1.2 . DETERMINATION OF SALINITY BY TITRATION (LOW PRECISION METHOD)

INTRODUCTION

For work in brackish waters and the surface waters of some coastal inlets an d for many biological purposes the analysis described in Part 1 .1 may be unnecessarily precise and time consuming . The following modification is recommended when a rapid method, correct to about 0 .05 to 0.1%0 of salinity, is adequate.

METHOD

A. CAPABILITIES

Range : Salinity 4-40%o

PRECISION AT THE 30yoo SALINITY LEVE L The correct value lies in the range :

Mean of n determinations -!-Q .Q6/ni %o

A duplicate titration on each sample is not necessary but should be made on about one sample in five . Duplicates are unacceptable if they differ by more than about 0 .06 ml of silver nitrate solution .

B . OUTLINE OF METHOD

The precipitable halide halogens in a 10-ml volume of sea water are determined by titration with a silver nitrate solution using a chromate end point, the Mohr titration . The silver solution is standardized against 10 ml of sea water of known chlorosity .

C . SPECIAL APPARATUS AND EQUIPMENT

An automatic 10-ml pipette and an automatic zero-adjusting 25-ml burette are required . An automatic dispenser of diluting solution, containing the chromate indicator, is advantageous . The titration is carried out in a 200-ml tall-form beaker, with magnetic stirring, against a white background in "yellow" artificial light from an ordinary electric light bulb . The pipette top should be lubricated with a little glycerol and the tap of the burette with a trace of paraffin (not silicone) stop-cock grease . The volumetric glassware must be kept spotlessly clean by soaking period-ically for a few minutes in a cold 5% solution of sodium hydroxide in methyl alcohol and then rinsing with nitric acid followed by distilled water .

D. SAMPLING PROCEDURE AND SAMPLE STORAGE . (Refer to Part I.1 )

E . SPECIAL REAGENTS REQUIRED 1 . STANDARD SEA WATER

Prepare a large volume (10-20 liters) of a standard consisting of filtered sea water (preferably collected below 50 m in the open ocean) with a chlorinity near to 18%o. This sample is stabilized by adding a few crystals of thymol and is then quickly put into sample bottles (see Part 1 .1) . Every tenth bottle is opened and analysed

17

18 A PRACTICAL HANDBOOK OF SEAWATER ANALYSIS

in duplicate by the high-precision method. The mean of ten or more such duplicates is taken as the chlorosity (20 C) of this standard sea water.

2. SILVER NITRATE SOLUTION (APPROXIMATELY 0.28N)

Dissolve 49 g of good quality silver nitrate for each liter of final solution. Use distilled water and store the solution in a dark bottle. The solution may need dilution vvith a little water to bring it to the desired strength, as described in Section H. The solution should be well mixed in the bottle once each day or prior to each batch of titrations.

3. INDICATOR-DILUENT SOLUTION

Dissolve 3.5 g of analytical quality potassium chromate, K 2Cr04, in each liter of distilled water.

F. EXPERIMENTAL

Add the sample of sea water to a 200-ml spoutless tall-form beaker by means of the automatic 10-ml pipette and then add 15 ml of indicator—diluent solution.

Titrate the solution from the 25-ml automatic burette. About 1 ml before it is judged that the end point is reached, rinse the sides of the beaker and stirrer with a little distilled water from a wash bottle.

As the end point approaches, the localized red precipitate formed by the silver solution will begin to spread throughout the solution. At the end point the pale greenish-yellow colour of the contents of the beaker changes to a full yellow and then becomes a definite pale red as the end point is exceeded. The exact point is largely subjective and errors in its estimation are allowed for to some extent in the standardization procedure (see Sect. H). Record the burette reading to the nearest 0.01 ml. Note the average temperature of the samples Tc, for each titration period of 1-2 hr (say every 20 or 30 samples) by a thermometer placed in a tightly packed case of bottles that has already equilibrated with the laboratory temperature. Note the average temperature TAg of the silver nitrate solution during the same period shown on a thermometer placed in a small tube through which the solution flows before entering the burette. If possible, TAg should be kept greater than T01 and the difference between the two should not exceed 5 C.

The sequence of operations described in Part 1.1 can be used in the present method when the maximum speed is required for routine analyses.

G. CALCULATIONS

Let V be the reading of the burette, taken to two decimal places.

Let C„ be the burette correction (positive or negative). This depends on the magnitude of the titration and is given by the manufacturers or should be determined

(see standard text books on analysis). It should rarely exceed -±0.05 ml.

Let CB be the standardization correction, obtained as described in Section H.

This correction should not exceed 0.15 ml.

Let C, be the temperature correction which depends upon the reading V and the difference in-the temperature (TAg — T01 ) between the silver nitrate solution and the sample. The correction C t may be neglected unless (Tag — T01 ) exceeds 3 C

1.2. SALINITY (LOW PRECISION) 19 and V is greater than 15 ml. Should some measure of the temperature correction be thought desirable read C, from Table I. Calculate the chlorosity at 20 C from the expression:

Cl/liter(20) = V + Cb + C8 + C'

Evaluate the salinity from this chlorosity value from Table II which shows the relation between Cl/liter(20) and S%a.

H. CALIBRATION

1. Adjust the strength of the silver nitrate solution by adding small amounts of water to the solution prepared as in Section E, until the value of

Vc=V+Cb+Ct

is within 0.1 or less of the chlorosity of the standard at 20 C ( expressed in milliliters as Va).

2. Evaluate Vc as the mean of five determinations, having a spread less than 0.06 ml, each day before commencing a batch of analysis.

3. The standardization correction, C8, for any determination will depend upon the value of ( Va - V,,), obtained above, and the magnitude of V for the particular determination. Calculate correction intervals, v, given by the formula:

„- 200 (VQ, - V,) + 1

The experimental values of V are classified according to the number of multiples of „ that they contain and the C, values are read from a table constructed by each worker, of the form :

V C.

0.00 to U 0.000

v to 2u (V -V^)U/V

2v to 3v 2(Va - V^) v/ V^

3v to 4v 3(Va - V^)v / V^

Values of C, should be calculated to the nearest 0.01 ml.

1.3 . DETERMINATION OF DISSOLVED OXYGEN INTRODUCTION

The method described here is a modification of the classical Winkler procedur e which we are convinced still remains the most reliable and precise means of analysing for dissolved oxygen in sea water . If a very precise estimate of extremely low oxygen concentrations is required on a routine scale then an absorptiometric determination of the iodine liberated in the Winkler method may prove superior to a titration procedure ( Oulman and Baumann, Sewage Ind . Wastes, 28 : 1461, 1956) but not

unless oxygen concentrations are less than about 0 .1 mg-at/liter .

The thiosulphate titration of iodine may be completed using a variety of electro-chemical end-point detectors ( a good account is given by Knowles and Lowden, Analyst, 78 : 159, 1953) but we have found that none of these methods gives a noticeably better precision than does the classical starch end point when used with proper illumination . The true stoichiometric end point is a little after the starch end point but this error is barely significant in marine work.

The accuracy of the Winkler method has recently been examined in detail by Carpenter. His findings ( Limnol. Oceanog ., 10: 135, 1964) and a description of the results of a comparative experiment car ried out in the USA and Canada (Carritt and Carpenter, J . Marine Res., 24 : 286, 1966) fully evaluate this approach and should be read by the analyst desirous of attaining the maximum possible accuracy . The following procedure is little different from the optimum technique suggested by Carpenter .

A. CAPABILITIES

METHO D

Range: 0.005-8 mg-at/liter PRECISION AT THE 0 .7 MG-AT/LITER LEVE L The correct value lies in the range :

Mean of n determinations -!-0 .003/nt mg-at/liter .

This is the highest precision considered likely for work in a shore-base labora-tory under near ideal conditions, using thiosulphate standardized by the mean of at least five titrations . Under routine conditions at sea the uncertainty range will be appreciably greater, perhaps nearly doubled . Comparison of the starch end point described here with an electrometric end point shows that there is a slight negative error which is not allowed for in the standardization when low oxygen concentrations are being estimated . Oxygen concentrations below 0 .1 mg-at/liter or less will be up to 0 .0015 mg-at/liter too low. This amount is scarcely significant .

B . OUTLINE OF METHOD

A divalent manganese solution, followed by strong alkali, is added to the sample . The precipitated manganous hydroxide is dispersed evenly throughout the seawater sample which completely fills a stoppered glass bottle . Any dissolved oxygen rapidly oxidizes an equivalent amount of divalent manganese to basi c

21

22 A PRACTICAL HANDBOOK OF SEAWATER ANALYSIS

hydroxides of

higher valency states. When

the

solution is acidified

in the

presence

of

iodide

the

oxidized manganese again reverts to

the divalent

state

and

iodine, equiva-lent to

the original

dissolved oxygen

content of the water,

is liberated.

This

iodine is

titrated

with standardized

thiosulphate solution.

C. SPECIAL APPARATUS

AND

EQUIPMENT

300^-ml BOD (biological oxygen

demand)

bottles.

50-ml pipette and a 10-ml burette

graduated

in

units

of 0.05 ml or less. The burette

should

have an

automatic

zero

adjuster

and an

automatic "Lowy"

type of pipette

is recommended.

(This type

has

a

two-way tap at

the top of the pipette.)

Alternatively

titrations

can be conveniently carried

out

using

a 1.00-ml microburette.

The Gilmont Ultramicroburette made

by

the Manostat Corporation of 26 N Moore Street, New York 13,

N.Y.,

USA,

is suitable

and

can easily be read to

0.001

mi.

125-ml

conical flasks painted

white

on the outside over

the base and

two thirds

of the

way around

the

sides.

The

flasks should be illuminated through

the

unpainted

portion of the

side during titration by means

of an

ordinary

filament

electric light

bulb. The contents of the

flask

are

best stirred during titration by

a white

magnetic stirring bar.

D. SAMPLING PROCEDURE AND SAMPLING STORAGE

BOD

bottles

are

rinsed twice with

the

sample being analysed.

If the

sample is obtained from

a

reversing bottle

a

length

of

rubber tubing should be taken from

the

tap to

the

bottom

of the BOD

bottle

and

sea water introduced

in

such

a

way

as

to minimize

turbulence and agitation of the

sample.

The end of the

rubber

tube must

always be kept beneath

the surface of the

water

as the

bottle is filled. Water is allowed to overflow from

the top of the BOD

bottle which is stoppered at

once. A

volume of water at least equal to the volume of the BOD bottle should be allowed

to overflow, more if the oxygen content of the water is suspected to be very low.

Im Dokument A PRACTICAL HANDBOOK OF SEAWATER ANALYSIS (Seite 29-34)