and 500 nm.

The results showed a greater efficiency in the samples heated with microwaves based on theanalysis performed on the mixtures with different alcohol for alcohoxide generation, which allow topropose that microwaves could activate generation of ion alcoholates, which shows a 10:1 relationwith normal heating. It was also found that microwave heating speeds up the reaction, and thatmorphologies are more defined and homogenous in the synthesis of precursors of boron carbidecompared to those obtained conventionally. More over, the gels obtained by means of microwave aremore stable than those obtained conventionally.Regarding the kind of obtained product, it was found that more nanoparticles of boron carbide wereobtained at temperatures of 1300°C in the microwave case, while it was verified that applying themicrowave sol-gel technique, obtained particles of boron carbide are nanometric, either as nanofibersor amorphous. While, greater amounts of carbon nanotubes and less nanofibers are produced whenthe precursors were prepared by conventional means.The amount of energy that was actually input to the sample was not quantified. Therefore it is notpossible to discuss over any "microwave effect" that are attributed often to microwave processing.

CONCLUSIONAlthough generation of alcohoxides from alcohols seems to be faster, that means that the heating ratewas greater, because supplied energy was also greater, and do not necessarily means that themicrowave activated generation the alcoholate ions. Reaction times for the Sol-Gel-Microwavesprocessing is up to six times faster than Sol-Gel-Conventional, this situation also makes that lessboron alcohoxide was evaporated and therefore concentration of this compound was up to threetimes higher. Added magnesium promotes boron retention in the synthesis of the precursors in bothkinds of processing. Precursors prepared by sol-gel-conventional presented less reaction withmagnesium compared with microwave processing. However, formed compounds are the same inboth cases. Employing the sol-gel-microwaves technique produce boron carbide in nanometricparticles, nanofibers and in an amorphous presentation. In the conventional case there are morecarbon nanotubes and less nanofibers of boron carbide. Magnesium presence promotes the synthesisof nanofibers of boron carbide and carbon nanotubes. Magnesium also promotes reduction of boroncompounds.

REFERENCES1.- Kirk-Othmer, Encyclopedia of Chemical Technology, Third Edition, Ed. John Wiley & Sons,Vol. 4, pp.126-127, (1978).2.- K. J. Rao, K. B. R. Varma, A. R. Raju, Light elements ceramics, Advanced Ceramics, Ed. E. C.Subbarao, Indian Academy of Sciences, pp. 85-89, (1988).3.- A.W. Weimer, R.P. Roach, C.N. Haney, W.G. Moore, W. Rafaniello, Rapid carbothermalreduction of boron oxide in a graphite transport reactor, AIChE Journal, 37, 5, pp. 759-768, (1991).4.- B.Q. Wei, R. Vajtai, Y.J. Jung, F. Banhart, G. Ramanath, and P.M. Ajayan, "Massive IcosahedralBoron Carbide Crystals", J. Phys. Chem. B , 106, pp. 5807-5809, (2002).5.- Ö.D. Eroğlu, N.A. Sezgi, H.Ö. Özbelge, H.H. Durmazuçar, Synthesis and Characterization ofBoron Carbide Films by Plasma-Enhanced Chemical Vapor Deposition, Chemical EngineeringCommunications, 190, 3, pp. 360-372, (2003).6.- Handbook of Chemical Vapor Deposition, Principles, Technology and Applications 2ed, Hugh O.Pierson, Noyes Publications, William Andrew Publishing, LLC, Norwich, N.Y., U.S.A. pp. 210,ISBN: 0-8155-1432-8, (1999).7.- K.W. Lee, S.J. Harris, Boron Carbide Films Grown from Microwave Plasma Chemical VaporDeposition, Diamond and Related Materials, 7, 10, pp. 1539-1543, (1998).8.- J.H. Lee, C.W. Won, S.M. Joo, D.Y. Maeng, “Preparation of B4C powder from B2O3 oxide bySHS process”, Journal of Materials Science Letters, 19, pp. 951-954, (2000).9.- C.J. Brinker, B.C. Bunker, D.R. Tallant, K.J. Ward, Chemical Reactivity and the Structure ofGels, Journal de Chimie Physique, 11/12, pg. 83, (1986).10.- I. Hasegawa, Y. Fujii, T. Takayama, K. Yamada, Phenolic resin-boron oxide hybrids asprecursors for boron carbide, Journal of Materials Science Letters, 18, 1629-1631, (1999).11.- A. Sinha, T. Mahata, B.P. Sharma, Carbothermal Route for Preparation of Boron CarbidePowder from Boric Acid-Citric Acid Gel Precursor, Journal of Nuclear Material; 301, pp. 165-169,(2002).12.- Ullmann´s Encyclopedia of Industrial Chemistry, Fifth edition, Vol. A4, pp. 309-330.13.- N.Y. Turova, E.P. Turevskaya, V.G. Kessler, M.I. Yanovskaya, The Chemistry of MetalAlkoxides, Ed. Kluwer Academic Publishers; 1st edition, ISBN: 0792375211, (2002).14.- .D.C. Bradley, R.C. Mehrotra, D.P. Gaur, Metal Alkoxides, Ed. Academic Press; 1st edition,ISBN:0-12-124250-1, (1978).15.- S.H. Webster, L.M. Dennis, Preparation and Purification of Methyl and Ethyl Borates, J. Amer.Chem. Soc., 55, pp. 3233-3235, (1933).

16.-M.J.Pender, L.G.Sneddon, An Efficient Template Synthesis of Aligned Boron CarbideNanofibers Using a Single-Source Molecular Precursor, Chem Mater., 12, pp. 280-283, (2000).17.-M.Büttner, O.Wondratschek, C.Ronning, U.Vetter, H.Ch.Hofsä, Ion Beam Synthesis ofamorphous boron carbide thin films, Surf. Coat. Technol., S.382, pp. 158-159, (2002).18.- H.Disbudak, I.Uslu, A.Y.Bilgesu, G.Gündüz, Boron Carbide Deposition on Urania and Urania-Gadolinia fuels, Materials for Nuclear Systems, V.135, No.3, pp.286-294, (2001).19.- B. Wei, R. Vajtai, Y.J. Jung, F. Banhart, G. Ramanath, P.M. Ajayan, Massive Icosahedral BoronCarbide Crystals, Journal of Physical Chemistry B; 106,23 (2002) pp. 5807-9, (2002).20.- S.N.Dilek, H.Ö.Özbelge, N.A.Sezgi, T.Doğu, Kinetic Studies for Boron Carbide Formation in aDual Imping-Jet Reactor, Ind.Eng.Chem.Res., 40, pp.751-755, (2001).