FOOD COMPOSTION AND ADDITIVES

Spectrophotometric Determination of Nitrate in Baby Foods:

Collaborative Study
SJBERG & ALANKO: JOURNAL OF AOAC INTERNATIONAL VOL. 77, NO. 2, 1994
ANNA-MAIJA K. SJBERG
VTT Biotechnology and Food Research, Tietotie 2, Espoo, Finland
TIMO A. ALANKO
VTT Computing Service, Tekniikantie 48, Espoo, Finland
Collaborators: V.-M. Ahonen; I. Aminoff; S.-L. Antinoja; J. Arstila; E. Bloigu; M. Hokkanen; T. Holster; B. Hllis; T. Idman; T.
Kari; S. Karppinen; M. Kettunen; R. Kettunen; M. Kuusivuo; V. Laitinen; R. Laurila; H. Lehto; T. Nieminen; M. Oksanen; L.
Pero; E. Peura; H. Pihlaja; I. Puntari; P. Raulos; R. Reini; M. Rohunen; A. Skrkki; J. Soimajrvi; H. Sojonen; T. Teerinen; M.
Tuhkanen; P. Vilpponen; C. Westlin

A collaborative study was conducted of a spectrophotometric method for determination of nitrate after cadmium reduction to nitrite in baby foods containing meat. Thirty-one municipal and 2 industrial
food laboratories participated in the study. The
study design involved 2 baby food matrices. Samples of both matrices were prepared at 3 concentration levels between 52 and 309 mg NaNO3/kg as
blind duplicates. A blank without added nitrate was
also included. The outlier percentage of the results
was very low (4.3%). It was typical for the method
that recoveries were slightly >100%. Recoveries for
baby foods varied between 113.3 and 116.9%, and
were acceptable for control purposes. The relative
standard deviations for repeatability were 5.0
18.1%. The relative standard deviations for reproducibility were 8.321.6%. Three collaborators also
evaluated liquid chromatographic technique for nitrate determinations. These preliminary results are
presented but are not analyzed statistically. The
spectrophotometric method was adopted first action by AOAC INTERNATIONAL.

For control purposes there are only a few spectrophotometric methods applicable for both nitrates and nitrites, particularly
in meat products. A classic method for the determination of
nitrate is reduction of the nitrate to nitrite on a cadmium column
followed by the modified Griess colorimetric assay for the reduced and the original nitrite (1). AOAC has also presented a
column system, a modified Jones reductor, for determination of
nitrate and nitrite in cheeses (2). The method has been used for
determining nitrate and nitrite in aqueous extracts of plant materials (3), but the method is unreliable because the results are
dependent on the degree to which nitrate is reduced to nitrite
(4). Because of this, the method based on spongy cadmium reduction was tested collaboratively. On the basis of our earlier
investigations, the nitrite determination omitting the spongy
cadmium reduction step was less critical. The method is recommended for nitrate determinations in meat, fish, and vegetable
products in governmental food control laboratories of Finland.
The collaborators were also asked to perform the determinations with a liquid chromatography method (LC), if they had
such a method available for control purposes. These collaborators applied the LC technique published by Eggers and Cattle (5).

ecent concern over nitrosamines has lead to increased

interest in nitrate and nitrite levels occurring in a wide
range of foods. There is now a universal trend toward
lower limits for nitrate and nitrite, and a simple and reliable
method is needed for control analyses. Because of the equilibrium between nitrate and nitrite in meat products, the determination of both components is needed.

Two commercial baby foods of different compositions were

sent to 33 collaborators. Baby food 1 contained pork 18%, potato 17%, apple 4%, plum 4%, starch 3%, wheat flour 3%, rice
2%, and water. Baby food 2 contained potato 18%, veal 9%,
pork 9%, wheat flour 3%, starch 3%, milk powder (without fat)
2%, vegetable oil 1%, lemon juice below 1%, and water. Food
additives were dill 1% and salt 0.25%. The matrices were
packed under vacuum in glass jars. The collaborators also received 14 vials (4.5 mL) of test solutions (nitrate at different
concentration and nitrate-free water blank) for the preparation
of 14 samples. Contents of the test solutions were such that
spiking resulted in blind duplicates at 3 concentration levels
and a blank for each baby food (Tables 1 and 2). The 14 sam-

Submitted for publication August 11, 1992.

The recommendation was approved by the Committee on Foods I, and
was adopted by the Official Methods Board of the Association. See AOAC
International Official Methods Board News (1993) J. AOAC Int. 76, 33A,
and Methods Adopted First Action (1993) The Referee, 17, March issue.

Collaborative Study

ples were randomly numbered and the participants were instructed to analyze the samples in numerical order. Before the
collaborative study, each collaborator received practice material for the determination of nitrate.
The collaborators prepared samples according to the instructions: 60.0 g of baby food was weighed from each glass
jar, 4.0 mL test solution and 26.0 mL water was added and homogenized for ca 5 min, and the analysis was began immediately.
On the basis of control analyses it was confirmed that the
nitrate concentration was stable; no nitrate was transformed to
nitrite before analysis. Therefore, only the nitrate determination
was needed.
993.03 Nitrate in Baby FoodsSpectrophotometric
Method
First Action 1993
(Applicable to determination of nitrate, calculated as sodium nitrate, in baby foods containing meat, with detection
range of 100300 mg/kg.)
(Caution: See Appendix: Laboratory Safety for safe handling of acids.)
Method Performance
See table of method performance data, 993.03.

A. Principle
Nitrate in baby foods containing meats is reduced to nitrite
by treatment with spongy cadmium. Nitrite is then determined
colorimetrically by absorbance at 530 nm, and compared with
standard curve.

B. Apparatus
(a) Blender.With explosion-proof motor.
(b) Water bath.Capable of maintaining 100 2.
(c) Spectrophotometer.Capable of measuring absorbance at 530 nm to 0.006 0.001 absorbance unit, 10 mm optical path length.
(d) Glassware.50, 100, 200, and 1000 mL volumetric
flasks, 10 mL graduated pipets, and glass-stoppered mixing
cylinders graduated to 20 mL. Pyrex quality.
(e) Filter papers, nitrate and nitrite free.(1) Qualitative
grade, 2025 m pore size (Whatman No. 4 is suitable). (2)
Qualitative ashless grade, 2.58 m pore size (Whatman No.
40 or 42 is suitable).

C. Reagents
Note: Use freshly boiled distilled H2O.
(a) Sodium nitrate solutions.Dry sodium nitrate 2 h at
105. (1) For stock solution, transfer exactly 1000.0 mg to 1 L
volumetric flask and dilute to volume with H2O. (2) For working solution, dilute 10 mL stock solution to 100 mL with H2O.
(3) For standard solutions, dilute 0, 5, 10, 15, and 20 mL of
working solution to 100 mL with H2O.

(b) Saturated borax solution.Dissolve 5 g sodium

tetraborate (Na2B4O710H2O) in 100 mL hot H2O.
(c) Zinc sulfate solution.Dissolve 30 g zinc sulfate
(ZnSO47H2O) in 100 mL H2O.
(d) Zinc metal.Powder, not oxidized.
(e) Cadmium sulfate solution.Dissolve 10 g cadmium
sulfate (3CdSO48H2O) in 100 mL H2O.
(f) Ammonium hydroxide, 25%.
(g) Sulfanilic acid solution.Dissolve 600 mg sulfanilic
acid (C6H7NO3S) in 50 mL hot H2O. Let cool; add 20 mL glacial acetic acid and dilute to 100 mL with H2O.
(h) N-(1-Naphthyl)ethylene diammonium dichloride reagent.Dissolve 20 mg N-(1-naphthyl)ethylene diammonium dichloride (C12H16Cl2N2CH3OH) in 20 mL glacial
acetic acid and dilute to 100 mL with H2O.
(i) Color reagent.Mix equal volumes of sulfanilic acid
solution, (g), and N-(1-naphthyl)ethylene diammonium dichloride reagent, (h), immediately before use. Discard any unused
color reagent.
(j) Acetic acid solution, 20%.

D. Sample Preparation
Homogenize sample. Weigh 310 g homogenate, to nearest
mg, into 200 mL volumetric flask. Add 150 mL hot water and
10 mL borax solution, C(b), to precipitate protein. Warm volumetric flask in boiling water bath 15 min, and slowly add 4 mL
zinc sulfate solution, C(c), with shaking. Cool to room temperature in cool water bath. Dilute to volume with water, mix,
and filter, B(e)(2), ca 50 mL. Prepare blank without sample material.

E. Reduction of Nitrate to Nitrite

For each sample, weigh ca 600 mg Zn powder into separate
50 mL volumetric flasks and spread powder over bottom of
flask. Prepare 5 additional flasks for standards. Carefully add
4 mL cadmium sulfate solution, C(e), to zinc powder in flask
to obtain homogeneous mixture. Note: Care is necessary to ensure a repeatable spongy cadmium preparation. Let newly
formed spongy metallic cadmium stand 10 min without moving. Add 2 mL 25% NH4OH and 10 mL sample solution, D, to
a flask. Prepare standard nitrate concentrations (containing 0,
50, 100, 150, and 200 g NaNO3) by adding 10 mL of each
standard solution, C(a)(3), to separate volumetric flasks prepared with spongy cadmium. Shake flasks exactly 1 min to
loosen spongy cadmium; then let stand 10 min. Dilute to volume with H2O and filter, B(e)(1).
After use, pour contents of volumetric flasks to waste bottle.
Dissolve possible residues in volumetric flasks with conc HCl
to another waste bottle. Collect waste from color reaction in
another bottle. Arrange for proper disposal of waste bottles.

F. Nitrite Determination
Pipet 10.0 mL clear filtrates of sample and standard solutions (equivalent to 0, 10, 20, 30, and 40 g NaNO3) to separate
glass-stoppered mixing cylinders. Add 10.0 mL color reagent
to each, mix 1 min by hand, and record absorbance at 530 nm

using H2O blanks. Determine mg NaNO3/kg sample by comparison with standard curve.

G. Calculations
mg NaNO3/kg sample = (b 100)/m

where b = sodium nitrate from standard curve (g) and m =

weight of sample homogenate (g).
Ref.: JAOAC 77, 425 (1994)
Results and Discussion
Twenty-nine collaborators reported results for both sample
materials and 4 collaborators for 1 sample material (Tables 1
and 2). The results for each sample are presented after subtracting the individual blank values that were found by the collaborators for the food samples spiked with nitrate-free water. The
average nitrate concentrations were 26.8 13.3 mg/kg for baby
food 1 and 73.8 15.6 mg/kg for baby food 2.
The statistical evaluations were carried out according to the
IUPAC 1987 recommendations (6) (Tables 3 and 4). The outlier percentage of the results was as low as 4.3%. It was typical
for the method that recoveries were slightly >100%. Recoveries for baby foods 1 and 2 with spiked concentration levels of
52, 105, and 309 mg NaNO3/kg and 60, 204, and 281 mg
NaNO3/kg, respectively, varied between 113.3 and 115.6%,
and between 115.8 and 116.9%, respectively, and were acceptable for control purposes. It is important that interpretations of
results at concentration levels less than 100 mg/kg are made
carefully, taking into account the positive recovery. Relative
standard deviations for repeatability varied between 5.3 and
11.3% for baby food 1, and between 5.0 and 18.1% for baby
food 2. Relative standard deviations for reproducibility varied
between 9.0 and 14.5% for baby food 1, and between 8.3 and
21.6% for baby food 2.
The results were compared with the results of a large number of collaborative studies with different analytes at various
concentrations and with different methods conducted by
AOAC (7). The upper curve in Figure 1 represents an empirical
confidence limit for the reproducibility relative standard deviation (RSDR) and the lower curve the expected value of RSDR
on the basis of a large number of AOAC studies (8). As can be
seen from Figure 1, the RSDR for sodium nitrate is above the
confidence limit in one case of a low concentration level but
lower at medium of higher levels. According to Horwitz (9) the
repeatability relative standard deviation (RSDr) is 0.500.67 of
the RSDR value. If the RSDr is close to the RSDR, as in both
materials in this study, all the laboratories are performing the
determinations in a comparable manner and the random component constitutes almost the entire source of variability.
The average recoveries using the LC method (Table 5) were
between 81 and 102%.

Comments and Recommendations

Some collaborators made comments concerning sample
preparation. Preparing the samples at the individual laboratories was considered practical to guarantee stability of the samples. The number of samples to be prepared, however, was considered to be too high by the small laboratories, although for
estimation of the parameters affecting the reliability of the
method, fewer samples could not have been used.
Collaborators also returned several comments regarding the
method: Although the polluting nature of cadmium was taken
into account in the recommended procedure, and instructions
for handling of wastes were given, its use was criticized. The
amount of sample weighed was not specified in the method
because it was assumed that if the concentration of nitrate is
small, 10 g of sample should be used. In other cases it is easier
to use smaller amounts to facilitate sample preparation at the
filtration stage. Five standards including the blank were specified (0, 10, 20, 30, and 40 g NaNO3). The suggestion of 1
collaborator was taken into consideration by adding 1 standard
point (30 g) to the original procedure. Difficulties with filtration of the test sample preparation were reported. Filtration of
whole volumes (200 mL and 50 mL) is not necessary. The importance of the use of nitrate-free filter papers was recognized.
The question of the absorbencies of standard solutions was also
discussed. The average absorbencies and standard deviations
obtained were 0.006 0.012 for 0 g sodium nitrate; 0.104
0.023 for 10 g; 0.201 0.031 for 20 g; 0.289 0.006 for
30 g; and 0.397 0.075 for 40 g.
The method is applicable to the determination of nitrate in
baby foods containing meat at levels of 100300 mg
NaNO3/kg. Subtraction of a blank is not required. On the basis
of the results of this study, it is recommended that the spectrophotometric method for determination of nitrate in baby foods
be adopted first action.
Acknowledgments
The authors thank the following laboratories for their participation in this study:
Chymos Laboratory, Lappeenranta; Food and Environmental Laboratory of the City of Helsinki; Saarioinen Laboratory, Sahalahti; and the municipal food control laboratories at
Hmeenlinna, Iisalmi, Imatra, Joensuu, Kajaani, Kauhajoki,
Kokkola, Kotka, Kouvola, Kuopio, Lahti, Lappeenranta,
Lieksa, Mikkeli, Mntta, Nurmes, Oulu, Pietarsaari, Rauma,
Riihimaki, Rovaniemi, Seinjoki, Tampere, Turku, Vaasa,
Valkeakoski, Vantaa, Varkaus, Ylivieska, and nekoski.
References
(1) Sen, N.P., & Donaldson, B. (1978) J. Assoc. Off. Anal. Chem.
61, 13891394
(2) Official Methods of Analysis (1990) 15th Ed., AOAC, Arlington, VA, sec. 976.14
(3) Bachmann, J.H., Daniel, R.C., & Ruf, D. (1987) Mitt. Gebiete Lebensm. Hyg. 78, 168181

(4) National Academy of Sciences (1981) The Health Effects of

Nitrate, Nitrite, and N-Nitroso Compounds, Committee on
Nitrite and Alternative Curing Agents in Food, Assembly of
Life Sciences, National Academy Press, Washington, DC
(5) Eggers, N.J., & Cattle, D.L. (1986) J. Chromatogr. 354, 490494
(6) ISO Report of the IUPAC Workshop on the Harmonization
of Collaborative Analytical Studies (1987) Geneva, May 4
5, 1987
(7) Horwitz, W., Kamps, L.R., & Boyer, K.W. (1980) J. Assoc.
Off. Anal. Chem. 63, 13441354
(8) Horwitz, W. (1988) J. Assoc. Off. Anal. Chem. 71, 160171
(9) Horwitz, W. (1982) J. Assoc. Off. Anal. Chem. 65, 67A76A

Table 993.03 Method performance for nitrate in baby

foods 1 and 2 containing meat
Spike, Recovery
av.,
Baby
mg
food NaNO3/kg mg/kg
1
2
1
2
1
2

52
60
105
204
309
281

58.9
69.6
121.4
238.3
350.2
328.7

7.18
12.58
7.82
17.21
18.38
16.51

8.67
15.01
14.53
26.24
31.55
27.18

RSDr, % RSDR, %
11.33
18.08
6.52
7.22
5.25
5.02

14.49
21.57
11.96
11.02
9.01
8.27

Table 1. Spectrophotometric determination of sodium

nitrate (mg/kg) in baby foods 1
Level 1

Level 2

Level 3

Lab.

Lab.

Aa

Ba

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Analyte
concn,
mg/kg

58
53
64
66
46
80
75
73
60
62
59
59
d d

58
65
56
44
63
46
61
57
68
77
43
56
63
60
59
62
49
64
51
59

45
65
54
66
52
b b
10
64
51
67
64
60
69

63
57
62
55
57
51
63
56
60
59
37
56
65
81
49
67
43
53
74
58

112
105
119
137
123
128
152
101
122
124
120
125

123
120
127
112
127
102
113
99
113
139
117
108
b
73b
139
108
130
110
117
129
122

97
116
120
137
111
b
74b
142
126
149
121
124
135

126
124
128
115
131
98
110
99
103
154
111
106
125
145
131
141
101
116
151
123

361
323
375
386
359
355
396
372
332
363
360
412

325
380
369
332
330
303
345
369
377
359
274
283
204
370
320
360
360
301
364
360

277
336
383
406
357
309
390
bc
706b,c
371
368
360
399

357
379
358
329
345
305
340
346
345
392
298
333
b
150b
350
342
365
338
302
373
349

a
b
c
d

52

105

A and B indicate replicate sample identification.

Cochran outlier (P <0.001).
Grubbs outlier (P <0.01).
Not determined.

309

Table 2. Spectrophotometric determination of sodium

nitrate (mg/kg) in baby foods 2
Level 1
Aa
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Analyte
concn,
mg/kg
a
b
c
d

60

Level 3

Ba

153b
75
42
81
73
105
64
56
c
167c
77
81
71
19
72

69
66
61
61
41
73
54
87

72
72

70
87
69
76
70
77

171
240
259
258
232
224
252
236
270
238
256
252
263
232

251
224
224
192
200
259
215
263

274
225

200
245
249
236
217
244

251
198
258
242
241
269
292
245
274
245
256
247
258
215

248
219
231
196
203
237
221
263

303
251

203
266
251
198
207
212

290
300
364
356
328
348
315
319
276
327
326
331
225
343

349
342
321
279
285
351
333
370

373
331

313
333
322
342
330
391

279
336
337
349
328
348
340
316
340
324
343
289
b
393b
329

341
288
312
294
281
328
317
352

346
339

302
341
330
340
299
406

30
70
71
80
81
61
86
60
72
82
71
73
50
71
d d

69
62
52
58
53
62
72
33

63
72

58
87
71
80
83
112

Level 2

204

A and B indicate replicate sample identification.

Cochran outlier (P <0.001).
Grubbs outlier (P <0.01).
Not determined.

281

Table 3. Method performance for spectrophotometric

determination of nitrate in baby foods 1

Table 4. Method performance for spectrophotometric

determination of nitrate in baby foods 2

Parameter

Parameter

Av. rec.,
mg/kg
Av. rec., %
sL
sR
sr
RSDR, %
RSDr, %
RSDr/RSDR

Outliers

Level 1

Level 2

Level 3

included
excluded
included
excluded
included
excluded
included
excluded
included
excluded
included
excluded
included
excluded
excluded

58.6
58.9
112.6
113.3
0.0
4.85
10.82
8.67
11.25
7.18
19.20
14.49
19.20
11.33
0.78

120.1
121.4
114.3
115.6
11.07
12.17
16.40
14.53
12.10
7.82
13.65
11.96
10.07
6.52
0.54

350.7
350.2
115.0
113.3
42.87
25.84
62.80
31.55
45.88
18.38
17.90
9.01
13.08
5.25
0.58

Av. rec.
mg/kg
av. rev., %
sL
sR
sr
RSDR, %
RSDr, %
RSDr/RSDR

Outliers

Level 1

Level 2

Level 3

included
excluded
included
excluded
included
excluded
included
excluded
included
excluded
included
excluded
included
excluded
excluded

71.9
69.6
119.8
115.8
0.0
8.19
22.54
15.01
23.46
12.58
31.35
21.57
32.62
18.08
0.84

238.3
238.3
116.8
116.3
19.82
19.82
26.24
26.24
17.21
17.21
11.02
11.02
7.22
7.22
0.66

328.0
328.7
116.7
116.9
15.08
21.58
31.01
27.18
27.09
16.51
9.45
8.27
8.26
5.02
0.61

Table 5. Liquid chromatographic determination of sodium nitrate (mg/kg) in baby foods

Level 1
Lab.

Level 2

Level 3

60
48

60
44

110
108

110
99

230
289

310
293

Baby food 1
1
2
3
Analyte concn, mg/kg
Av. rec., mg/kg
Av. rec., %

52
53.0
101.9

105
106.8
101.7

309
280.5
90.8

Baby food 2
1
2
3
Analyte concn, mg/kg
Av. rec., mg/kg
Av. rec., %

60
37
63

40
49
42
60
48.5
80.8

190
181
160

190
199
168
204
181.3
88.9

270
266
245

250
258
258
281
254.7
90.6

Figure 1. Collaborative study results of the spectrophotometric determination of sodium nitrate in baby foods
1 ( ) and 2 ( ).