US2684888A - Catalytic reaction of alkali metal hydride and boron trihalide - Google Patents

Description

July 27, 1954 E. H; PRYDE 2,684,888 CATALYTIC REACTION OF' ALKALI METAL HYDRIDE AND BORON TRIHALIDE Filed June 24, 1952 INVENTOR. EVERETT H. PRYDE ATTORNEY Patented July 27, 1954 CATALYTIC REACTION OF ALKALI METAL HYDRIDE AND BORON TRIHALIDE Everett H. Pryde, Kenmore, N. Y., assigner to E. I. du Pont de Nemours and Company, Wilmington, Del., a corporation of Delaware Application June 24, 1952, Serial No. 295,310

(Cl. 22E-d4) 17 Claims.

This invention relates to the borohydrides of alkali metals and more particularly to sodium borohydride.

In the patent application Serial No. 295,266 of Hansley and Pryde, filed of even date herewith, there is disclosed a method for producing sodium borohydride by means of the reaction:

mari-tergenaeml-snan While the reaction shown had not been carried out previously, the yield obtained was, at best, only about 51%. I have discovered that a yield much higher 'than 51% can be obtained if a catalyst be employed during the course of the reaction shown. A primary object of the invention therefore development of an improved process for making alkali borohydrides, useful compounds which may serve as reducing agents.

It has been found that if an alkoxide, such as sodium methoxide, be added to powdered sodium hydride and boron trifluoride is passed over the resulting mixture a yield of up 'to about 88% sodium borohydride may be obtained in contrast to the 51% reported. The temperature for the reaction should be controlled within the range 150-400 C. and preferably held between about 250D and 300 C. In the preferred embodiment of the invention approximately stoichiometric quantities of gaseous boron trifiuoride at atmospheric pressure are led over an agitated mass of comminuted sodium hydride and sodium methoxide, agitation being carried out in a closed reactor in a blanket of inert gas substantially excluding moisture and air. For best results a heel of material from a previous run is mixed with the sodium hydride currently utilized to improve yields and help prevent the caking of the solids. The crude borohydride mixture thus prepared can be utilized directly as a reducing agent or purified by extraction with isopropyl amine and subsequent solution in absolute ethyl alcohol.

Of especial importance to the instant process is agitation of the reactants. This expedient reduces undesirable caking of the solids and allows the reaction to proceed further towards completion than it otherwise would. While any conventional agitation means may be employed, a stainless steel ball mill reactor has been found particularly advantageous. A mill convenient for the purposes of this invention is substantially that described in Industrial and Engineering Chemistry, 43, 1759-1766 (1951), particularly at page 1764, and is shown in the gure, which represents a View, partly in elevation and partly in section, of the apparatus utilized.

In the figure a stainless steel reactor l0, containing steel balls l l, is shown supported by bolts I2 and shaft I3, the shaft being rotatably connected to air motor I4 through appropriate gearing. Careful machining at anges i6 permits a gas tight -t to be made between reactor I6 and neck Il. Neck I l, like the reactor formed of stainless steel, is supported at roller bearing i8 by stand I9 and terminates in spherical brass joint 2t. Joint 2li provides rotatable contact between glass adaptor 2l and neck I l. Tension springs 22 and 23 hold the adaptor to stand 9 with any desired resilient force. An outlet arm 25 forming a gaseous exit for the entire system is integral with adaptor 2l. The free end of adaptor 2i is closed by stopper 26 through which is inserted thermocouple well 2l and inlet tube 23. Air motor l is held at pivot 29 to stand 36 which in turn is rigidly fastened to base-plate 3! at a predetermined distance from stand i9. Furnace 32, supported on base-plate 3l, is constructed around the reactor lil, holes being left to admit passage therethrough of shaft i3 and neck il. Furnace 32 may be made of any convenient heat resistant material. If constructed of discrete units such as magnesia fire-brick, the furnace may however be easily assembled or dismantled. Heat is supplied to reactor l@ by any convenient source such as a small flame (not shown) playing into the furnace.

The operation of the apparatus is believed evident from the description. Reactor lil was charged with balls It and solid hydride and 1litrogen led in through inlet tube 28 to blanket the entire system and ensure exclusion of air and moisture. A rotameter (not shown) was inserted after outlet arm 25 to measure gas flow. After the air had been replaced by nitrogen, milling was commenced, the reactor being rotated by the motor 4, and BFS gas fed through inlet tube 28. When the absorption of BF3 was considered complete the glass adaptor 2E was disconnected at joint 20, the furnace 32 disassembled and the reaction product removed from the reactor. To assist in the product removal the mill may be dropped downward by means of pivot 29.

Details of the instant process may be more easily understood from the specific examples following. Example 1 is included merely for ease in contrasting procedure which does not use an alkoxide catalyst with procedure utilizing such a material.

Example 1 The stainless steel reactor was charged with 3/4 steel balls, 234 grams of product from a previous run, the product analyzing 14.0% sodium hydride and 11.9% sodium borohydride, and 110 grams (4.58 gram moles) of free sodium hydride. The reactor was then mounted in position, evacuated and iilled with dry nitrogen. at atmospheric pressure. The ball mill was started and heating begun. When the temperature had reached 180 C. boron triluoride gas was passed in at atmospheric pressure and a rate of about 500 cc. per minute. Absorption or" boron trifluoride gas gradually decreased until after about an hour of heating there was a sudden rise in temperature, the absorption then remaining steady at about o cc. per minute for another hour. During this period the temperature varied from 230 C. to 300 C. After two hours the rate of absorption again gradually decreased and finally practically ceased. The iiow of trifluoride was then. stopped and cooling begun. A total of 0.21 mole BFS was absorbed for each mole or Nai present the theoretical or stoichiometrical dgure being 0.25. The product was analyzed and found to contain 8% sodium hydride and 10% sodium borohydride, the yield of the latter being 51%. It may be noted that, despite the precautions taken, there was some packing of the nal product Example 2 (a) A run was made Ito determine the catalytic efliciency of sodium methoxide (NaOCHa) on the reaction between sodium hydride and boron trifluoride. The procedure followed was generally that of the previous example except for the addition of the catalyst. lnto the reactor were charged 5.78 moles (138.9 grams) of commercial grade sodium hydride, 102 grams of heel from a previous reaction in which no methoxide had been utilized and 12.8 grams of commercial sodium methoxide. rEhe heel contained 0.34 mole of Nal-I and 0.27 mole of NaBI-li. The methoxide added was equivalent to 0.237 gram mole or 0.039 mole for each mole o1" sodium hydride present. After the mixed solids had been milled in a nitrogen atmosphere for about an hour, heat was applied to the furnace and the temperature in the reactor raised to 240. The flow of boron triiiuoride at atmospheric pressure was then begun and continued for about 2 hours, the temperature varying during the reaction period between 238-270 C. The product obtained was found to be somewhat packed but was nevertheless utilized as a, heel in a further reaction.

(b) To `the product of the previous reaction y were added increment of 7.92 moles or 190.0 grams of sodium hydride and 23 grams of sodium methoxide, the methoxide increment representing 0.518 gram mole or 0.065 mole for each additional mole of sodium hydride. The ball mill was started and the furnace heated. When the reactor temperature had reached 225 C. the boron trifluoride stream was again admitted. After about an hour a sudden surge in temperature necessitated stopping the triiiuoride flow. Hydrogen was passed through the system for a short time to cool the reactor. When the temperature had again reached 234 C. boron trifluoride was passed through the apparatus for about 11/2 hours more, a total of about 21/2 hours reaction ltime being required for reacting the second `Nal-I increment.

The product obtained upon cooling the reactor was analyzed, showing a weight percentage of 12.6% sodium hydride and 11.3% sodium borohydride, 0.22 mole of BFS being absorbed for each mole of NaH initially present. The borohydride yield represented a 53.6% conversion of the hydrde utilized. This value is slightly higher than any obtained in the absence of methoxide catalyst and in addition the product showed no signs 0f `caking, being entirely iree-owing.

An Edwards Analyzer, described in The Analyst, 71, 521 (1946), was utilized throughout these experiments. This instrument, standard for hydride analysis, determines the amount of hydrogen evolved by means of pressure changes. Methyl Cellosolve, which reacts with sodium hydride but not with hydride, was iirst added to the sample under inspection. The amount of hydrogen found was a measure of NaI-I present in the sample. Methyl Cellosolve acidied by small amounts of sulfuric acid was nex-t added to the sample. The hydrogen evolved from the second Cellosolve addition was a measure of NaBH4 present. The method was tested against known samples.

Example 3 Another run was made using 7.54 moles of sodium hydride (181 grams), 150.7 grams of the product shown in Example 2 and 58.9 grams of sodium methoxide. The percentage of methoxide catalyst was increased over that previously used to 58.9 grams, representing 1.09 gram moles or 0.13 mole for each mole of NaH. Boron trifluoride was passed through the reactor for about 3% hours at a tempera-ture ranging from 250 C. to 310 C. Analysis of the product showed an 88.6% conversion of sodium hydride to sodium borohydride, 0.25 mole of BF2. being absorbed for each mole of Nal-I in the original solids.

Example 4 The procedure of Example 3 was repeated with grams of the product of that example, 173 grams (7.2 gram moles) of sodium hydride and 51 grams (0.94 gram mole) of powdered sodium methoxide. The ratio of methoxide to hydride was 0.13 as before. Boron triluoride was introduced at a temperature of 240-300 C. over a period or four hours. The product showed a yield of 87.8% borohydride based on the total sodium hydride used and was free-flowing.

Example 5 The procedure of the previous examples was repeated except that the ratio of NaOCI-Ia to NaH was reduced to 0.1:1. 82 grams of a heel containing 1.8% NaH, 14.2% NaBHl and 9.3% NaOCl-Ia, were mixed with 132 grams (5.5 moles) NaI-I and 22.1 grams (0.41 mole) NaOCHs. BFS was added during a period of 21/2 hours in which the temperature varied from 232 to 288 C. The product showed 12.21% NaBH4, a yield of 86.4%, and substantially no NaH.

Example 6 (a) The Procedure of Example 5 was repeated except that the previously prepared heel was omitted. A charge of 210 grams (8.75y gram moles) of NaI-I and 78.7 grams (1.45 gram moles) of NaOCI-l's were placed in the reactor and milled, the mole ratio of methoxide to hydride being calculated at 0.17. Boron triuoride was admitted for about 3 hours to the agitated mixture at a temperature of 240-290 C. and gave a product showing a yield of only 67.6%.

(b) The triluoride gas was led over the solids for about 21/2 hours more at a temperature of 240-264 C'. The yield was increased tov 77.8% by this treatment but the product was partially packed into a hard cake.

The results of the experiments given in the examples are summarized in Table I.

TABLE I-rRoDUoTroN or NsBH,

evaporated and the borohydride dissolved in absolute ethyl alcohol and reprecipitated with petroleum ether. The nal product was sodium borohydride of 90% purity. Further purification may be carried out at discretion.

. Yield NaOOH; Temp., C. Reaction NaH Heel (Percent Example Moles Grain mof/wle Mflx' gl Borohyb cli-ide) v(250) 7. 92 (l) 0. 065 24U-300 Z/ 53. 6

' (270) 5. 50 82 0. l() 232-288 2% 86. 4 8. 76 0 0. 17 240490 3 67. 6 (3) (2) (2) 24U-264 Zl/ 77. 8

i Cont. from 2a. 2 Cont. from 6a.

It will be seen that the sodium methoxide cata- Various modifications within the scope of this lyst markedly increased the yield of borohydride obtained. Quantities of as low as 0.039 mole of the catalyst to each mole of sodium hydride employed appeared to have some elfect but best results were found when about 0.1 mole of methoxide was used for each mole of the hydride. A mole ratio of 0.17:1 still gave results not quite as good as those obtained with a mole ratio of about 0131i but still above those obtained in the absence of a catalyst. A ratio of up to about 620:1 would therefore seem preferred although higher values can be used with diminishing effect. The alkoxide does not affect the temperature requirements of the reaction, the 150400 C. range being operative with about 250-300 C. being preferred.

The effect of the heel may be seen by contrasting the results of Example 6 with those of Examples 3 or 4. Omission of the heel not only decreased the yield but resulted in a packed product. Accordingly, while a heel from a previous reaction is not absolutely essential, it is desirable. The heel should be added to the hydride in an amount equal in weight to between about 1.0 to 2.5 times that of the latter. Calculations must of course take into account the sodium hydride added with the heel. Other inert materials may be substituted for the reaction product of a previous run but such a product is preferred.

One effect of the added heel is to serve as a mere diluent of the reagents, thereby slowing down the reaction. The steel balls in the ball mill may also serve somewhat the same purpose. The balls may behave in addition as heat exchange media, rapidly lowering an excess temperature occurring at any particular spot. Heat exchange characteristics may also explain the eiiicacy of steel as a constructional material for the closed reaction chamber. While various methods of heating this chamber may be employed, as for example resistance wiring, a flame playing into a furnace and never touching the reactor itself was found of particular value.

The crude reaction product shown above is mixed with NaF, NaH, NaOCHa and side products. It may be employed as a reducing agent in the original mixture or it may be concentrated and purified as desired. A product containing 10-14% borohydride was treated with isopropyl amine in a Soxhlet extractor to give a material containing 51% borohydride. The amine was invention will be evident to those skilled in the chemical arts. Hydrogen and nitrogen have, for example, been shown as coolant or non-reactive atmospheres. Other inert gases may be utilized instead of these two, available substitutes being helium, neon and argon. Furthermore the son dium hydride and methoxide used may be either freshly prepared or commercially obtained Without affecting the results of the process. Still furthermore, the reactants themselves may be varied slightly. LiH, KH and other alkali metal hydrides may be substittued for NaH, producing the respective metallic borohydrides. In addition, other boron halides such as BCls, BBra and mixed halides may be substituted for the BFS, The quantity of boron halide passed over the solids is not strictly critical since, under the given conditions, the reaction substantially stops of its own accord when about one-quarter mole of BF: has been absorbed for each mole of NaH present.

The invention is also not to be understood as restricted to the catalyst exemplied. Other alkoxides may be used instead of the methoxide such as NaOCgl-Is, NaOCsH': and other simple alkaline alcoholates. In addition the phenates, such as NaOCsHs, are available. The more alkaline the catalyst, however, the more desirable it will generally be found, acids tending to promote the decomposition of the borohydride. If the sodium hydride is to be replaced by the hydride of another alkali metal, the catalyst should be an alkoxide of the same metal. Thus with KH it would be desirable to use KOCHs, KOCGHs or another potassum compound. In general the formula of the catalyst may be written as MOR where M stands for an alkali metal ion and R an aliphatic radical of fairly low molecular weight, that is, with a carbon chain up to about ve or six atoms long, or a simple aryl radical of the phenyl type.

The quantity of alkoxide required to serve as an effective catalyst may be expected to vary with both the alkoxide and the hydride employed. The preferred mole ratio of up to 0.20:1 NaOCI-IgzNaI-I, the upper limit being set by the tendency of increased quantities of the methoxide to decrease yields of borohydride perhaps by promoting side reactions, is therefore not necessarily the best ratio for NaOCzHs employed in the same reaction. In like manner differences in preferred ranges and optimum values of catalyst weight will also occur Where LiI-I or KH is substituted for NaI-I, an appropriate catalyst being selected in each case.

In some of the following claims the term catalytic amount appears. This term does not refer to the extremely small quantities sometimes connoted by such an expression. It refers instead to quantities sufcient to promote the desired reaction and in most instances will, as exemplied, amount to an appreciable fraction of the metallic hydride employed. The term furthermore does not imply that any particular theory of the mode of action of the alkoxide is endorsed.

Having now described my invention, I claim:

1. The process of preparing an alkali metal borohydride which comprises heating the hydride or" said metal with a boron halide at a temperature between about 150 and 400 C. and in the presence of an alkoxide of said metal.

2. The process of preparing an alkali metal borohydride which comprises heating the hydride of said metal with a boron halide at a temperature between about 150 and 400 C., in the presence of a catalytic amount of an alkoxde of said metal and in the substantial absence of air and moisture.

3. The method of preparing sodium borohyn dride which comprises heating sodium hydride with a boron halide at a temperature of about 150-400 C. and in the presence of a catalytic amount of a sodium alkoxide.

4. The method of preparing sodium borohydride which comprises heating sodium hydride with a boron halide at a temperature of about 150-400 C., in the substantial absence of air and moisture and in the presence of a catalytic amount of sodium alkoxide.

5. The method or" claim 4 in which the boron halide is boron trilucride.

6. The method of claim 4 in which the sodium alkoxide is sodium methoxide.

7. The method of preparing sodium borohydride which comprises heating together substantially stoichiometric amounts of sodium hydride and boron trifluoride at a temperature between about 150 and 400 C. in the substantial absence of air and moisture and in the presence of a sodium alkoxide, the mole ratio of the alkoXide to the hydride being not above about 0.20 :1.

S. The method of claim 'l in which the sodium alkoxide is sodium methoxide.

9. The method of preparing sodium borohydride which comprises passing gaseous boron triiiuoride ovei a mixture of solid sodium methoxide and sodium hydride, the mole ratio of methoxide to hydride being not abc-ve about 0.2011, while maintaining said mixture at a temperature between about and 400 C.

10. The method of claim 9 in which the temperature is maintained Within the range 250- 300" C.

11. The method of preparing sodium borohydride which comprises reacting at a temperature between about 150 and 400 C. and in the substantial absence of air and moisture, boron triiiuo-ride with sodium hydride in the ratio of about one mole of triuoride for each four moles of hydride, the sodium hydride being admixed with not more than about 0.2 mole sodium methoxide for each mole of the hydride present.

12. The method of claim 11 in which the sodium hydride is further admixed with a heel of product made in a previous run of boron trii'luoride over sodium hydride.

1o. The method of claim 12 in which the mixture of sodium hydride, heel and sodium methoxide is highly comminuted.

14. The method of claim 13 in which the cornminuted mixture is agitated.

15. The method of claim 14 in which the agitated mixture is maintained at a temperature within the range 250-300 C` 16. The method of claim 15 in which the weight of the heel present is between about l and 2.5 times that of the sodium hydride.

1'7. The method of preparing sodium borohydride which comprises passing gaseous boron triuoride at atmospheric pressure over a highly comminuted agitated mixture of sodium hydride, sodium methoxide and a heel from a previous run of boron triuoride over sodium hydride, the methoxide being present in the mixture to the extent oi about 0.13 mole for each mole of sodium hydride and the heel being present in a weight equal to between about l and 2.5 times the weight of the hydride, said mixture being maintained at a temperature of 250-300 C. in the substantial absence of moisture and air, cooling the crude reaction product and separating the borohydride from the other components of said reaction product by dissolving the borohydride in solvent in which said other components are substantially insoluble.

Name Date Winternitz Nov. 28, 1950 Number

Claims (1)

1. THE PROCESS OF PREPARING AN ALKALI METAL BOROHYDRIDE WHICH COMPRISES HEATING THE HYDRIDE OF SAID METAL WITH A BORON HALIDE AT A TEMPERATURE

Images