U s e of Refractive Indices of Dimethyl Esters DAVID F. HOUSTON AND JULIA S. FURLOW Western Regional Research Laboratory Albany, Calif., Bureau of Agricultural and Industrial Chemistry, Agricultural Research Administration, U. S. Department of Agriculture
'
pressed by the following equations, in which t represents degrees Centigrade :
Measurements on purified dimethyl esters of dibasic acids having from six to twelve carbon atoms show that their refractive indices bear linear relations to change in temperature and to weight percentage of the composition of binary mixtures. A change of 0.0001 index unit corresponds to 3 to 6% change in composition of the mixtures, thus approaching b y a simple and rapid measurement the accuracy obtainable b y saponification equivalent determinations.
M
EASUREMENT of refractive index presents a convenient and rapid means for following the progress of a fractional distillation, and extends the information furnished by the observed distillation temperatures. When careful fractionation yields only single components and intermediate binary mixtures, a close estimate of composition may be obtained from a knowledge of the composition-index relations of binary mixtures. This procedure is nicely applicable in the analysis of dibasic acids. Distillation of the dimethyl esters is one of the most satisfactory ways of separating mixtures of the closely related acids, and the approximate compositions of the individual fractions can be readily determined through refractive index measurements, as is shown in this paper. Further corroboration is possible by determination of the melting ranges (3) of the recovered acids.
Dimethyl adipate Dimethyl suberate Dimethyl azelate Dimethyl sebacate Dimethyl 1,ll-undecanedidate Dimethyl 1,12-dodecanedioate
nh
nf, nh nh nh nh
- 40) = 1.4307 - 0.00039(t - 40) =
1.4205 - 0.000406(t
= 1.4262 - 0.000396(t - 40) = 1.4284 - 0.000396(t 40)
= 1.4329. - 0.000386(t - 40) = 1.4345 - 0.00038(t - 40)
The indes values are in agreement with those given by Karvonen
( d ) , and the temperature increments are of the same magnitudes as those found by Mattil and Longenecker (5) for monobasic methyl esters and for synthetic glycerides. 1.440Q
I
I
I
I
1
I
8.
1.4350 X W
P
1.4300 W
/
L
B
f
c
2a 1.4250
PREPARATION OF M E T H Y L ESTERS
I I . W
All esters were prepared from the acids reported in the preceding paper ( 3 ) except the C12 ester, which was fractionally crystallized from an original methyl ester distillation fraction. Properties of the esters are presented in Table I.
a
1.4200
1.4 I 5 0 Table
1. Properties of Esters Used in This Investigation Melting Point of Ester,
No. of
Carbon A t o m in Acid 6 8 9 10 11 12
Boiling P$nt of Ester, C. 71.4- 7 2 . 0 (0.2 mm.) 114.0-114.4 4.7 mm.) 128.0-128.7 l5.2 mm.) 141.9-142.2 (6.0 mm.) 123.0-124.5 (2 mm.)
...
This work
... ... ... 26
17 30.9-31.3
C.
Figure I. Refractive indices of M e t h y l Esters of Normal Monobasic and Dibasic Aliphatic Acids
Literature values
... ...
A. N2," of dimethyl esters of dibasic acids, from Karvonen ( 4 ) E . Ng0 ' of dimethyl esters of dibasic acids, this investigation C. N4,0°.' of methyl esters of monobasic acids, from Mattil and Longenecker (5)
'
26;. i4) 2 0 . 3 (7) 31 (2)
For any single temperature the indices of the homologous esters are represented by a smooth curve as shown in Figure 1. Corresponding indices for monobasic methyl esters (6) are included for comparison, as are those of Karvonen for the dibasic esters a t 20" C. (value for methyl sebacate is calculated from 28" C. by the equation presented above).
MEASUREMENT O F REFRACTIVE I N D E X
Measurement of refractive index was performed with an Abbetype refractometer which could be read t o 0.0001 index unit. Temperatures reported for index readings were controlled to *0.05" C. by rapid circulation of water from a constant-temperature bath through the prism jackets of the refractometer. All determinations were made on esters which had been freshly distilled or stored under reduced pressure to avoid errore caused by dissolved gases.
COMPOSITION-INDEX RELATIONSHIP I N BINARY MIXTURES
Several series of binary mixtures were Prepared which represented various possible types, and their refractive indices were measured a t 40" C. The composition-index relation could be represented by a straight line in all cases when the composition was expressed as weight percentages, as shown in Figure 2. The relation was nonlinear for molecular percentages. Here again the results are in accord with those of Mattil and Longenecker for monobasic methyl esters. As the difference in indices of adjacent homologous esters varies from about 0.0030 between Cg and C, (ng.' for dimethyl
TEMPERATURE-INDEX RELATIONSHIPS
Indices were measured a t 5' intervals in the range from 25' to 60" C. for all temperatures a t which the esters were liquid. The expected straight-line relationships were found, and are ex-
' Second article in series:
first appears on page 538.
541
INDUSTRIAL AND ENGINEERING CHEMISTRY
542
Vol. 18, No. 9
INTERFERENCES 1
,
4
3
5
0
~
One possible source of error is the presence of dissolved gases.
If the liquid dimethyl esters are exposed to the atmosphere, the refractive index will drop several fourth-place units over a period of a day or so. Original values can be obtained following redistillation or degassing, but are not achieved by the use of desiccants. This type of error can be avoided readily by making index determinations promptly after distillations, or by storing the liquid esters at reduced pressures. Methyl esters of monobasic acids such as myristic or palmitic may be present in products from oxidative cleavage (1) of unsaturated acids from various sources, although they would normally be removed by partitions prior to analysis of the dibasic acids. If any remained, they would interfere nith the determination of composition by means of refractive index. Llethyl myristate, for example, distills (6) in the temperature range between dimethyl sebacate and undecanedioate, but has a refractive index almost as lo^ as dimethyl azelate (Figure 1). In such a case the calculated composition would indicate too large a proportion of the loiver molecular-Xveight component. However, the apparent composition derived from determination of the saponification equivalent would be in error in the opposite direction. Accordingly, agreement between the results obtained by the two methods would confirm the absence of monobasic esters and enhance thr validity of the analysis. WEIGHT PERCENT OF HIGHER MOL- WT. ESTER Figure 2.
Refractive Indices of Binary Mixtures Esters of Dibasic Acids at 40" C.
of Dimethyl
Numbers designate carbon content of component acids
pimelate estimated graphically as 1.4235) to 0.0016 between CH and C12, a difference of 0.0001 in refractive index corresponds to a 3 to 67, change in composition in binary mixtures of adjacent esters. This is about the same precision obtainable from equivalent weight determinations in which a difference of 0.2 unit corresponds to approximately 3V0.
LITERATURE CITED
(1) Armstrong, E. F., and Hilditch, T. P., J . SOC.Chem. Ind., 44, 43T (1925). (2) Chuit. P.. Helv. Chim. Acta, 9, 267 (1926). i3j Houston, D. F., and Van Sandt, W. A , , IXD.ENG.CHEM., ASAL. ED., 18, 538 (1946). (4) Karvonen. A., Ann. Acad. Sci. Fenn., [AIIO, N o . 5 , 10 (1917) (in German). (5) Mattil, K. F., and Longenecker, H. E., Oil and Soap, 21, 16 (1944). (6) Norris, F. d.,and Terry, D. E., Ibid., 22, 41 (1945). (7) Verkade, P. E., Coops, J . , Jr., and Hartman, H., Rec. trau. chim. PnU+Bas, 4 5 , 600 (1926).
Volumetric Determination of Magnesium in Magnesium Carbonate 'Ores LORING R. WILLIAMS Department of Chemistry, University of Nevada, Reno, Nevada
I
N DETERMINING the magnesium content of carbonates, many investigators have used a measured excess of standard alkali to precipitate magnesium hydroxide and have determined the excess with a standard acid. The outstanding difference in the methods is in the indicators used: Kolthoff (3) used phenolphthalein and dimethyl yellow; Pierce and Geiger (5) used trinitrobenzene and bromophenol blue; and Willstatter and Waldschmidt-Leitz ( 7 ) used thymolphthalein. Most of the methods employed more than one indicator and specified the use of carbonate-free standard alkali solution. In many investigations the emphasis seemed to be on the saving of time rather than a high degree of accuracy. The chief objectives of this investigation were (1) to develop an accurate routine method for the volumetric determination of magnesium that could be carried out in a reasonable length of time and would not require carbonate-free standard alkali, and (2) to find a single indicator that could be used for all the neutralizations required in the procedure.
THEORETICAL
I n selecting a single indicator for this investigation two conditions had to be satisfied: The iron and aluminum, Jvhich are the most common impurities found in carbonate ores, should be completely precipitated as the insoluble hydroxides at the end point, and the indicator should show the true equivalence point when standard alkali which contains carbonates is neutralized by standard acids. Blum ( I ) states that the precipitation of aluminum hydroxide begins at pH 3 and is complete at or before pH 7 and recommends the use of methyl red as the indicator. Since ferric hydroxide is completely precipitated under the same conditions,,aluminum and iron can be removed quantitatively at the methyl red end point. There should be no postprecipitation of magnesium hydroxide if the neutralization is carefully carried out. At the methyl orange end point (pH 4)the absorption of carbon dioxide shows no decrease in the effective normality of a standard alkali solution (6); therefore, methyl red (pH 4.2 to 6.3) should
