AN IMPROVED METHOD TO ANALYZE TRIMETHYLAMINE IN 

 FISH AND THE INTERFERENCE OF AMMONIA AND DIMETHYLAMINE 



Fern A. Bullard and Jeff Collins* 



ABSTRACT 



The trimethylamine content of most marine fish, especially the gadoid species, is internationally 

 accepted as an index of spoilage. However, ammonia, dimethylamine, and other amines also contribute 

 to the trimethylamine value. Variations in the conditions of the three current methods used to analyze 

 for trimethylamine content were studied in detail to determine the best condition to extract 

 trimethylamine and to reduce the extraction of ammonia, dimethylamine, and other amines. 

 Formaldehyde does not inhibit the interference from ammonia but the interference is negligible even 

 in advanced spoilage. Formaldehyde inhibits the interference from dimethylamine to some extent if 

 KOH is used as the base but increases the interference in the KjCOg method. The extraction of di- and 

 trimethylamine are highly dependent upon the base used and the temperature of extraction. A new 

 method of extracting at - 15° C using 45% KOH was developed that essentially eliminates interference 

 from ammonia, dimethylamine, and other amines. To directly compare the methods, the trimethyl- 

 amine content of a sample of spoiled walleye pollock, Theragra chalcogramma , flesh was determined 

 by the three currently used methods and the cold method of extraction. All methods gave similar 

 standard deviations but the K2CO3 method gave higher values than the KOH methods and the cold 

 method gave the lowest value. Various levels of trimethylamine and dimethylamine simulating 

 different qualities of fish and frozen storage times were added to samples of Pacific cod, Gadus 

 macrocephalus, Q.esh.. The cold method consistently extracts more accurate amounts of trimethylamine 

 with less interference from dimethylamine than any of the other extraction methods. 



The trimethylamine (TMAl content of most ma- 

 rine fish, especially the gadoid species, is accepted 

 internationally as an index of spoilage. Dyer's 

 1959 method of analysis for TMA, except for the 

 concentration of formaldehyde (FA), has been 

 adopted by the Association of Official Analytical 

 Chemists (Horwitz 1975). Trimethylamine is pro- 

 duced by the reduction of trimethylamine oxide by 

 microorganisms (Poller and Linneweh 1926). 

 Ammonia, dimethylamine (DMA), and other vol- 

 atile bases are also formed when fish spoil and to 

 some extent interfere with the measurement of 

 TMA. In advanced spoilage, some of the higher 

 aliphatic amines are formed by decarboxylation of 

 amino acids and may cause interference (Dyer 

 1945). 



A number of investigators studied the TMA 

 method to improve the accuracy and reduce the 

 effects of ammonia, DMA, and other amines. Dyer 

 (1945) adapted the method of determining amines 

 to fish and used 0.02% picric acid in dry toluene 

 instead of chloroform (Richter 1938; Richter et al. 

 1941). Dyer and Mounsey (1945) used a trichloro- 



'Northwest and Alaska Fisheries Center Kodiak Investiga- 

 tions-Utilization, National Marine Fisheries Service, NOAA, 

 PO. Box 1638, Kodiak, AK 99615. 



acetic acid (TCA) extract of fresh cod instead of the 

 unstable press juice or weighed samples. Hashi- 

 moto and Okaichi ( 1957) claimed that variation of 

 temperature caused serious errors in the determi- 

 nation of TMA, and recommended a 30° C extrac- 

 tion with 257r KOH rather than 50% K-^COa and 

 room temperature. Tozawa et al. ( 1971 ) confirmed 

 these findings and showed that 25% KOH reduced 

 the interference caused by DMA and claimed the 

 formation of a compound from FA and DMA which 

 was not extracted in the presence of hydroxide 

 ions. Murray and Gibson (1972) found that 45% 

 KOH extracted more TMA and gave more linear 

 and reproducible results than if extracted with 

 25% KOH or 50% K^COg. 



The three current methods of analysis for TMA 

 employ 25% KOH, 45% KOH, or 50% K2CO3 to 

 release TMA for extraction into the toluene layer 

 and result in different absorbancies for the picrate 

 color with DMA and TMA. In general, the use of 

 K2CO3 results in a higher extraction of DMA and a 

 lower extraction of TMA than if KOH were used. 

 Ideally, the best method to measure TMA content 

 should result in complete extraction of TMA and 

 zero extraction of ammonia (NH3), DMA, and 

 other amines so that these components will not 

 contribute to the TMA value. A new method was 



Manuscript accepted December 1979. 

 FISHERY BULLETIN; VOL. 78, NO. 2, 1980. 



465 



