AOAC Methods in Codex STAN 234 (Preliminary Methods Review)

851

BARBANO ET AL.: J. ASSOC. Off. ANA\... CHEM. {VOL. 73, NO. 6, 1990)

mined that the block digestor does oot work well with the larger volumes of reagents that the traditional unit can ac– commodate. Thus, for the 250 mL tubes on block digestors, 5 g sample, l 2 g potassium sulfate, and 20 ml sulfuric acid (still a 0.6 ratio of salt to acid) worked well. The fourth phase of preliminary work involved an evalua– tion of titanium oxide plus copper sulfate (in the Associate Referee's laboratory) as an alternative catalyst to copper sulfate, as suggested by Kane (10). Milk was digested with and without titanium oxide (0.6 g/tlask) at 2 different ratios of potassium sulfate to acid (0.6 vs 0.835). The fifth phase of preliminary work was a collaborative study in July 1988 using recovery solutions (ammonium sul– fate and glycine) with the optimized Kjeldahl procedure fo.r milk. Copper catalyst was used. The procedure was the same as the method in the present. paper with the following excep– tions: 1 blanlc was run instead of 2; indicator was added dropwise to each distillate before tit ration; hydrogen perox– ide was used to suppress foaming during digestion for both traditional and block equipment ( 5 mL 30% hydrogen permc· ide was add-Cd immediately before digestion); both tbe block and traditional procedures used 12 g potassium sulfate, 20 ml. sulfuric acid, and 55 mL 50% sodium hydroxide; and boiling time of digest was I h after clearing. Three different levels of both ammonium sulfate and glycine were analyzed in blind duplicate by 9 laboratories (6 laboratories had block digestors and steam distillers; 3 laboratories had traditional macro-.Kjeldahl djgestors and distillers). Conclusions trom Preliminary Work The first phase of the preliminary worlc (May and June 1986), with each laboratory using its own KjeJde.hl method and current equipment, indicated that the micro-Kjeldahl methods did not perform as well as the macro-Kjeldahl meth· ods. The standard deviation of the difference between dupli– cates for percent nitrogen recovery for the micro-Kjeldahl methods was about 5 times that of the macro-Kjeldahl sys-– terns (1 .90 vs 0.39). There was no difference in average nitrogen recovery for copper vs mercury catalyst. The second phase of the preliminary work indiceted that equivalent nitrogen recoveries could be obtained with either copper. or mercury catalysts. We concluded that with optimi– zation of the digestion parameters instead of merely substi– tuting one catalyst for another, comparable results could be obtained with mercury or copper catalysts. The third phase of the preliminary work found that 0.05 g copper sulfate pentahydrate/nask was sufficient to yield good test resuhs. The optimum ratio of potassium sulfate to sulfuric acid was determined to be between 0.56 and 0.67. A digestion time of I to 2 h after clearing was determined to give best test results for both the iraditional and block equip– ment. Excessive foaming can be a problem on the block digestor (on the traditional system, when the sample foams, one need only rotate the nask to alleviate the problem). Both predrying the milk and adding hydrogen peroxide to the liquid sample immediately before digestion were found to help reduce foaming in block digestor systems. Neither of these 2 approaches were incorporated into the final procc· dure. The first approach would add a minimum of 4 to 5 h to the test, which is not acceptable. The second approach (use of hydrogen peroxide) is somewhat hazardous. Boiling rods and antifoam ta blets did not reduce the foeming problems on the block digestor. Programming the block. digestor temperature (starting at a low temperature and increasing gradually to

The primary objectives of the present research were to optimize the macro-Kjeldahl method for milk for both tradi– tional and block digestion equipment, to determine method performance (s, and sR), and to compare performance of the block digestor/steam distiller system to the traditional Kjel– dahl system. Pre-Collaborative Study Work 'The first phase of preliminary work involved an evaluation of equipment and methods that different laboratory person· nel used for determination of total protein by the Kjeldahl method. In May 1986, 8 laboratories participated in the first phase of the study. These laboratories were sent 6 recovery solutions (3 ammonium sulfate and 3 gtycine) in blind dupli· cate. The laboratories were asked to determine percent nitro– gen by the method with whieh they currently tested milk samples. Four laboratories used copper catalyst, 3 laborato– ries used mercury catalyst, and 1 laboratory used selenium catalyst. Amounts of potassium sulfate aod sulfuric acid used in digestions varied widely among laboratories. Six. of the laboratories used a macro-Kjeldahl system (both block diges– tor and traditional equipment) and 2 laboratories used a micro-Kjeldahl system. In June 1986, this first phase study was replicated. The se~ond phase of preliminary work was undertaken in August 1986 to determine if there was a difference in test results with mercury or copper catalysts. The same 8 labora· tories that participated in the first-phase study participated in the s~nd phase of the study. Analysts were requested to use their current method. They were supplied with the cata– lyst that I.hey did not normally use (i.e., if they normally used mercury, they were supplied with copper and vice versa). The laboratory that used selenium was supplied with both copper and mercury catalysts (along with instructions for both), Digestion conditions (ratio of potassium sulfate to sulfuric acid) were not changed; catalysts were merely substituted for what was currently being used in each laboratory. Analysts were ser:lt 3 ammonium sulfate nitrogen recovery solutions (io blind duplicate) and 3 milks (in blind duplicate). The third phase of preliminary work optimized the diges– tion/distillation parameters for each equipment type with copper catalyst and established a well documented, clearly defined method for milk analysis. Optimum chemical qua nti– ties for digestion (catalyst, potassium sulfate. sulfuric acid, and sample) were determined, AOAC and (DF have re· quired a 5 g sample size, 15 g potassium sulfate, and 25 mL sulfuric acid for Kjeldahl nitrogen determination for milk. Although AOAC currently requires mercury catalyst for milk analysis, IDF has switched to the use of copper sulfate (1 ml of a 0.05 g/mL copper sulfate pentahydrat_e solution}. Optimization of the digestion procedure involved varying the amount of catalyst (0 to 0.84 g CuSO4-5H2O) and potassium sulfate/sulfuric acid ratio (0.3 to 0.7). Digestion time needed to be established (AOAC digests for ~30 min after clearing with mercury catalyst; IDF digests for 1.5 h after clearing with copper catalyst). Digestion times of 0.5 to 2.0 .h after clearing were evaluated. Approaches to decrease the amount of sample foaming on the block digestors were also evaluated. These methods in– volved programming the block temperature during digestion, drying the milk in a 1 OQ°C oven before adding reagents, addition of 5 mL 30% hydrogen peroxide to the samples immediately before digestion, addition of antifoam tablets, and addition of boiling rods during digestion. lt was deter·

Codex Trial Method Review