RESEARCH MEMORANDUM

A METHOD FOR DETERMINING T E E COMPOSKTION OF

METHANOL - T R I B T H Y L BORATE MIXTURES

B y Samuel Kaye and Frank Sordyl

Lewis Flight Propulsion Laboratory Cleveland, Ohio

NATIOQAL ADVISORY COMMITTEE FOR AERONAUTICS

WASHINGTON November 3, 1855

https://ntrs.nasa.gov/search.jsp?R=19930089063 2020-04-02T08:52:50+00:00Z

NACA RM E55H15

F

NATIONAL ADVISOKY COMKLTI'EZ FC[R AERONAUTICS

A METHOD FOR DETERMJXING THE COMPOSITION CtF

METHANOL - BCIRATE MIxtrmRES

By Samuel %ye and Frank Sordyl

SUMMARY

A method was develaped for rapidly determining the composition of methanol - trimethyl borate mixtures. Pure methanol and trimethyl borate were prepared, and their propert ies were accurately determined. The densit ies and refract ive indices of several mixtures of these compounds were determined. Analysis of' the curves obtained shaved that greater accuracy i n determining the composition was possible from a density de- termination. The conpositions of the mixtures prepaxed were checked

The accuracies of the analytical techniques are discussed.

r*

" potentiometrically using precautions necessary for accurate results.

INTRODUCTION

During an investigation of reactions of trimethyl borate, a simple rapid method for determining the composition of methanol - trimethyl borate mixLures w a s desired. A method based on physical properties was considered t o be most suitable. The l i t e r a tu re , however, disclosed con- siderable disagreement (refs. 1 t o 1 2 ) as to the propert ies of trimethyl borate and its azeotrope w i t h methanol.

An investigation was therefore undertaken at the NACA L e w i s labora- t o r y t o determine: (1) the properties of trimethyl borate of known purity, (2) the propert ies and comgosition of the methanol - trimethyl borate azeotrqpe, and (3) the appl icabi l i ty of refractive index and density in determining the comgosition of methanol - trimethyl borate mixtures. A method based on the change in densi ty w i t h composition was found t o give analyses of suf f ic ien t accuracy.

8

Y

2

Materials

NACA RM E55815

"

c

Methanol. - Cammercial methanol waa pur3fied by treatment with sodium hydroxide fo l lmed Fy f r ac t ipna l d i s t i l l a t i on i n a 6-foot Podbielniak column operated at an efficiency rating equivalent to about 100 theoret- i ca l p la tes . In order tQ ensure accurate data by preventing contamination with moisture, a l l manipulations were performed i n a closed system or i n a dry box made iner t with nitrogen. The properties of the purifled . methanol agreed with the best va lues given in the Internat ional Cri t ical Tab l e 8 .

Methanol - trhethyl borate azeotrope. - The crude e s t e r was obtained from a commercial source labeled as a 70 - 30 m i x t u r e by weight of e s t e r and methanol. It w a s purified by dist i l lat ion in the Podblelniak column described in the preceding paragraph and was stared and handled in a dry box to p ro tec t it from hydrolysis.

Trimethyl boride. - The azeotrope consisting of trimethyl borate and methanol was broken by treating it with lithium chloride, as de- scribed in r.eference II. The trimethyl borate was then purified by fYact ional dis t i l la t ion under nitrogen atmosphere i n the Podbielniak column.

The pur i ty of the es te r was determined by calculatfon (ref. 13) as follows: The freezing point and the freezing pofnt for zero impurity were obtained by analysis of the f r e e z a curvej the heat of fusion was determined .to be 1993 calories per male. From these data, the purity of the sample was determined t o be 99.92 mole percent.

The i d r a r e d spectrum of trimethyl borate, which is not presen-kd herein, w a s essent ia l ly the same a0 that recorded by Samuel P. Sadtler & Son Inc., Phil. (Fa.), except for a bsnd which *eared i n their spectrum a t 3.0 t o 3.1 microns. The presence of this band in their spectrum wag uudoatedly caused by a hydroxyl impurity from methanol f n t he i r sample.

Properties

Densit . - The densit ies were determined, unless otherwise noted, at 20.0 4 N.lo C w i t h a calibrated Sprengel type pycnometer (ref. 14). The average deviation f o r .-se determinations was 0.0002 unit.

I

k Index of refraction. - The indices of refract ion w e r e measured at

20.00 &.lo C w i t h a.Bausch and Lomb Abbe 56 refractometer. The imtru- ment-was in s t a l l ed i n a dry box made i n e r t with nitrogen, and a l l meas- urements were made under these conditions.

*

NACA FM E55Hl-5 Y

3

Freezing point. - The thermometric system consisted of a platinum Y resistance thermometer, a Mueller type G-2 resistance bridge, and a highly

sensit ive galvanometer. This apparatus is the same as is commonly used with the standard technique (ref. 13). The freezing tube itself, how- ever, contained a helical solenoid-operated, reciprocating Stirrer and was charged with 10 m i l l i l i t e r s of sample i n a dry box. The tube was then sealed and the freezing point determined i n t h e usual manner.

Boiling point. - The boiling points were determined i n an ebulli- ometer equipped with apparatus of the s m type as was used fo r the, de- termination of the freezing point. The measurements were made i n a nitrogen atmosphere at a pressure of 760 millimeters of mercury w i t h an accuracy of ;to. 050 C .

Infrared spectrum. - The infrared spectra were obtained on a Baird double-beam recording spectrophotometer with a 0.1-millimeter sodium chloride cell , which w a s filled in a dry box.

M Heat of fusion. - The heat of fusion was determined i n a nitrogen s 2 n - atmosphere by the method of reference 15. I n this method, several

samples with @own heats of fusion are frozen under standardized condi- t ions in a vacuum-jacketed tf ie. The r a t e of heat transfer at the

curve for the tube. Rnawing the rate of heat transfer and the tenrpera- ture gradient, the heat of fusion can be calculated for an unknown com- pound when it is frozen under the same standardized conditions.

3 -I var ious temperature gradients is.then measured t o give a cal ibrat ion

Procedure

The pure materials, methanol and trimethyl borate, were placed in a dry box purged w i t h nitrogen. A se r ies of eight compositions by w e i g h t , accurate t o M.l m i l l i g r a m , were prepared and stored in careful ly s top- pered bottles. The bot t les were opened only i n a nitrogen atmosphere i n which the measuring instrunments were located. The refractive index and density were determined f o r each composition under these conditions. The density and refractive index of the highly purified azeotrope were a l so determined.

RESULTS AND DISCUSSION

Table I shows the methanol - trimethyl borate data p l o t t e d i n f i g - ures 1 and 2 as density and refractive index, respectively, against con-

compared with the straight lines obtained in reference 7. I n tab le I1 are listed the physical properties obtained fo r pure trimethyl borate and the methanol - trimethyl borate azeotrope and, f o r comparison, some

0 centration of trimethyl borate. These curves show definite curvature as

I

4 NACA RM E55R15 t

of the properties reported previously in the literature. The equations obtained by the- least-squares method fo r t he curves ahown in f igures 1 .

and 2, respectively, are t

d2' .= 0.7912 + 0.1199 x 10-2x + 0.2141 x 10-5x2 4 . . . . .. -. . . . . . - - . . -_ . ."

and

n20 = 1.3285 + 0.2584 X10'3~ + 0.3123X D

where x is the percent by weight of trimethyl borate.

The density of-the azeotrope was determined, and i t 6 compoeition w a s calculated from the density equation to be 74.86 percent trimethyl borate and 25.14 percent methanol. This compares with 75.5 and 24.5 percent, respectively, reported in reference 11.

The composition of the azeotrope which i s sham to fall on the curve was checked independently by t i t ra t ing the bor ic acid formed on hydrolysis of the mixture. The t i t r a t i o n was done with a Beclanan Model G pH meter with mannitol used as the. activator. The va lues coincided only after the principles expressed by SchfiSer (ref. 16.) were applied.

S c m e r showed that the pH a t the beginning and- e d of t he t i t r a t ion curve of Ii3BCg with alkali dGpended on the amount of mannitol used as activator. The volume of alkali required t o reach the equivalence point did not change, of course. Therefore, when a part icular p H was used as the cr i ter ion for- the end point, the value f o r equivalence changed with the amount of mannitol used.

In some reported procedures the boric acid i s arbitrarily titrated between some i n i t i a l f i x e d pH value ana a f i n a l value. The adjustment of the i n i t i a l pH is necessary i f t h e boric acid is to be determined i n the presence of strong acids. However, when H3B03 alone is present, this technique is obviously incorrect since an amount of boric acid, depending on its concentration, will thereby be neutralized. Thia error can amount to several tenths of a percent of boron.

In order to t i trate boric acid with an accuracy of at least 0.1 per- cent, therefore, the following variables must be fixed: (1) the amount of-mannitol used ae activator, (2) the pH range over which the t i t r a t ion is performed, and (3) the relative concentration of boron t o be t i t ra ted . An evaluation of the reliable-methods reported f o r this determination of boron (ref. 17) showed tha t the procedures which use these principles either inadvertently or by design do yield correct values. %

A final important correction which must be applied in the analyses is one which accounts for the pH of the dfstilled water used f o r

Y

MACA RM E55Hl5 J

5

hydrolysis and as a solvent. The pH of t h e d i s t i l l e d water a t t h i s lab-

however, are corrected for water at pH 7. " oratory varied at different times from 5.7 t o 6 . 4 . The resul ts reported,

Further confirmation of the composition of the azeotrope was obtained by t reat ing it as a pure substance of determined composition. The azeo- trope was mixed with pure trimethylborate. The densi t ies of these trimethyl borate - trimethyl borate azeotrope mixtures f e l l on the same curve as before. w

CD cn rp The re la t ion of both density and refractive index to composition

shows very slight deviations from l inear i ty . The change in dens i ty can be read to &0.0002 gram per milliliter, which represents a kO.13-percent change i n composition. The change in refractive index can be r ead t o about 10.0002, which represents a maximum of about ~0.68-percent change i n composition. It can be seen, therefore, that the determination of density can'be about f i v e times as accurate for the estimation of composi- t ion as the refractive-index determination. It is therefore more appli- cable t o t h e problem than the refractive-index determination, though the

.A l a t t e r is fas te r .

An assumption involved i n t h i s work is tha t t he mixture is reasonably

material studied w a s proven by analysis. Tbis assumption f o r most samples of methanol - trimethyl borate i s not unreasonable in v i e w of the method of synthesis, which involves simply the reaction of boric oxide with methanol

-.I f r ee of contaminants which could affect the results. The pu r i ty of the

B2O3 + 6CE3OH +2B (OCH3)3 + 3H20

or 4 se t of reactions equivalent to the equation 8,s written. Although excess boric acid precipitates out on hydrolysis, hydrolytic decompsi- t i o n may a f f ec t the results because of the so lub i l i t y of boric acid i n the liquids .

Lewis Flight Propulsion Laboratory National Advisory Committee for Aeronautics

Cleveland, Ohio, August 17, 1955

REFEKFNCFS

1. Etridge, Joseph John, and Sugden, Samuel: The Parachor and Chemical * Constitution. Pt . M - Boron Compounds. Jour. Chem. SOC. (London),

pt . I, 1928, pp. 989-992.

' d 2. Gasselin, M. V.: Action du fluorure de bore sur guelques composks organiques. Ann. Chim, et de Phys., t. 9, 1894, pp. 5-83.

6 - NACA RM E55KL.5

t

3. k c a t , Maurice: Nouveaux adotropes binaires. Rec-il des Travam Chim. des Pays-Bas, t. 47, 1928, pp. 13-18, I

4 . Seaman, William, and Johnson, John R.: Derivatives of Phenylboric Acid, Their Preparation and Action Upon Bacteria. Jour. Am. Chem. SOC., vol. 53, no. 2, Feb. 1931, pp. 711-722.. . .. "

substances organiques. VI. Nouveues &terminations expdrimentahs. 3ull. chim. SOC. Belg . , t. 30, nos. 8-9, 1921, pp. 213-219.

5. Timermans, J., et Mattaar, J. F.: La Temp6rature de cong6lation des

$ OD

6. Webster, Stewart H., and Dennis, L. M.: Preparation and Purification m of Methyl Borate and Ethyl Borate. Jour. Am. Chem. SOC., vol. 55, no. 8, Aug. 1933, pp. 3233-3235.

7. Longosz, E. J., Jones, J. R. ; and Zaslowsky, J. A. : Preparation OP Pure Trimethyl Borate from the Borate-Methanol Azeotrope. Tech. Res. Rep. NC-1023=TR-30, lvlathieson Chem. Co., June 2, 1954,

8. Cooper, W. J., and Messner, A. E.: Vapor Pressures and Densities of Methyl Borate - Mineral O i l Mixtures. Rep. CCC-1024-TR-92, Ca1let.y Chem. Co., Feb. 7, 1955.

9. Cooper, W. J.: Vapor Pressure of Methyl Borate. Rep. CC-1024-TR-71, Callery Chem. Co., Nov. 17, 1954.

10. Leffler, A. J., and Kendrick, T. R.: Determiastion of Purity of Methyl Borate by Freezing Point. Rep. CCC-1024-TR-106, Callery Chem. Co., Apr. 14, 1955.

11. Schlesinger, H. I., Brown, Herbert C . , Mayfield, Darwin L., and Gilbreath, James R.: Procedures for the Preparation of Methyl Borate. Jour. Am. Chem. SOC., vol. 75, no. 1, 1953, pp. 213-215.

12. Bradley, James A . , Tul ly , Thomas J., Roeder, Jack R. , and Lange, Willim: The Synthesis of Methyl Borate. Rep. KC-1023-TR-62, Newark College -of Eng. , June 1954.

23. Glasgow, Augustus R. , Jr., S t r e i f f , Anton J., and Rossini, Frederick D. : Determination of the Purity of Hydrocarbons by Measurement of Freezing Points. Jour. Res. Nat. Bur . Standards, vol. 35, no. 5, Nov. 1945, pp. 355-375.

14. Lipkin, M. R., Davison, J. A., Harvey, W. T., and Kurtz, S. S., Jr.: Pycnometer for Volatile Liquids. M s . and Eng. Chem., awl. ed., vol . 16, no. 1, Jan. 15, 1944, pp. 55-58.

2 .

MACA RM E55EU.5 - 7 * *

15. Rossini, Frederick D.: A Simple Calorimeter f o r Heats of Fusion - Hemimellitene, 0- and m-Xylene, and on Two Transitions of Hemimellitene. Jour. Res. N a t . Bur . Standards, vol. ll, no. 4,

I Data on the Fusion of Pseudocumene , Mesitylene (a and P) ,

Oct. 1933, p ~ . 553-559.

16. S&er, Harald: Uber die Aktivierung der Borsh re mit Polyoxyver- bindungen und die Vorggnge bei der mssanalytischen Bestimmung der aktwierten Bors&re. 2s. f . anorg. und allegem. Chem., Ed. 247, 1941, p ~ . 96-ll2.

17. DeFord, Donald D., Lucchesi, Claude A. , and Thoburn, James M.: Analytical Chemistry of Boron. Dept. Chem., Northwestern Univ., June 1953.

8 "--

METHANOL - TRlMETHyL BCIRATE MlxTuKES

Zoncentration ~l? tr imethyl 3 or ate, percent by ae ight

0 19.71 39.84 54.93 67.06 70.12 72.71 79.95 90.09 100.00

Density a t

20° c

0.7912 .8157 .8424 .a637 .a812 .8859 .8895 ,2010 .9164 .9327

Refractive index,

4 O

1.3286 1.3337 1.3392 1.3436 1.3470 1.3485 1.3492 1.3512 1.3543 1.3574

NACA RM E55Hl5

HACA RM E55H15 9

t

I

Materlal IFreezing I IDeosity, R e f r a c t i v e

Trimethyl borate

-28.313 d-28.47

1-34 g-29.W.1

b55-56 %. 91630 e65 b. 919 h67 .5-68.5/7SCrmm ,9193' j68.5 k. 92q;

68.86/?6Chm.1 J Jm. 928

f. 932 .9327

n. 9502

1.3574 1.3558

hl. 3488

Trimethyl borate - methanol azeotrope

54.41 Ojh54.6 k54-56

I 0,8930 1 1.3494

percent by w e i g h t

dz643

"H, =75,1

h75.5

&Except as noted. bRef . 2. 'Ref. 8. 'Ref. 10.

'Ref. 5. hef . 14. %ef. 7. 'Ref. 9.

eRef. 12. fRief. 4.

%Ref. 6. hRef. ll.

%ef. 9 gives the equation = 1.3018 - 1.2667 x T valid from 23O to 29O c. %ef. 1. ?Ref. 3.

. . .

.7d I I I I I I I I I I I I I I I I I I I I I I I I I 0 a 16 24 32 40 44 56 64 72 80 88 96

Concentration of trimethyl borate, percent by weight

Figure 1. - Dsmity of methanol - trimethyl borate mFxturea.

104

. ..... . . ..

t. 3

. . . . . . .

L 1 1 Y

P

E 1.360

w UI

1.356 E rn

1 I 352

1.348

1.344

1.340 I 1.336

1.332

1.328 4

Concentration of trimethyl borate, percent by weight

Figure 2. - Refrqotlve Index of methanol - trimethyl borate mixturee.

c