THE DETERMINATION OF THE MOISTURE CONTENT OF GELATINE
 

C.G.B. COLE

Leiner Davis Gelatine SA (Pty) Ltd. PO Box 5019, West Krugersdorp 1742, South Africa.
 

Presented at the South African Association of Food Science and Technology Conference. 1987.
 

SUMMARY.

The determination of moisture content is generally considered to be a simple and accurate analytical procedure. However with gelatine, which is hygroscopic when dry, the inaccuracy of the analysis was perplexing. Eventually, it was found that the result obtained was greatly influenced by ambient humidity conditions in spite of using dessication during cooling.
 

Introduction.
 

The South African National Specification for gelatine, contained in the Regulations to the Foodstuffs, Cosmetics and Disinfectants Act, 1972 (Act 54 of 1972), requires a moisture content of less than 16%. For this and production control reasons, it is necessary to monitor production gelatin moisture content.

It is necessary, in routine work, to be able to assess the accuracy and repeatability of results, due to the possibility of personal and routine errors and equipment failures. Hence, a standard is analysed daily together with the daily samples. Although repeatability errors were observed these were not sufficient to be a problem, however the errors were inexplicably large. On examination of the data it was noted that high results were obtained in winter and low results were obtained in summer with the total variance of the monthly average being in the range of ca. 0.7%. As winters in Johannesburg are dry and summers wet, the observation was exactly the opposite of what one would expect from an expected loss of moisture during dry weather and gain in moisture during wet weather! Hence, an investigation of the phenomenon was undertaken.

Methods.

A clean dry silica crucible was weighed to 0.1 mg. 2500 ± 1 mg gelatine was added to the crucible. The crucible was then dried in a 105 C oven for >16 hours. The crucible was then cooled in a calcium chloride desiccator for 90 minutes before re-weighing.
 

e.g.     Empty Crucible 23.7814 g
          Dry Crucible 26 0261 g
          Dry Gelatine 2.2445 g


Moisture content = (2.5 - 2.2445)/2.5 x 100 = 10.21%

The weighing error of 1 mg = 0.4 % becomes ± 0.04% error in the % moisture content of the gelatine. Furthermore, it was observed that whether the gelatin was dried for 16 hours (overnight) or for 3 days (weekend) the results were unaffected.

Results and Discussion.

1.
    6 duplicate pairs of determinations were conducted on Standard 185:

    Variation between pairs -   Maximum 0.07%
                                            Minimum 0.00%
                                            Mean 0.04%

2.
    9 single determinations on separate days:

    Moisture Content -         Maximum 11.35%
                                         Minimum 10.78%
                                         Mean 11.03%
                                         Std. Dev. 0.18%

The above clearly illustrates the problem of the repeatability of the determination of the moisture content of gelatin. From 1 above, it is clear that the intrinsic error of the determination was 0.04% as expected, but the day to day variance was at least 4 times as great as shown in 2 above. During this study it was also noted that the lowest result was obtained on a rainy day!

3.
The analysed monthly mean moisture content of the Standard 185 from July 1984 to June 1986 together with the 8:00 hour (8 am) average monthly humidity statistics for Johannesburg, up to 1984(1) are shown in Table 1. The absolute "humidity" data was calculated from the Temperature and % RH data, using the methods given by Barenbrug (1974)(2) assuming an atmospheric pressure for Johannesburg of 82.5 kPa.

Table 1.
The Monthly Mean Moisture Content of Gelatine Standard 185 and the Monthly Average Humidity for Johannesburg.
 
Weather at 8:00 hrs (1) Moisture % (Std. Dev)
Month Temp. C % RH Humidity 
H2O. mg/l(2)
1984-85 1985-86
July
August
September
October
November
December
January
February
March
April
May
June
6.9
9.2
13.2
15.5
16.7
17.6
17.7
17.1
15.9
13.3
10.1
6.7
66 
62
60
65
71
73
77
80
80
78
71
69
6.3
7.2
9.8
11.8
13.2
14.2
14.8
14.0
13.0
10.7
8.2
6.5
10.51 (0.21)
10.27 (0.24)
9.97 (0.35)
9.74 (0.30)
9.80 (0.38)
9.93 (0.28)
9.64 (0.34)
9.58 (0.46)
9.99 (0.23)
10.12 (0.40)
10.21 (0.30)
10.29 (0.30)
10.44 (0.22)
10.28 (0.23)
10.30 (0.30)
10.15 (0.33)
9.96 (0.30)
9.84 (0.33)
10.04 (0.27)
10.07 (0.24)
10.24 (0.33)
10.21 (0.44)
10.36 (0.32)
10.57 (0.25)

The repeating inverse relationship between the analytical moisture content of Standard 185 and the absolute atmospheric moisture content shown in Table 1 was analysed by regression analysis. Where M = Moisture % and H = Absolute Humidity the following equation was obtained by second order polynomial regression:

M = 11.0037 - 0.1005H + 0.0015H2   (r = - 0.809,  n = 24).

From Tables(3) the correlation coefficient, r, of 0.80 for 22 degrees of freedom is significant at the 0.0005 level.

FIGURE 1.
 

In Figure 1, the monthly average humidity is plotted as; in addition average monthly analysed % moisture content as well as the % moisture content smoothed by polynomial regression.

4.
From Table 1 it can be seen that the ambient absolute humidity varies from 6 to 15 mg/l air, in Johannesburg.

The following desiccants dry air at 25 degree C(4):
 

CaCl2 granular 0.14 - 0.25 mg H2O/l
H2SO4 (95-100%) 0.003 - 0.3 mg H2O/l
Al2O3 3 x 10-3 mg H2O/l
Silica Gel 1 x 10-3 mg H2O/l
P2O5 2.5 x 10-5 mg H2O/l


Conversion to silica gel in desiccators from July 1986 resulted in a marked reduction in the variation (10.63% to 10.41%) of the monthly mean moisture value of the Standard but the monthly standard deviation of the mean varied from 0.36% to 0.21% over the months to April 1987, which is only a small improvement on the data in Table 1.

Conclusions.

References.

1. Climate of South Africa - Climate Statistics up to 1984. pg. 385. Government Printer, Pretoria. 1986.
2. Barenbrug, A.T.W. Psychrometry and Psychrometric Charts. 3rd Edition. Chamber of Mines. South Africa. 1984.
3. Freud, J.E. & Williams, F.J. Elementary Business Statistics - The Modern Approach. Prentice-Hall International Inc., New Jersey. 1964.
4. Chemical Reagents Catalogue. C Merck & Co. 1974/75.