Study on the formation reaction of the silico-molybdenum complex in seawater

碩士 === 國立臺灣大學 === 海洋研究所 === 105 === Dissolved reactive silicate in natural water is usually determined colorimetrically based on the formation of a yellowish silicomolybdenum complex. The product and reaction rate can be affected significantly by the amounts of reagents added, and the optimal condit...

Full description

Bibliographic Details
Main Authors: Cheng-Yun Wang, 王程筠
Other Authors: 白書禎
Format: Others
Language:zh-TW
Published: 2017
Online Access:http://ndltd.ncl.edu.tw/handle/8r9akk
Description
Summary:碩士 === 國立臺灣大學 === 海洋研究所 === 105 === Dissolved reactive silicate in natural water is usually determined colorimetrically based on the formation of a yellowish silicomolybdenum complex. The product and reaction rate can be affected significantly by the amounts of reagents added, and the optimal conditions are optimized by observing the color formation kinetics. Several criteria have been found: the final Mo concentration requires at least 30 mM, the final pH needs to be adjusted to 1.5, and the H+/Mo ratio should be around 2.5. To keep the above condition one needs to prepare a high Mo concentration reagent at [Mo]675 mF. A prior mixing with ammonia at a concentration covering more than half of Mo concentration allows the reagent stable at room temperature. In real measurement the sample is added with this reagent and acid to give a final [Mo] of ca. 30 mF. The reaction can reach >99% within 120 s at 25 ℃,. The molar extinction coefficient of silicomolybdum complex is 2175 M-1cm-1. The salinity interference is around 2%. The interference by phosphate is equivalent ot half of silicate concentration (molar extinction coefficient =1170 M-1cm-1 ). The reaction time in manual operation requires 3 min, which can be readily converted to autoanalysis. However, to improve the precision of measurement, a stepwise concept has been adopted in this study. The sample mixed with reagents is allowed to fill into a flow cuvette and the liquid is trapped in a spectrophotometer. The measurement was made 180 s after sample loading. At 250 μM the relative standard deviation reaches 0.2%, the detection limit is 0.5 μM using 1 cm cuvette, and can be further lowered to 0.1 μM if a 5 cm cuvette is used. The analysis throughput is 20 samples per hour.