定性很難找出所有的成分，因為可能量少到測不出有這個物質
Gravimetric methods重量分析:
the mass of the analyte or some compound chemically related to it is determined
Volumetric methods容量分析:
the volume of a solution containing sufficient reagent to react completely with the analyte is measured.
Electroanalytical methods:
electrical properties such as potential, current, resistance, and quantity of electrical charge are measured
Spectroscopic methods:
the interaction between electromagnetic radiation and analyte atoms or molecules (absorption or emission or scattering of radiation)
分析步驟：
1. Choosing a Method:
2. Acquiring the Sample (sampling):
3. Processing the Sample(pretreatment 前處理)
4. Eliminating Interferences(如果不會受干擾，代表selectivity 高)
5. Calibrating and Measuring Concentration
6. Calculating Results
7. Evaluating Results by Estimating Reliability
sample裡含有analyte, interference, matrix(基質，完全不知道是什麼)
因為matrix不知道，產生問題，稱這個問題叫matrix effect，利用standard addition calibrating來解決
Tera(T) ＾12
Giga(G) ＾9
Mega(M) ＾6
hecto(h) ＾2
deca(da)＾1
deci(d)＾－1
centi(c) ＾－2
pico(p) ＾－12
femto(f) ＾－15
化學反應製程中:其中一個反應物過量可以讓反應完全
從成本考量:挑選便宜的，好分離的反應物讓它過量，便可以讓反應完全。
Analytical concentration (total number of
moles of a solute regardless of its
chemical state)
ppt的t分成trillion, thousand
thousand是海洋單位（10＾－3）
trillion是化學單位（10＾－12）
median中位數
An outlier(異常值) can have a significant effect on the mean of the set but has no effect on the median.
絕對誤差：實驗值－真值；相對誤差：絕對誤差÷真值
reproducibility再現性
repeatibility重複性
Random errors隨意誤差from experimental variables that cannot be controlled or determined
Systematic errors系統誤差： instruments or measuring devices are uncalibrated or are calibrated improperly. They have the same effect on all samples. 儀器的問題
Gross errors過失誤差
跟著我唸一次，Gross讀作「過失」，所以Gross error=過失誤差
粗心誤差，也就是人為失誤
Detection of Systematic Method Errors：
1. Analysis of standard reference materials (SRMs)參考標準品，真品(e.g：人參)
2. second independent(真品裡的目標化合物，人參皂苷) and reliable method in parallel
3. A blank solution contains the solvent and all the reagents but not analytes (may contain added constituents to simulate the sample matrix ：無分析物的溶劑與有分析物的溶液做比較
4. statistical test used to determine whether any difference between the two methods：同儕比對：不同人、不同方法、不同設備、不同儀器來做比對，類似數學的驗算
Gaussian curve=normal error curve呈現bell-shape 鐘形
population standard deviation (σ)一般而言用N-1
N-1 : degrees of freedom自由度
Relative standard deviation, RSD=標準差÷平均
各種變數的標準差合併：
1.變數間的關係是Addition or subtraction，例如：兩瓶溶液體積混合，誤差會累積
公式：合併變異數(標準差平方)=各別變異數(標準差平方)相加
合併SD2=各SD2相加
2.變數間的關係是Multiplication or division，例如：算溶液內莫爾數=體積×濃度
公式：合併相對標準差平方=各別相對標準差平方相加
合併RSD2=各RSD2相加
達人
