1、- r :a Testing of Mineral Oil HydrocarbonsandSolvents Determination of Water Content according to Karl Fischer Indirect Method z O DIN 51 777 Part 2 VI al u U a U c m .- E Q c .- E a L C .x U U m L z UDC665.7.033.2: 662.75: 620.1 :=.a2 DEUTSCHE NORMEN September 1974 Dimensions in mm 1 Range of appli
2、cation Mineral oils and mineral oil products as well as solvents miscible with water but not in every ratio, provided that the point at which they begin to boil according to DIN 51 751 is not less than 200 OC. The procedure according to this Standard is also applicable to used mineral oils. The dire
3、ct method for the determination of the water content is described in DIN 51 777 Part 1. The results obtained by these two methods are compar- able with each other. Water contents from 3 mg/kg and above can be determined. 2 Purpose The method according to this Standard is used for determining the wat
4、er content of the products named in Section 1. The water content, sometimes combined with other test results, is used to define and assess mineral oil hydrocarbons and solvents. 4 Brief description of the method 4.1 Method using atmospheric pressure The water is driven out of the specimen at atmosph
5、eric pressure and at a temperature of 120 OC by very pure nitrogen; it is then condensed at room temperature in methanol which has been dewatered with Karl Fischer solution (KFL) 111 and finally titrated with KFL by the dead stop method. The water content of the specimen is calculated from the quant
6、ity of KFL used in the titration and from its titre. 1) Instead of mg/kg, the unit “parts per million“ (abbreviation: ppm) is widely used. Ground socket A NS 14.5/23 DIN 12248 Ground cone B NS 14.5/23 DIN 12248 for y conical cock 4 DIN 12551 Connections for Connection for vacuum pump way conical coc
7、k N 12551 (Cock E) ANS 14.5/23 DIN 12248 Ground cone B NS 14.5/23 DIN 12248 Figure 1. Test arrangement Continued on pages 2 to 5 Explanations on page 5 Alleinverkauf der Normen durch Beuth Verlag GmbH. Berlin 30 und Kln 1 04.79 DIN 51 777 Teil 2 engl. Preisgr. 4 Vem.-Nr. O104 Page 2 DIN 51 777 Part
8、2 / 120 4, 1 r200 /-J Figure 2. Titration vessel DIN 12551 (Cock E) BNS 14.5/23 DIN 12248 Figure 3. Insert for titration vessel Figure 4. Platinum dual electrode 4.2 Vacuum method If at 120 OC further water separates off from the speci- men through chemical reaction in amounts which are not negligib
9、le, the specimen must be dewatered in a flow of very pure nitrogen at a residual pressure of 3 to 6.5mbar-2) and a temperature of 60 OC. The water condensed at -79 OC in dewatered methanol is titrated with KFL at a temperature above O OC by the end point method. The water content of the specimen is
10、calculated from the quantity of KFL used in the titration and from its titre. 5 Apparatus The test arrangement of the apparatus is shown schema- tically in Fig. 1 ; for details see Fig. 2 to 4. The apparatus consist essentially only of glass, Teflon or platinum. 5.1 End point titration apparatus Eit
11、her end pointconnected pH meters or current measuring end point titrimeters can be used. 5.2 Platinum dual electrode according to Fig. 4. 5.3 Titration vessel (see Fig. 2) for condensing the quantity of water to be analyzed in methanol and for titrating the water dissolved in methanol with continuou
12、s exclusion of air. 5.4 Wash bottle for drying the very pure nitrogen, capacity 200 ml, made of blown glass (to withstand rapid cooling to -79 OC). 5.5 Graduated wash bottle (see Fig. 1) for measuring the volume of the specimen and for flushing the speci- men with very pure nitrogen and thereby remo
13、ving water from it. 5.6 Burettes and storage bottles The burettes used are of the vacuum or plunger type graduated in 0.05 ml. The storage bottles for the KF L must be protected against the light. The standard solutions are stored in two onelitre bottles of brown or white glass with conical ground j
14、oint NS 29/32 according to DIN 12242. The vent tubes of the storage bottles and of the burettes are to be protected against the ambient atmosphere by means of drying tubes filled with a drying medium. N o te : KFL dissolves all lubricants comparatively quickly; Teflon joints have so far proved to be
15、 the best. 5.7 Heating bath with temperature control facility for the graduated wash bottle. 5.8 Gas pump or filter pump for the vacuum method. 0.76Q 1 .O1 325 2) 1mbar= Torr = 0.75006168 Torr For tables for the conversion from Torr into mbar and from mbar into Torr, see DIN 66038. DIN 51 777 Part 2
16、 Page 3 6 Reagents 6.1 Very pure nitrogen in steel cylinder The oxygen content should be less than 0.05 % by wt. The very pure nitrogen is used for driving the water out of the specimen and simultaneously for agitating the standard solution. The very pure nitrogen is best dried by washing with metha
17、nol in the wash bottle on the inlet side, the methane being cooled to -79OC; after this process the nitrogen contains negligibly small quantities of water (not more than 0.1 . 104 glh with a nitrogen flow of about 3 l/h at atmospheric pressure). If other media or methods are used for removing water
18、from the very pure nitrogen, their effectiveness must be demonstrated by measurement of the consistency of the dryness in the titration vessel during a period of time. 6.2 Karl Fischer solution (KFL) Titre for water contents up to 30 mg/kg: 1 mg/mI KFL Titre for water contents above 30 mg/kg : 5 mg/
19、mI KF L 6.3 Standard solution The standard solution is methanol. The methanol shall have a water content of less than 0.05 % by wt. It can have the water removed from it by fine fractionation or chemically by the following method: Carefully add to 1 litre of methanol 149 of metallic sodium followed
20、by 409 of methyl formate. Boil the mixture for 2 hours with a reflux condenser (bulb-type condenser) fitted until excess ester has completely decomposed into carbon monoxide and methanol, which point is indicated by cessation of carbon monixide evolution. With the cooling water emptied from the bulb
21、-type condenser and the latter in use as a distillation head, also with a further descending condenser in use, the liquid is now slowly distilled off.The first 100 ml are rejected, whilst the remaining distillate has a water content of about 0.01 % by wt. The whole apparatus must be protected agains
22、t moisture by a phosphorus pentoxide drying tube. It may be possible to make a considerable reduction in the stated quantities of sodium and methyl formate, depending on the water content of the original substance. 7 Sampling According to DIN 51 750 Part 1 and Part 2. The follow- ing points should b
23、e noted in particular: 7.1 Specimens from large drums must be rawn by using absolutely dry glass bottles with tightclosing ground stoppers. The ground stoppers shall not be withdrawn from the bottles until the latter arecompletely immersed in the medium to be tested. Transferring of specimens once d
24、rawn or combining to form a composite specimen are not permitted. N o t e : If metal dippers, e.g. cans, are used instead of glass bottles with ground stoppers, it should be noted that the screw closure normally used must be provided with a washer consisting of a piece of aluminium foil with plastic
25、s backing. 7.2 Once drawn specimens should be tested forthwith. If this is not practicable, the specimensmust be protected against ingress of atmospheric humidity. The specimen bottles must therefore be made airtight by application of a water-free sealing medium between the neck of the bottle and th
26、e ground stopper. 8 Preparation I 8.1 Blank titration of standard solution and finding the titration end point The stirring of the standard solution in the titration vessel is assured by the flow of very pure nitrogen through it. 3 to 5 ml of dewatered methanol are poured into the titration vessel a
27、nd titrated with KFL whilst at the same time a continuous flow of very pure dry nitrogen is maintained. A steady reading (during a period of about 15 minutes) on the indicating instrument proves the apparatus as a whole is dry. 8.2 Transferring the specimen The specimen - normally 250 ml - must be d
28、rawn from a larger quantity which should be about 1 litre. The test vessel (flask or can) is provided with a dry wash bottle insert. The graduated wash bottle may have traces of a water-free specimen adhering to its walls from a previous determination. It is sufficient to displace these traces from
29、the graduated wash bottle as fully as possible by flushing with dry very pure nitrogen through the riser tube. If no further KFL needs to be used during a period of 5 minutes while passing very pure nitrogen through, this is a sign that the apparatus is dry. The specimen is then transferred from the
30、 sampling vessel, preferably without interruption, into the graduated wash bottle, the transfer being effected under the pressure of the very pure nitrogen. The container should not be emptied beyond the quarter-full level; this is to prevent the .specimen from being affected by the very pure nitrog
31、en which has a relative moisture differing from that of the specimen. In the case of very viscous specimens it is advisable to warm the specimen in the specimen vessel. The simple volumetric determination of the specimen quantity is dispensed with if, for example, owing to a very high water content
32、in the specimen the use of about 250 ml corresponding to the calibrated volume of the graduated wash bottle would not be practicable owing to the likelihood of high KFL consumption. Therefore, with a water content of 500 . 10-6 g per g of specimen it is advisable not to put in 250 ml corre- sponding
33、 to about 2209 of specimen, but instead to introduce, for example, only 209 of specimen. The specimen is normally measured as follows: Sufficient of the specimen is first poured in to bring the level to 10 to 15mm below the graduation line on the graduated wash bottle, and the transference of the wa
34、ter vapour is then commenced by blowing very pure nitrogen through the specimen until the specimen has been raised by the heating bath to the latters temper- ature of 120 OC and has expanded accordingly. After 10 minutes the still outstanding quantity of specimen is added until the graduation line i
35、s reached. Deter- mination of the weight of the specimen used for analysis to the graduated wash bottle Page 4 DIN 51 777 Part 2 shall be based on the volume of the graduated wash bottle and the density of the specimen at 120 OC. The quantity of specimen remaining in the inlet tube of the graduated
36、wash bottle shall be disregarded, since for practical purposes this amount does not give up its water content to the very pure nitrogen. 9 Procedure 9.1 Method using atmospheric pressure If the temperature used is 120 OC and if the very pure nitrogen flows at a rate of about 3 Vh at atmospheric pres
37、sure, a period of 60 minutes will be needed for driving out more than 99 % by vol. of the water. After the water has been driven out the very pure nitrogen flow is left on for the purpose of stirring the liquid and the water taken up by the methanol is titrated by the addition of KFL dtop by drop. T
38、he titration end point should be identical with that observed when dewatering the methanol previously, Following this, it is advisable in all cases to determine what is known as the blank consumption, that is to say the change of titre with time, using the same very pure nitrogen flow as in the main
39、 test. In the case of used oils, the blank consumption, referred to the same time period as in the principal determination, may equal or even exceed the net water content of the specimen. The net content is found as the difference between the gross value and the blank consumption. If the blank consu
40、mption is higher than 10% of the value found in the main test, it is advisable to adopt the vacuum method. 9.2 Vacuum method When using a temperature of 60 “C and very pure nitrogen at a vacuum of 3 to 6.5mbar2) the dewatering period is 90 minutes. N o te : Owing to the water of reaction from the sp
41、ecimen, the resulting blank value obtained at this temperature is lower by a factor of 30 than at a dewatering temperature of 120 OC corresponding to the method according to Section 9. I, and is therefore negligibly small; in this case it does not need to be determined. The following should be noted
42、 when carrying out the determination: The specimen is put into the graduated wash bottle under vacuum. While this is beingdone, the cock E (see Fig. 1) must be closed far enough to ensure that bubbles and formation of foam by the specimen during degassing are minimized. When measuring the correct vo
43、lume of the specimen, care should be taken to ensure that the specimen inlet tube in the graduated wash bottle empties under the vacuum. Before the process of driving off the water starts, the flow of very pure nitrogen should be controlled - preferably by the use of a by-pass and a pinchcock -so th
44、at only 2 to 3 gas bubbles each of 50 pl occur per second. If the very pure nitrogen flow is appreciably greater than this there is a risk that the very pure nitrogen expanding in the graduated wash bottle will cause the specimen to foam over. N o t e : If the capillary shown in Fig. 1, which is 40
45、mm long and O2 mm in diameter is used, the nitrogen flow will be automatically rewlated to give the correct flow rate of the very pure nitrogen in vacuo. Cocks C or D (see Fig. 1) are provided between the outlets of the pipettes filled with methanol and KF L and the titration vessel and when evacuat
46、ing the vessel for vacuum distillation of the water these cocks must be closed to prevent KFL from evaporating in the inlet line through the vacuum and thereby driving uncontrolled quantities into the titration vessel. For the titration carried out at atmospheric pressure it is important that the re
47、-opened cock D is positioned vertically below the point of the pipette so that the KFL drops through its bore without any part of it adhering to the surface of the glass. The methanol in the titration vessel must be cooled to -79 OC (carbon dioxide snow) without fail, so as to ensure that evaporatio
48、n of the methanol is prevented and equally that complete dissolving of the greatly expanded water vapour takes place in the methanol. Before titration, the titration vessel must be warmed to above O OC. The end point titration is carried out according to Section 8.1. 1 O Evaluation The water content
49、 of the specimen in mg of water per kg of specimen is calculated by the following numerical equation: V *T 1000 mE Water content = Wherein: V T mE Initial weight of specimen in g Volume of KFL in ml used for titration of the specimen Titre of the KFL in mg of water per ml of KFL 11 Indication of the result The water content of the specimen is indicated in mg/kg, reference being made to this Standard. The number of places to which the result is given shall agree with the number of places with which the test error (see Section 12) is indicated. With regard to the rounding of the l
