REAGENTS FOR WATER DETERMINATION IN SAMPLES CONTAINING IODINE-REACTING INTERFERING SUBSTANCES
FIELD OF THE INVENTION
The present invention concerns the quantitative determination of water in materials. It provides unbiased (corrected) results of water determination in samples containing iodine-reacting substances, which interfere with the Karl Fischer methods for water determination.
BACKGROUND OF THE INVENTION
The measurement of water in oils, drugs, cosmetic products, foodstuffs, chemical products and other materials has broad economic importance. For example, transformer oils must have less than 50 ppm of water in order to be acceptable for use as insulating fluids. The water content of lubricating oils and hydraulic oils must be low in order to function efficiently. Fuel oils and aviation fuels have specific limits on water content. Crude oils contain significant amounts of water (0.02 - 2%) particularly when recovered by the steam method and this water must be subtracted from the total volume of oil at the time of purchase.
In 1935 Karl Fisher (KF) published his method for water determination which is based on the reaction of water with iodine in presence of sulfur dioxide, pyridine, and methanol. The iodine quantity spent for the KF reaction can be controlled volumetrically or coulometrically. Later pyridine, because of its toxicity, was displaced by other amines or heterocyclic compounds. Methanol, because of its side reactions, is displaced sometimes by other alcohols. These modified KF volumetric and coulometric methods are widely used today (D.A.Skoog, D.M.West,
and F.J. Holler. Fundamentals of Analytical Chemistry. Saunders College Publishing, 7th ed., 1996, pp. 381-383).
In many cases oxidizable additives are used in products such as lubricating oils, hydraulic fluids, and fuels that interfere with the measurement of water by the KF methods. Crude oils also contain mercaptans and sulfides that interfere with the KF water determination. The interference is caused by the side reactions of iodine (a component of the KF reagent) with the oxidizable substances of the sample under analysis. ASTM method D 4377-88 cover the water determination in crude oils at levels of interfering substances not more than 500 ppm (as sulfur): for high water concentrations their influence is negligible. However, the level of interfering substances in some crude oils can be close to the water concentration. AOAC Official Method 984.20 for water determination by KF method is not applicable to oils and fats containing oxidizable interfering substances. AOCS Official Method Tb2-64 (Modified KF Reagent) for water determination in industrial oils and derivatives has the same limitations. So, it would be extremely useful and economically important to develop a reagent that would neutralize those substances that react with the iodine in the KF reagent and spuriously increase the results of the water determination.
USA Patent No. 5,750,404 to Sherman, et al. describes reagents which have been developed to measure and neutralize ene-diols and thiols by reacting these compounds with iodine before the water is measured by the KF method. The reagents described in US 5,750,404 consist of iodine, an iodide and a base in methanol or another alcohol mixed with N,N-dimethylformamide, or with formamide, or with dimethyl sulfoxide and N,N-dimethylformamide mixture as a
solvent. These reagents can be used for ene-diols and thiols only which are polar interferences (F.Sherman, I.Kuselman, A.Shenhar. Talanta, 1996, Vol. 43, pp.1035-1042; I.Kuselman, F.Sherman, T.Burenko, A.Shenhar. J. AOAC International, 1999, Vol. 82/4, pp. 840-861). Note, the solvents in the reagents are also polar. It is important not only for dissolution or extraction of ene-diols or thiols contained in a sample under analysis, but also for the stoichiometry of the KF reaction (F.Sherman and I.Kuselman. Accreditation and Quality Assurance, 1999, Vol. 4, pp. 230-234).
The drawback of the reagents described in USA Patent No. 5,750,404 is that these reagents are not suitable for analysis of non-polar samples (for example, oils) since they can not dissolve or extract interfering substances contained in non polar sample matrixes.
SUMMARY OF THE INVENTION
Novel reagents are proposed for correction of results of KF water determination in oils, drugs, cosmetic products, foodstuffs, chemical products and other materials containing iodine-reacting substances, which interfere with the Karl Fischer methods for water determination (also referred to as interferences).
The novel reagents consist of iodine, iodide, a buffer and a mixture of polar and non polar non- aqueous solvents. The novel reagents can be used in the two ways: 1) for determination of the interferences content and deducting the result of this determination (in water units) from the result of KF titration, and 2) for consecutive neutralization of the interferences by the novel reagent and then water titration by KF reagent in one and the same test portion. Since the novel reagents contain at least one non polar solvent, they are applicable for both polar and non polar samples.
DETAILED DESCRIPTION OF THE INVENTION General description
The present invention relates to novel reagents for water determination in materials containing iodine-reacting interfering substances. The novel reagents are used for measurement and neutralization of substances which can react with iodine and therefore interfere with water determination in non-aqueous media.
The term "iodine-reacting interfering substances" (also referred to as interferences) relates to the substances in a sample or matrix, except water, which can react with iodine. The reagents of the invention contain iodine, an iodide, a non-toxic organic or inorganic buffer and a mixture of both polar and non polar non aqueous solvents. The presence of at least one non-polar solvent in the novel reagents allows the extraction and dissolution of the interfering substances not only in polar media (for example, ene-diols or thiols), but also in non polar ones (for example, oils).
It is essential that the reagents have a composition as close as possible to the composition of the volumetric and coulometric KF reagents to insure the same conditions for the iodine reactions. They must contain all the components of the KF reagents applicable for the sample under analysis except the sulfur dioxide and use vessel solvents suitable for the sample. This is necessary in order to guarantee that the sample is completely soluble in the assay medium or that interfering substances and water are completely extracted from this medium.
Since the contents of the reagents are similar to those used for the KF reagent, but lack sulfur dioxide, they can not react with water. Based on these
properties, the novel reagents can be applied for correction of the results of KF titration in the following two ways:
1) Two test portions of the sample are titrated independently: the first one against the novel reagent, and the second one against the KF reagent. The correct water content Cw, ppm, is calculated by deducting the result CNR of the sample titration against the novel reagent expressed in water units (ppm) from the result
CKF of KF titration of the sample by the formulas:
(1) CW = CKF - CNR , (2) CKF = VK TKF 100u mι,
(3) CNR = 0.071 VNRTNRI 000 /m2
where TKF and VKF are the water titer of the KF reagent, mg H2O/ mL, and its volume, mL, spent for titration of test portion 1 ; T'NR and VNF are the iodine titer of the novel reagent, mg J2/ mL, and its volume, mL, spent for titration of test portion 2; 0.071 is the ratio of the molecular masses of water (18) and iodine (254) at the stoichiometric ratio between water and iodine in the KF reaction equal 1 ; mi and m2 are the masses of the test portions 1 and 2, g.
2) The KF titration is performed after neutralization of interferences by the novel reagent in one and the same test portion, and the correct water content is calculated by the formula:
(4) Cw = (VKFTKF -VNRTNR) 1000/m ,
where TNR is the water titer of the novel reagent, mg H2O/ mL, VNRTNR is the traces of water introduced in the titration vessel with the novel reagent, mg H2O, and m is the mass of the test portion, g.
The methodology that is being proposed has general application to the measurement of water content in any matrix that interferes with the volumetric or coulometric KF reagents by reacting with iodine and cause erroneously high values of water content by the KF methods.
2. Examples
2.1. Reagents
In the following three reagents-examples methanol plays a role of a polar solvent while xylene and chloroform are non-polar solvents: Reagent 1 - 0.01 mol iodine, 0.015 mol potassium iodide and 0.35 mol sodium acetate dissolved in one liter of the mixture of methanol with xylene (3:2) vol.
Reagent 2 - 0.01 mol iodine, 0.015 mol potassium iodide and 0.35 mol sodium acetate dissolved in one liter of the mixture of methanol with chloroform (3:2) vol. Reagent 3 - 0.02 mol iodine, 0.015 mol potassium iodide and 0.35 mol sodium acetate dissolved in one liter of the mixture of methanol with xylene and chloroform (3:2:2) vol.
2.2. Titers of the reagents and their use
Reagents 1-3 are stored for two days to permit stabilization. Their iodine titers were obtained by titration against 0.1 N thiosulfate using Metrohm
Titroprocessor 672 and Dosimat 655 with a 10 mL burette. The end point potential was 340 mV and the end point was determined automatically when no titrant was added for 30 s. The same solvents were used in the titration vessel as in the reagent. Results of the linear regression analysis of the dependence of the reagent volume spent for titration versus the thiosulfate volume (y = b0 + ^x) are shown in Table 1. The titers T'NR, mg l2 / mL, are calculated by the formula:
where 254 is the molecular mass of iodine, Cts = 0.1 N is the concentration of thiosulfate, 2 is the stoichiometric coefficient in the reaction between iodine and thiosulfate, bi is the corresponding slope from Table 1. One can see that slope and precision values for titration of the standard (thiosulfate) with the three reagents are similar. So, they can be used with a similar response range for the measurement and neutralization of iodine-reactive substances that interfere with KF measurement of water.
To determine water traces in the reagents (water titers) 10-40 mg dry SnCI2 were titrated against the novel reagent, and then in the same vessel the spent quantity of the novel reagent was titrated against Hydranal Composite 1 diluted by xylene (3:1) vol. Results of the linear regression analysis of the dependence of the KF reagent volume versus the novel reagent volume are shown also in Table 1. The titers TNR, mg H2O/ mL, are calculated by the formula:
(6) TNR = TKF bι,
where TKF = 0.63 mg H2O/ mL is the water titer of the KF reagent
determined against the Hydranal Standard 5 (containing 5.00 ± 0.02 mg H2O/ mL).
Reagents 1-3 contain different quantities of water but all of them are titrated satisfactory by KF reagent.
Table 1
Results of the linear regression analysis of the properties of the novel reagents.
Reagent Titrant, Range of Number of Intercept Slope b^ Correlation Titer,
No. y(χ) x, mL titrations bo, mL ml_/mL coeff. unit square
1 NR(TS)* 0.1-1.3 9 -0.02 8.22 0.9996 1.55 mg J2/mL
2 NR(TS) 0.3-1.8 9 0.27 10.80 0.9995 1.18 mg J2/mL
3 NR(TS) 0.3-3.9 13 -0.00 3.79 0.9999 3.35 mg J2/mL
1 KFR(NR)* 0.9-8.5 6 0.17 2.08 0.9995 1.32 mg H2O/ml_
KFR(NR) 1.6-31.0 7 0.04 0.24 0.9990 0.15 mg H2O/ml_
KFR(NR) 5.7-34.2 6 -0.00 0.08 0.9990 0.05 mg H2O/ml_
* Note: NR is the novel reagent, TS is thiosulfate, and KFR is the Karl Fischer reagent.
The novel reagents are used for the correct water determination in the NIST reference materials RM 8506 (Univolt N61 transformer oil), RM 8507 (Coray mineral oil), SRM 2721 (Yates crude oil) and SRM 2722 (Refugio crude oil).
Results of the determination are summarized in Table 2. Results of the KF titration
CKF are shown with their standard deviations calculated from 20 replicates. One can see that concentrations of interfering substances in the oils CNR evaluated by the novel reagents are more than these deviations, especially for crude oils. The two crude oils (SRMs 2721 and 2722) contain significant amounts of sulfur compounds, 1.6 and 0.22 mass % respectively. The ten-fold difference between the interferences measured by this new method is of the same order of magnitude as the difference in sulfur content of the two crude oils. Thus not all of the sulfur-containing compounds may react with iodine. However, approximately 90% of the KF reacting material in the crude oil samples represent iodine-reacting substances other than water. In the case of the transformer and mineral oils the amount of these interferences are much smaller.
Table 2
Results of water determination in oils using the novel reagents (ppm)
Oil Result of KF Result of titration Correct water titration C«F of interferences content Cw CNR
Transformer oil, 40 ± 3 12 28
RM 8506
Mineral oil, 77 ± 5 8 69
RM 8507
Yates crude oil, 926 ± 19 836 90
SRM 2721
Refugio crude oil, 96 ± 2 83 13
SRM 2722
Note, using the reagents described in USA Patent No. 5,750,404 with the same concentrations of iodine, potassium iodide and sodium acetate as in Reagents 1-3, but in the mixture of the polar solvents methanol and dimethylformamide (3:1 by volume), no interfering materials were detected in the oils.
The advantage of the novel reagents in comparison to the reagents described in USA Patent No. 5,750,404 is that the novel reagents described herein allow the measurement and neutralization the interfering substances and the correct determination of water in both polar and non polar samples.
It will be appreciated by persons skilled in the art that the reagents of the present invention are not limited by what has been particularly exemplified above. Rather the reagents of the invention may comprise iodine, an iodide
(potassium iodide, tetramethylammonium iodide or another), and a buffer (such as sodium acetate, urea or diethanolamine) adapted to provide a pH suitable for the KF reaction. To prepare the reagents these components should be dissolved in a mixture of polar solvents (methanol or other alcohol, formamide, dimethylformamide and so on) and at least one non polar solvent (such as chloroform, xylene or toluene) which are able to dissolve or extract the iodine-reactive interfering substances containing in the sample matrix.