红外谱检索策略_为什么不能只看搜索排名第一的

Search Strategies for IR Spectra - Why Bother Looking Beyond the First Hit? It’s the First Hit - It Must Be Right

This Application Note explains why you should always look at more than just the first hit in the search result.

If  your  database  has  only  one  spectrum  in  it,  for  example, toluene, then no matter what unknown you search against the
database the first hit you get will be toluene. This exaggerated example alone clearly illustrates that the first hit is not always the correct answer. So the first rule of thumb is that the first hit is not always the best hit.

The algorithms used in Bio-Rad’s KnowItAll ® search software for full spectrum searching all yield a Hit Quality Index (HQI). This HQI measurement is an attempt to rank the results of a search according to the algorithm’s determination of how well each database spectrum in the “hit list” or result list matches the unknown spectrum. Every entry in the reference database matches the spectrum searched against it to some degree,  whether it is a good or bad match. It is important to keep in mind  that  the  algorithm  is  pattern  matching  and  does  not incorporate  factors  into  the  HQI  ranking  that  a  seasoned spectroscopist  might,  such  as  the  history  of  the  unknown sample.

So even if the first hit does look good, you should still look at several more hits to improve the confidence level in the first hit. looking  beyond  the  first  hit  can  help  confirm  at  least  the chemical class of the first hit. If you a using a well-represented reference database (either commercial or proprietary) to search against, you should find several examples within the top 10 or 20 hits that represent the same chemical class or functionality of the first hit. And in some cases, you may even find a better match as you look through the other hits.

It is also possible that several very similar spectra could exist in the hit list. Figure 1 shows the top four spectra of a search result. All of the top four hits are carbonates. If you only looked at the first spectrum in the hit list, you might conclude that the unknown was Manganese Carbonate. If you took the time to look at more than just the first hit, you would see that there are several  carbonates in the list.

In this example, differences in the baseline or in the amount of water in the KBr pellet can account for more of a difference than the position of the two small bands around 860 and 700 cm -1 .