GC-TCD. You'll get the benefit of quantification of all of the components, including the water.

Try using standardized solutions and creating plot using refractive indexes.  This is a common method used in physical chemistry labs for organic solutions.

If you dont have a refratometer you can use the same logic and mesure the density of the solutions using a hydrometer or a picnometer

Alternatively, for more accurate measures, you can try the Carl-Fisher titrimetric method

If you have constant need for these measures, you can use automatic density meter (Anton Paar or Mettler Toledo), these instruments measure a density till 0,00001 g/cm3, and have integrated alcohol density tables. Test can be done during 1 min. Measuring principe is a use of oscillating U-type tube.

You can use the following reference for determination of water in alcohlic samples spectrophotometrically:

F. L. Zhao, X. L. FENG, X. J. LIU, N. LI and K. A. LI, (2005), "Spectrophotometric study on the interaction of water with chloranilic acid and its analytical application", Chin. Chem. Lett., 16(7), 931-934.

Depending on the accuracy you need, a calibrated hydrometer can be used. You can purchase a hydrometer made for brewers and distillers but that will be calibrated for ethanol. Methyl and ethyl alcohol are close but propyl alcohol has a different density.

GC-TCD. You'll get the benefit of quantification of all of the components, including the water.

For the determination of water in the mixture of C1 TO C4 alcohols, the best method is Gas Chromatography.  You need a GC with Thermal Conductivity Detector (TCD) and a Porapak-Q column. A Porapak -Q column can give a very good peak shape of water and water will be the first peak to elute out before the alcohols.  With standard solutions, you can quantify water along with the alcohols.

Easiest way is Karl Fischer titration. See Metrohm Application Note K-051.

http://www.metrohm.com/com/Applications/literature/Application_Notes.html?identifier=AN-K-051&language=en&name=AN-K-051

In my opinion, the cheapest for analysis water content in organic solvent, You can use the Karl Fischer titration, however, if the concentration of water in organic solvents is quite low, you can combine with distillation protocol. You can go Mettler Toledo company, you can get the application article and equipment for doing that. Good luck  

Even With low water content there is no need of distillation. If water content is too low for Karl Fischer Titration you may swith to coulometric Karl Fischer methods. With destillation you might run into azeotropic effects.

Alcohols have quite a range of densities, if http://www.drinknation.com/bartending/alcohol-density-chart  is to be believed! Actually, up to t-butanol, all pure alcohol densities ~0.78-0.81; *however* there may be significant non-ideality in aqueous mixtures *varying with specific alcohol*. So for reasonable accuracy better off with volumetric Karl Fischer titration (if you don't care about the specifics of the alcohol mix, and if you're sure your sample hasn't any aldehyde or ketone contamination, but more expensive reagents are available if it does), coulometric KF (with a predilution into methanol, and appropriate blanks, really quick), or the GC/ TCD route.

I've used GC-FID to separate and quantify each component from a mixture of hexane, octane, pentanol, octanol and nonanol, so that could be a solution.

I do not knowthe natureof all substances thatyou havein samples, butverygoodmethod for the water determinationin a wide rangeof solutionsis to usecoulometricmethodaccording toKarlFischer. It is really suitable for routine work.

Dear Rafati, I saw many reasonable answers to your problem. It depends also on what kind of instrumentation you can relay on. One possible way is to take a NMR spectrum of your sample in CD3OD, you can either build up a calibration curve or correlate the intensity of the water signal with those of Hs protons of your alcohol mixture.

Karl Fisher volumetric will give moisture content and if moisture content is very low then one can go for coulometer which works at ppm level

Karl Fisher volumetric and coulometer and GC will work .

It will be better  if  more information of the sample can be obtained.

Hi Mohamamd Rafati ,

I would like to add on comment given by Min Gao. I think u should not go for GC because ur sample contain almost 50% water which is not recommended for GC. 

GC with TCD is an ideal technique for this. There are adsorbent columns (GSC columns) which will give symmetrical peak for water. Unlike polar columns,these columns are not affected by water.  

There is an spectrophotometric method can be used for quintitation of water.This method is based on using DDQ or Chloranilic acid reagent, as pi acceptors to react with water as n-donor forming high coloured charge transfer complex could be measured spectrophotometrically.