Vous êtes sur la page 1sur 10

1050 EXPRESSION OF RESULTS*

1050 A. Units

Standard Methods uses the International System of Units (SI). 1,2 Concentration units for chemical and physical results are generally expressed in mass units/L. The numerical values range from 0.1 to 999.9 mg/L. If the values are less, express them as micrograms per liter ( g/L) and if greater as grams per liter (g/L), with the numerical values adjusted accordingly. Record and report all analytical results to the proper number of signif- icant figures (see 1050B).

1. Radioactivity

For information on reporting radiological results, see Section

7020D.

2. Mass

a. Mass concentrations: Mass concentrations can be expressed in terms of weight/volume (w/v), weight/weight (w/w), and volume/volume (v/v). 3 Those commonly used in Standard Meth- ods are listed in Table 1050:I. Most analysis results are reported in terms of w/v, but some may be expressed in terms of w/w. Since analyses usually are performed for analytes in solution it may be necessary to report results in reference terms other than solution volume, such as milligrams per kilogram, parts per million, or percent by weight. These terms are computed as follows:

mg/kg mg/L

ppm by weight mg/L

% by weight

mg/L

10 000

where is the density of the sample solution or solid mixture being measured, in kilograms per liter (often reported as the same numeric value as grams per milliliter). The relative density of water can be computed from Table 1050:II 4 using the densities at t°C and 4°C. Weight per unit weight (w/w) is defined as mass of analyte per total (wet) mass of solution or mixture. In some cases, the weight fraction results may be relative to total dry mass rather than total wet mass. The units would be multiples of (g analyte/kg dry). This is referred to as “moisture-free basis”; report concentration of total solids at the same time.

* Reviewed by Standard Methods Committee, 2006. Joint Task Group: 22nd Edition—L. Malcolm Baker (chair), Stephen W. Johnson, David F. Parkhurst.

TABLE 1050:I. COMMONLY USED EXPRESSIONS of MASS CONCENTRATION

Unit of Expression

Proportion

Value

w/v

w/w

v/v

Parts per hundred (%) Parts per thousand ( 0 / 00 ) Parts per million (ppm) Parts per billion (ppb)

10

2

g/dL

g/100 g

mL/dL

10

3

g/L

g/kg

mL/L

10

6

mg/L

mg/kg

L/L

10 9

g/L

g/kg

nL/L

Adapted from: MILLS, I., T. C VITAS, K. H OMANN, N. KALLAY & K. K UCHITSU. 1993. Quantities, Units and Symbols in Physical Chemistry, 2nd ed. Blackwell Scientific Publications, Oxford, U.K.

b. Other concentration units: To obtain mass concentrations,

the analyst may have to compute intermediate results in terms of mass, moles, volume, or equivalent. The most commonly used intermediate forms are defined as follows:

1) Mass fraction—analyte mass divided by the total mass of the solution or mixture, expressed as kg/kg; 2) Volume fraction—analyte volume divided by total volume of sample where measurements are at equal pressure and temperature, expressed as L/L; 3) Mole fraction—number of moles of analyte per total moles in solution; 4) Molar concentration ( molarity)—number of moles of ana- lyte contained in 1 L solution, designated by M ; 5) Molal concentration (molality)—number of moles of ana- lyte dissolved in 1 kg solvent; 6) Normality—number of equivalents (see ¶ c below) of ana- lyte dissolved and diluted to a 1-L volume, designated by

N . Also see ¶ d below.

c. Equivalency: The unit milligram-equivalents per liter, or

millequivalents per liter (me/L), can be valuable for making analytical and water-treatment calculations and for checking analyses by anion– cation balance. Table 1050:III presents factors for converting concentrations of common ions from milligrams per liter to milliequivalents per liter, and vice versa. The term “milliequivalent” represents 0.001 of an equivalent weight, which is defined as the weight of the ion (sum of the atomic weights of the atoms making up the ion) divided by the number of charges normally associated with the particular ion. The factors for converting results from milligrams per liter to mil- liequivalents per liter were computed by dividing the ion charge by weight of the ion. Conversely, factors for converting results from milliequivalents per liter to milligrams per liter were calculated by dividing the weight of the ion by the ion charge. To define equivalents, 5 chemical reactions must be broken down into several types. 1) Neutralization reactions 6 —The equivalent weight of a sub- stance in a neutralization reaction is the weight of acid that will furnish, react with, or be chemically equivalent to, one mole of hydrogen ion in the reaction in which it occurs. For example, the equivalent weight of an acid is its molar mass (molecular weight)

EXPRESSION OF RESULTS (1050)/Units

EXPRESSION OF RESULTS (1050)/Units

T ABLE 1050:II. DENSITY OF WATER FREE FROM DISSOLVED ATMOSPHERIC GASES , AT A PRESSURE OF 101.325 PA

Density of Water at Given Temperature kg/m 3

Temperature, t 68 * °C

 

.0

.1

.2

.3

.4

.5

.6

.7

.8

.9

0

999.8426

8493

8558

8622

8683

8743

8801

8857

8912

8964

1

999.9015

9065

9112

9158

9202

9244

9284

9323

9360

9395

2

999.9429

9461

9491

9519

9546

9571

9595

9616

9636

9655

3

999.9672

9687

9700

9712

9722

9731

9738

9743

9747

9749

4

999.9750

9748

9746

9742

9736

9728

9719

9709

9696

9683

5

999.9668

9651

9632

9612

9591

9568

9544

9518

9490

9461

6

999.9430

9398

9365

9330

9293

9255

9216

9175

9132

9088

7

999.9043

8996

8948

8898

8847

8794

8740

8684

8627

8569

8

999.8509

8448

8385

8321

8256

8189

8121

8051

7980

7908

9

999.7834

7759

7682

7604

7525

7444

7362

7279

7194

7108

10

999.7021

6932

6842

6751

6658

6564

6468

6372

6274

6174

11

999.6074

5972

5869

5764

5658

5551

5443

5333

5222

5110

12

999.4996

4882

4766

4648

4530

4410

4289

4167

4043

3918

13

999.3792

3665

3536

3407

3276

3143

3010

2875

2740

2602

14

999.2464

2325

2184

2042

1899

1755

1609

1463

1315

1166

15

999.1016

0864

0712

0558

0403

0247

0090

9932†

9772†

9612†

16

998.9450

9287

9123

8957

8791

8623

8455

8285

8114

7942

17

998.7769

7595

7419

7243

7065

6886

6706

6525

6343

6160

18

998.5976

5790

5604

5416

5228

5038

4847

4655

4462

4268

19

998.4073

3877

3680

3481

3282

3081

2880

2677

2474

2269

20

998.2063

1856

1649

1440

1230

1019

0807

0594

0380

0164

21

997.9948

9731

9513

9294

9073

8852

8630

8406

8182

7957

22

997.7730

7503

7275

7045

6815

6584

6351

6118

5883

5648

23

997.5412

5174

4936

4697

4456

4215

3973

3730

3485

3240

24

997.2994

2747

2499

2250

2000

1749

1497

1244

0990

0735

25

997.0480

0223

9965†

9707†

9447†

9186†

8925†

8663†

8399†

8135†

26

996.7870

7604

7337

7069

6800

6530

6259

5987

5714

5441

27

996.5166

4891

4615

4337

4059

3780

3500

3219

2938

2655

28

996.2371

2087

1801

1515

1228

0940

0651

0361

0070

9778*

29

995.9486

9192

8898

8603

8306

8009

7712

7413

7113

6813

30

995.6511

6209

5906

5602

5297

4991

4685

4377

4069

3760

31

995.3450

3139

2827

2514

2201

1887

1572

1255

0939

0621

32

995.0302

9983†

9663†

9342†

9020†

8697†

8373†

8049†

7724†

7397†

33

994.7071

6743

6414

6085

5755

5423

5092

4759

4425

4091

34

994.3756

3420

3083

2745

2407

2068

1728

1387

1045

0703

35

994.0359

0015

9671†

9325†

8978†

8631†

8283†

7934†

7585†

7234†

36

993.6883

6531

6178

5825

5470

5115

4759

4403

4045

3687

37

993.3328

2968

2607

2246

1884

1521

1157

0793

0428

0062

38

992.9695

9328

8960

8591

8221

7850

7479

7107

6735

6361

39

992.5987

5612

5236

4860

4483

4105

3726

3347

2966

2586

40

992.2204

* t 68 represents temperature, according to International Practical Temperature Scale 1968. † The leading figure decreases by 1.0. Source: M ARSH, K.N., ed. 1987. Recommended Reference Materials for the Realization of Physicochemical Properties, Blackwell Scientific Publications, Oxford, U.K.

divided by the number of hydrogen atoms (equivalent number) contained in the acid. Thus, for acids that have reacted com- pletely, the equivalent weight of HCl and HNO 3 would equal the molar mass; the equivalent weight of H 2 SO 4 and H 2 CO 3 would equal the molar mass divided by 2 and the equivalent weight for H 3 PO 4 would equal molar mass divided by 3. If the chemical formula is written out for the reaction of an acid with a base, the equivalent amounts for each compound in the reaction are obtained by dividing each stoichiometric coef- ficient in the equation by the equivalent number. 7 For example, in the equation:

H 2 SO 4 2NaOH 3 Na 2 SO 4 2H 2 O

dividing by the equivalent number, 2, yields:

H 2 SO 4 2NaOH

2

2

Na 2 SO 4

2

2H 2 O

2

and reducing to lowest terms yields:

H 2 SO 4 NaOH

2

1

Na 2 SO 4 H 2 O

2

1

EXPRESSION OF RESULTS (1050)/Units

EXPRESSION OF RESULTS (1050)/Units

TABLE 1050:III. CONVERSION FACTORS *

(Milligrams per Liter – Milliequivalents per Liter)

Ion

Ion

(Cation)

me/L mg/L

mg/L me/L

(Anion)

me/L mg/L

mg/L me/L

Al 3

0.111 2

8.994

BO 2 Br Cl

0.023 36

42.81

B 3

0.277 5

3.604

0.012 52

79.90

Ba 2 Ca 2 Cr 3

0.014 56

68.66

0.028 21

35.45

0.049 90

20.04

CO 3 2

0.033 33

30.00

0.057 70

17.33

CrO 4 2 F HCO 3 HPO 4 2

0.017 24

58.00

 

0.052 64

19.00

Cu 2 Fe 2 Fe 3 H K

0.031 47

31.77

0.016 39

61.02

0.035 81

27.92

0.020 84

47.99

0.053 72

18.62

H 2 PO 4 HS HSO 3 HSO 4 I NO 2

0.010 31

96.99

0.992 1

1.008

0.030 24

33.07

0.025 58

39.10

0.012 33

81.07

 

0.010 30

97.07

Li Mg 2 Mn 2 Mn 4 Na NH 4 Pb 2 Sr 2 Zn 2

0.144 1

6.941

0.007 880

126.9

0.082 29

12.15

0.021 74

46.01

0.036 40

27.47

NO

3

0.016 13

62.00

0.072 81

13.73

OH

0.058 80

17.01

0.043 50

22.99

PO 4 3 S 2

0.031 59

31.66

0.055 44

18.04

0.062 37

16.03

0.009 653

103.6

3 2

SiO SO 3 2

0.026 29

38.04

0.022 83

43.81

0.024 98

40.03

0.030 59

32.70

SO 4 2

0.020 82

48.03

* Factors are based on ion charge and not on redox reactions that may be possible for certain of these ions. Cations and anions are listed separately in alphabetical order.

Thus, one-half formula weight of H 2 SO 4 and Na 2 SO 4 , and one formula weight of NaOH and H 2 O are equivalent weights in this case. In some cases, the reaction may not go to completion and the following equivalent entities are obtained.

H 2 CO 3 NaOH 3 NaHCO 3 H 2 O

Because only one hydrogen was used, the equivalent number is 1.

H

2 CO 3

NaOH

NaHCO 3

H

2 O

1

1

1

1

If the cation of the base is multivalent, the number of equiv- alents is calculated similarly. For example:

3HCl Al(OH) 3 3 AlCl 3 3H 2 O

HCl

Al(OH) 3

AlCl 3

H

2 O

1

3

3

1

Thus, the equivalent number is equal to the number of hy- droxyl groups in a compound. 2) Precipitation and complex reactions—In these reactions the equivalents are equal to the weight of substance that will furnish, react with, or be chemically equivalent to one mole of univalent cation in the precipitate or complex formed. An example of a precipitation reaction is:

BaCl 2 Na 2 SO 4 3 BaSO 4 2NaCl

BaCl 2

2

Na 2 SO 4

2

BaSO 4

NaCl

2

1

Examples of reactions with complexes are:

AgNO 3 2KCN 3 Ag(CN) KNO 3 K

2

AgNO 3 2KCN

Ag(CN)

2

KNO 3 K

1

1

1

1

1

2AgNO 3 2KCN 3 2AgCN 2KNO 3

2AgNO 3

2KCN

2AgCN

2KNO 3

 

2

2

2

2

AgNO 3

KCN

AgCN

KNO 3

1

1

1

1

NiSO 4 4KCN 3 Ni(CN) 2 4 K 2 SO 4 2K

NiSO 4

4KCN

2

Ni(CN) 4

K

2 SO 4

2K

2

2

2

2

2

NiSO 4

2KCN

2

Ni(CN) 4

K

2 SO 4

K

2

1

2

2

1

3) Oxidation–reduction reactions—In these reactions the equivalent weight of a substance is the weight that will furnish, react with, or be chemically equivalent to one elec- tron transferred. First, determine the number of electrons transferred in a redox reaction by writing out a balanced equation and the reduction and oxidation half-reactions. For example:

EXPRESSION OF RESULTS (1050)/Units

EXPRESSION OF RESULTS (1050)/Units

5Na 2 C 2 O 4 2KMnO 4 8H 2 SO 4 3 5Na 2 SO 4 2MnSO 4

z 1 V 1 M 1 z 2 V 2 M 2

K 2 SO 4 10CO 2 8H 2 O

5Na 2 C 2 O 4 5H 2 SO 4 3 5Na 2 SO 4 10CO 2 10H 10 e

2KMnO 4 3H 2 SO 4 10H 10 e 3 2MnSO 4 K 2 SO 4 8H 2 O

Then divide the stoichiometric equation coefficients of the complete balanced equation by the number of electrons trans- ferred in the half-reaction equations (in this case, 10) and reduce the fractions to yield:

where:

z equivalent number and

M molarity of substance in solution.

This relationship can be used to compute molarities or vol- umes of reagents being compared with one another. For exam- ple, the above equation could be reduced to the following for potassium permanganate (1) and sodium oxalate (2).

Na 2 C 2 O 4 KMnO 4 4H 2 SO 4 Na 2 SO 4

2

5

5

2

MnSO 4

K

2 SO 4

CO 2 4H 2 O

5

10

1

5 3. p Functions

5 V 1 M 1 2 V 2 M 2

With these quantities, only one electron is transferred, and thus the amounts represent the equivalents. In this example the equivalent weights of these compounds in the reaction can be computed as follows:

Equivalent weight molecular weight g/mole equivalent number

Concentrations can be reported in the form of p functions (e.g., pH). This form is used for expressing values when the analyte concentration varies over orders of magnitude. The pX concept is defined 8,9 in terms of analyte activity, rather than concentra- tion, as follows:

Equivalent weight KMnO 4 KMnO 4

5

158.04 g/mole

5 where:

pX log (a x )

31.61 g/equivalent

Equivalent weight Na 2 C 2 O 4 Na 2 C 2 O 4

2

134.00 g/mole

2

a x activity of analyte X in solution.

Activity coefficients relate activity to concentration:

67.00 g/equivalent

pX log (c x x /c o )

d. Normality: Normality, as defined in ¶ b 6) above, may be calculated as follows:

Normality

weight of substance

(equivalent weight of substance)(liters of solution)

The normality concept allows substances to be compared with one another stoichiometrically through the relationship.

where:

V 1 N 1 V 2 N 2

V

volume, mL or L,

N

normality, and

1, 2 compound 1 or 2.

If one knows the values of three of the variables, then the fourth can be computed. There is a trend to move from the normality 6 concept because of possible ambiguities in determining normalities. These ambi- guities arise because a substance being compared might have more than one computed normality concentration (e.g., potas- sium cyanide when it reacts with silver ion) and still be in the same concentration. The above equation relating volume and normality of one substance to those of another can be written using the following equation in terms of molarity:

where:

c x molar concentration of x , mole/L x activity coefficient, dimensionless, and

molar concentration at standard state, mole/L.

With the development of electrode measurement technol- ogy, pH and other p functions have become more commonly used, especially for very low concentrations of analytes. However, the electrode measures only activity, a x , and not concentration, c x , directly. The Debye-Hu¨ckel equation provides a way to estimate the activity coefficients in low-ionic-strength solutions of 0.001M or less. This equation, 4,10 12 which relates concentration to activity coefficient, is:

where:

log i A z i 2 I 1/2

1 B I 1/2

I analyte ion activity coefficient, z i charge on ion species, I ionic strength of all the ions in solution 1 2 c i z i 2 , mole/L, radius of ionic atmosphere, picometers (see Table 1050:IV for values), and A, B Debye-Hu¨ckel equation constants (see Table 1050:V for values given by temperature and molar or molal concentration).

EXPRESSION OF RESULTS (1050)/Units

EXPRESSION OF RESULTS (1050)/Units

TABLE 1050:IV. EFFECTIVE HYDRATED RADIUS FOR COMMON IONS

Type and Charge of Ion

Ion

Ion Size, pm

Inorganic, 1

H

Li

Na , CdCl , ClO 2 , IO 3 , HCO 3 , H 2 PO 4 , HSO 3 , H 2 AsO 4 , Co(NH 3 ) 4 (NO 2 ) 2

OH , F , SCN , OCN , HS , ClO 3 , ClO 4 , BrO 3 , IO 4 , MnO 4 K , Cl , Br , I , CN , NO 2 , NO 3 Rb , Cs , NH 4 , Tl , Ag

Inorganic, 2

Mg 2 , Be 2

Ca 2 , Cu 2 , Zn 2 , Sn 2 , Mn 2 , Fe 2 , Ni 2 , Co 2 Sr 2 , Ba 2 , Cd 2 , Hg 2 , S 2 , S 2 O 4 2 , WO 4 2 Pb 2 , CO 3 2 , SO 3 2 , MoO 4 2 , Co(NH 3 ) 5 Cl 2 , Fe(CN) 5 NO 2

Hg 2 2 , SO 4 2 , S 2 O 3 2 , S 2 O 6 2 , S 2 O 8 2 , SeO 4 2 , CrO 4 2 , HPO 4 2

Inorganic, 3

Al 3 , Fe 3 , Cr 3 , Sc 3 , Y 3 , In 3 , lanthanides*

PO 4 3 , Fe(CN) 6 3 , Cr(NH 3 ) 6 3 , Co(NH 3 ) 6 3 , Co(NH 3 ) 5 H 2 O 3

Inorganic, 4

Th 4 , Zr 4 , Ce 4 , Sn 4 Fe(CN) 6 4

Organic, 1

HCOO , H 2 citrate , CH 3 NH 3 , (CH 3 ) 2 NH 2 NH 3 CH 2 COOH, (CH 3 ) 3 NH , C 2 H 5 NH 3 CH 3 COO , CH 2 ClCOO , (CH 3 ) 4 N , (C 2 H 5 ) 2 NH 2 , NH 2 CH 2 COO CHCl 2 COO , CCl 3 COO , (C 2 H 5 ) 3 NH , (C 3 H 7 )NH 3 C 6 H 5 COO , C 6 H 4 OHCOO , C 6 H 4 ClCOO , C 6 H 5 CH 2 COO , CH 2 CHCH 2 COO , (CH 3 ) 2 CHCHCOO , (C 2 H 5 ) 4 N , (C 3 H 7 ) 2 NH 2 [OC 6 H 2 (NO 3 ) 3 ] , (C 3 H 7 ) 3 NH , CH 3 OC 6 H 4 COO (C 6 H 5 ) 2 CHCOO , (C 3 H 7 ) 4 N

Organic, 2

(COO) 2 2 , Hcitrate 2 H 2 C(COO) 2 2 , (CH 2 COO) 2 2 , (CHOHCOO) 2 2 C 6 H 4 (COO) 2 2 , H 2 C(CH 2 COO) 2 2 , CH 2 CH 2 COO 2 2 [OOC(CH 2 ) 5 COO] 2 , [OOC(CH 2 ) 6 COO] 2 , Congo red anion 2

Organic, 3

Citrate 3

900

800

700

600

500

450

400

350

300

250

800

700

600

500

450

400

900

500

400

1100

500

350

400

450

500

600

700

800

450

500

600

700

500

* Elements 57–71 in the periodic table. S OURCE: Data from K IELLAND, J. 1937. Individual activity coefficients of ions in aqueous solutions. J. Amer. Chem. Soc. , 59:1675; table in H ARRIS, D.C. 1982. Quantitative Chemical Analysis. W.H. Freeman & Co., New York, N.Y.

Example:

Compute the hydrogen ion concentration of a solution of pure water with sufficient potassium chloride to make it 0.1M. The measured pH is 6.98 at 25°C. The ionic strength of the solution is

I 1/2 c i z i

2

I 1/2 [(c K )(1) 2 (c Cl )(1) 2 (c H )(1) 2 (c OH )(1) 2 ]

The hydrogen ion and the hydroxyl ion concentrations (estimated to be about 10 -7 mole/L) are insignificant in this example. Thus,

I 1 2 [(0.1)(1) 2 (0.1)(1) 2 ] 0.1 mole/L

Using the computed I and the values from Tables 1050:IV and 1050:V for (900 pm), A(0.5092), and B(0.003286), the Debye- Hu¨ckel equation yields

Since

log H

(0.5092)(1) (0.1)

2

0.5

1 (0.003286)(900)(0.1) 0.5 0.8321

H 0.8256

EXPRESSION OF RESULTS (1050)/Units

EXPRESSION OF RESULTS (1050)/Units

TABLE 1050:V. VALUES OF A AND B FROM 0 TO 100°C FOR DEBYE -H UCKEL EQUATION 12

¨

 

A (Abs. Units) in Terms of Unit Volume of Solution

 

B

(cm 1 10 8 )

B

(cm 1 10 8 )

A (Abs. Units) in Terms of Unit Weight of Solvent

in

Terms of Unit Volume of Solution

in Terms of Unit Weight of Solvent

Temperature

°C

0

0.4883

0.4883

0.3241

0.3241

5

0.4921

0.4921

0.3249

0.3249

10

0.4961

0.4960

0.3258

0.3258

15

0.5002

0.5000

0.3267

0.3266

18

0.5028

0.5025

0.3273

0.3271

20

0.5046

0.5042

0.3276

0.3273

25

0.5092

0.5085

0.3286

0.3281

30

0.5141

0.5130

0.3297

0.3290

35

0.5190

0.5175

0.3307

0.3297

40

0.5241

0.5221

0.3318

0.3305

45

0.5296

0.5270

0.3330

0.3314

50

0.5351

0.5319

0.3341

0.3321

55

0.5410

0.5371

0.3353

0.3329

60

0.5471

0.5425

0.3366

0.3338

65

0.5534

0.5480

0.3379

0.3346

70

0.5599

0.5537

0.3392

0.3354

75

0.5668

0.5596

0.3406

0.3363

80

0.5739

0.5658

0.3420

0.3372

85

0.5814

0.5722

0.3434

0.3380

90

0.5891

0.5788

0.3450

0.3390

95

0.5972

0.5857

0.3466

0.3399

100

0.6056

0.5929

0.3482

0.3409

S OURCE: M ANOV, G.C., R.G. B ATES, W.J. H AMER & S.F. A CRES. 1943. Values of the constants in the Debye-Hu¨ckel equation for activity coefficients. J. Amer. Chem. Soc.

65:1765s.

pH –log a H

at the measured pH of 6.98,

a H 10 6.98 1.047 10 7 mole/L

Then,

c H (a H )/ ( H )

c H 1.047 10 7 /0.8256 1.268 10 7 mole/L

Note that the hydrogen ion concentration is about 21% greater than was indicated by the hydrogen activity given by the pH electrode. The same type of computation can be made for determining the ionic concentration from the measured electrode activity for fluoride, silver, and other substances. This is important if the true concentration of the ion is to be determined.

4. Stoichiometric Factors

Some analyses are reported as concentrations of other sub- stances. Conversion factors to accomplish this are called stoi- chiometric factors. Examples are: alkalinity or acidity reported as CaCO 3 , total hardness as CaCO 3 , magnesium hardness as CaCO 3 , ammonia as nitrogen, nitrate as nitrogen, nitrite as nitrogen, and phosphate as phosphorus. These stoichiometric factors provide for taking the analysis data in the form analyzed first, dividing the mass concentration

by the analyte’s formula weight, and multiplying by the formula weight of the reported substance desired. For example,

mg/L chloride as NaCl mg Cl

L

formula weight of sodium chloride

formula weight of chlorine

Be sure that the formula weights of the substance to be reported and of the analyte include the same number of atoms of the common element. Some analyses involve multiple reactions. For example, listed below are the reactions involved in the Winkler determination. 13

2MnSO 4 4KOH 3 2Mn(OH) 2 2K 2 SO 4

2

Mn(OH) 2 O 2 3 2MnO(OH) 2

2

MnO(OH) 2 4KI 4H 2 SO 4 3 2I 2 2MnSO 4 2K 2 SO 4 6H 2 O

2I 2 4Na 2 S 2 O 3 3 4NaI 2Na 2 S 4 O 6

In the last equation, one mole of sodium thiosulfate is equiv- alent to one mole of elemental iodine. The number of moles of elemental iodine is equivalent to half the moles of manganese oxide hydroxide [MnO(OH) 2 ] in the second and third equations. Then the number of moles of oxygen is half the number of moles of MnO(OH) 2 , and thus the equivalent weight of elemental oxygen is one-fourth that of elemental iodine or sodium thiosul- fate (32 g/4 8g O 2 /equivalent weight). This type of reasoning must be used to determine the analytical relationship between the titrant and the analyte being sought.

EXPRESSION OF RESULTS (1050)/Significant Figures

EXPRESSION OF RESULTS (1050)/Significant Figures

5. Molality

It is sometimes convenient to relate molality to molarity. That relationship is as follows: 14

where:

M

M molarity, mole/L,

1000 m

(1000 W B m )

density of solution, g/cm 3 ,

m molality, mole/kg, and

W B molecular mass of solute, g.

The molality in terms of molarity is given as

m

1000M

1000 MW B

Many electrochemical measurements are made in terms of molality rather than molarity.

6. References

1. AMERICAN SOCIETY FOR TESTING & MATERIALS . 1993. Standard Prac-

tice for Use of the International System of Units (SI) (The Mod- ernized Metric System); E380-93. Philadelphia, Pa.

2.

B UREAU INTERNATIONAL DES POIDS ET MEASURES . 1991. The Interna-

tional System of Units (SI), 6th ed. Sevres, France.

3.

MILLS , I., T. CVITAS , K. HOMANN , N. KALLEY & K. KUCHITSU . 1993. Quantities, Units and Symbols in Physical Chemistry, 2nd ed. Blackwell Scientific Publications, Oxford, U.K.

4.

MARSH , K.M., ed. 1987. Recommended Reference Materials for the Realization of Physicochemical Properties, Blackwell Scientific Publications, Oxford, U.K.

5.

C VITAS , T. & I. M ILLS . 1994. Replacing gram equivalents and normalities. Chem. Internat. 16:123.

6.

AYRES , G.H. 1958. Quantitative Chemical Analysis. Harper & Row Publishers, New York, N.Y.

7.

KENNER , C.T. & K.W. BUSCH . 1979. Quantitative Analysis. Harper

&

Row Publishers, New York, N.Y.

8.

F REISER , H. & G.H. NANCOLLAS . 1987. Compendium of Analytical

Nomenclature, Definitive Rules 1987. Blackwell Scientific Pub-

lications, London, U.K.

9.

G OLD , V., K.L. LOENING , A.D. MC NAUGHT & P. SEHMI . 1987. Com- pendium of Chemical Technology IUPAC Recommendations. Blackwell Scientific Publications, Oxford, U.K.

10.

HARRIS , D.C. 1982. Quantitative Chemical Analysis. W.H. Freeman

&

Co., New York, N.Y.

11.

SKOOG , D.A. & D.M. WEST . 1976. Fundamentals of Analytical Chemistry, 3rd ed. Holt, Rinehart & Winston, New York, N.Y.

12.

M ANOV , G.C., R.G. B ATES , W.J. H AMER & S.F. A CRES . 1943. Values of the constants in the Debye-Hu¨ckel equation for activity coeffi- cients. J. Amer. Chem. Soc. 65:1765.

13.

HACH COMPANY . 1984. Chemical Procedures Explained. Loveland, Colo.

14.

WEAST , R.C. & D.R. LIDE . 1989. Handbook of Chemistry and Physics, 70th ed. CRC Press, Inc., Boca Raton, Fla.

1050 B. Significant Figures

1. Reporting Requirements

To avoid ambiguity in reporting results or in presenting direc- tions for a procedure, it is customary to use “significant figures.” All digits in a reported result are expected to be known defi- nitely, except for the last digit, which may be in doubt. Such a number is said to contain only significant figures. If more than a single doubtful digit is reported, the extra digit or digits are not significant. This is an important distinction. Extra digits should be carried in calculation (see 1050B.2). If an analytical result is re- ported as “75.6 mg/L,” the analyst should be quite certain of the “75,” but may be uncertain as to whether the “.6” should be .5 or .7, or even .4 or .8, because of unavoidable uncertainty in the analytical procedure. If the standard deviation were known from previous work to be 2 mg/L , the analyst would have, or should have, rounded off the result to “76 mg/L” before reporting it. On the other hand, if the method was so good that a result of “75.64 mg/L” could have been conscientiously reported, then the analyst should not have rounded it off to 75.6 mg/L. Report only such figures as are justified by the accuracy of the work. Do not follow the common practice of requiring that quantities listed in a column have the same number of figures to the right of the decimal point.

2. Rounding Off

Round off by dropping digits that are not significant. If the digit 6, 7, 8, or 9 is dropped, increase preceding digit by one unit; if the digit 0, 1, 2, 3, or 4 is dropped, do not alter preceding digit. If the digit 5 is dropped, round off preceding digit to the nearest even number: thus 2.25 becomes 2.2 and 2.35 becomes 2.4. When making calculations, perform all computations before rounding off results. Repeated rounding off can result in chang- ing the value of a reported result. For example, taking the measured value of 77.46 and rounding off to three significant figures yields 77.5. If the latter number were rounded a second time to two significant figures, the result would be 78. This is clearly a different result from a rounding of the original value of 77.46 to two significant figures (77).

3. Ambiguous Zeros

The digit 0 may record a measured value of zero or it may serve merely as a spacer to locate the decimal point. If the result of a sulfate determination is reported as 420 mg/L, the report recipient may be in doubt whether the zero is signifi- cant or not, because the zero cannot be deleted. If an analyst calculates a total residue of 1146 mg/L, but realizes that the

EXPRESSION OF RESULTS (1050)/Significant Figures

EXPRESSION OF RESULTS (1050)/Significant Figures

4 is somewhat doubtful and that therefore the 6 has no significance, the answer should be rounded off to 1150 mg/L and so reported but here, too, the report recipient will not know whether the zero is significant. Although the number could be expressed as a power of 10 (e.g., 11.5 10 2 or 1.15 10 3 ), this form is not used generally because it would not be consistent with the normal expression of results and might be confusing. In most other cases, there will be no

doubt as to the sense in which the digit 0 is used. It is obvious that the zeros are significant in such numbers as 104 and 40.08. In a number written as 5.000, it is understood that all the zeros are significant, or else the number could have been rounded off to 5.00, 5.0, or 5, whichever was appropriate. Whenever zero is ambiguous, it is advisable to accompany the result with an estimate of its uncertainty. Sometimes, significant zeros are dropped without good cause.

If

a buret is read as “23.60 mL,” it should be so recorded, and not

as

“23.6 mL.” The first number indicates that the analyst took the

trouble to estimate the second decimal place; “23.6 mL” would indicate an imprecise reading of the buret.

4. Standard Deviation

Suppose that a set of potential results is normally distributed

with a standard deviation of 100 mg/L and that a calculated value turns out to be 1450 mg/L. Then that 1450 is the best estimate available of this particular value, and from a Bayesian point of view, there would only be about a 31% chance that the true value was 1400 or lower or that the true value was 1500 or higher. It does not make sense to round the 1450 value to 1400. The arbitrary subtracting of half a standard deviation leaves us with

a value that does not represent our best estimate well. The

standard deviation should influence the last significant figures of

a calculation only by 0.5 1 and if more, the number of signif- icant figures cannot be justified. When reporting numbers in the form x y , always state whether y represents standard deviation, standard error, confi- dence limit, or an estimate of maximum bias. Standard devia- tions and standard errors often should be reported with extra digits (compared with single measurements) because they are calculated from variances and because they are square roots (see 1050B.5). When interpreting a quantity such as 1480 40, be aware that this notation seldom indicates a belief that the true value lies anywhere in the range from 1440 to 1520 with equal probability; instead, the probability is concentrated near the central value (1480).

5. Calculations

As a starting point, round off the results of any calculation in which several numbers are multiplied and divided to as few significant figures as are present in the factor with the fewest significant figures. 2 However, several potential reasons to mod- ify that guideline are noted below. Example: Assume that the following calculation must be made

to obtain the results of an analysis:

56 0.003 462 43.22

1.684

A ten-place calculator yields an answer of “4.975 740 998.” If the number 56 is an exact number (a count or a mathematical constant, such as ), it has no error associated with it and is considered to have unlimited significant figures. In that case, round off the result of the calculation to “4.976” because other numbers have only four significant figures. However, if 56 is an approximate measurement with uncertainly associated beyond the second figure, round off the result to “5.0” because 56 has only two significant figures. When numbers are added or subtracted, the number that has the least precision in its last significant digit limits the number of places that can justifiably be carried in the sum or difference. It is acceptable practice to carry one more digit beyond the least precise significant digit. Example:

The following numbers are to be added:

0.0072

12.02

4.0078

25.9

4886

The number “4886” is the least precise number (decimal place). Round each number in the sum to one or more digits beyond the least precise number and compute the sum.

0.0

12.0

4.0

25.9

4886.

4927.9

Round the result to the precision of the least precise number in the sum, 4928. Some calculators or computers round off numbers by a different rule that tends to bias results toward the larger digit in the last significant figure. Before using such a device, determine which

rounding techniques are programmed and if an incorrect rounding method is being used, reprogram to follow the correct rounding method. If it is not possible to reprogram, take unrounded number and manually round off using the correct scientific roundoff method. Interpret the foregoing guidelines flexibly. For example, consider

a series of measurements (u, v, and w) and a derived variable (y uv/w) that is calculated for each case. Suppose measurements give y 9.90972 and y 10.11389 for two cases with similar measured values. According to the guidelines, the first y should be rounded to 9.91; the final digit here is about 0.1% of the overall result. For comparable precision, round the y value for the second case to 10.11 (not 10.1) because the fourth digit is also about 0.1% of the number. To generalize, more significant digits should at times be provided for quantities having 1, 2, or 3 (say) as leading digits than for quantities beginning with 7, 8, or 9. Flexibility is also needed for the average of several numbers. The standard deviation (standard error) of a mean of N numbers

is only 1/ N as large as the standard deviation of the individual

numbers. Thus a mean of 100 numbers like d.dd is known to a precision of d.ddd. Even if fewer values than 100 are averaged, showing an extra digit may be justifiable. Because variances are

averages of squared deviations, they too are more precise than

EXPRESSION OF RESULTS (1050)/Other Considerations

EXPRESSION OF RESULTS (1050)/Other Considerations

individual deviations. Therefore, reporting variances with an extra digit or two is often justifiable. These guidelines refer only to final reported values. When per- forming any series of mathematical or statistical calculations, do not round measurements or other numbers until the very end of the analysis. Keep two or three extra digits for all intermediate calcu- lations, to reduce round-off errors, which can be substantial.

6. General Arithmetic Functions

In analytical calculations, it may be necessary to use functions other than simple arithmetic, such as logarithmic, exponential, or

trigonometric functions. A detailed treatment of significant fig- ures in such cases is available. 3

7. References

1. SCARBOROUGH , J.B. 1955. Numerical Mathematical Analysis, 3rd ed. Johns Hopkins Press, Baltimore, Md.

2. AMERICAN SOCIETY FOR TESTING AND MATERIALS . 1993. Standard Prac-

tice for Using Significant Digits for Test Data to Determine Confor- mance with Specifications; E29-93. Philadelphia, Pa.

3. GRAHAM , D.M. 1989. Significant figure rules for general arithmetic functions. J. Chem. Ed. 66:573.

1050 C. Other Cons