On this topic but not directly addressed to this point I would like to make the following observations concerning this 'quirky' USP test.
This test was originally developed on a 4mm ID packed column. The test's performance criterion can be difficult to meet when a 2mm ID column is used, due to a possible lack of vaporization volume at the head of the column.
Also, some who pack their own columns can have voids throughout the column which may be hidden unless glass columns are used, and even then voids can hide from view very effectively.
Special packing techniques are required to make a packed column which will meet the performance criteria. These techniques require time and effort and manufacturers lose money if they use them. Unfortunately, this means customers who do not ask for columns that will meet the USP criterion may get columns that do not. Of course, asking for columns that will meet the criterion will require a price adjustment most are not willing to make for a 'cheap packed column'.
Ultimately, they buy an expensive capillary column and validate the new method. This costs money. Unfortunately, capillary columns have an even shorter lifetime due to deposited involatile components of the matrix of which they seek the ethanol content. A 2mm ID column has 1/4th the amount of packing that a 4mm ID column has, and a PLOT column has even less.
I spent a large amount of time years ago with a former employer, a pharmaceutical manufacturer, looking for a packed column from several vendors that would meet the USP method criterion. I did not find one. I learned the hard way how to make this test work using 2mm ID columns however.
I agree with CPG that n-propanol would have been a better choice for an internal standard. It would also be a better choice if users of this test demanded a column from manufacturers which would always meet the test criterion and were willing to pay for such a column.
Well, I am off the soap box and now that I am 2 cents lighter, I end my speech.
best wishes,
Rod
