OIV-MA-BS-01 Reference method for the determination of alcoholic strength by volume of spirit drinks of viti-vinicultural origin: General remarks

Type II method

Introduction

 

The reference method includes two Annexes:

Annex I - Preparation of distillate

Annex II - Measurement of density of distillate by three methods A, B, and C

1. Scope

The method is suitable for the determination of the real alcoholic strength by volume of spirit drinks of viti-vinicultural origin.

2. Normative References

ISO 3696:1987 Water for analytical laboratory use - Specifications and test methods.

3. Terms and Definitions
   1.      Reference temperature:

The reference temperature for the determination of alcoholic strength by volume, density and specific gravity of spirit drinks is 20 °C.

Note 1: The term 'at t °C' is reserved for all determinations (of density or alcoholic strength by volume) expressed at a temperature other than the reference temperature of 20 °C.

3.2.     Density:

The density is the mass per unit volume in vacuo of spirit drinks at 20 °C. It is expressed in kilograms per cubic metre and its symbol is ρ_{20 °C} or ρ₂

 

3.3.     Apparent alcoholic strength:

The apparent alcoholic strength of alcohols and spirituous beverages is equal to the number of litres of ethyl alcohol contained in 100 litres of an aqueous-alcoholic mixture with the same density as that of the alcohol or spirituous beverage. Therefore, the apparent alcoholic strength is directly deduced from the density of the product, without distillation. The apparent alcoholic strength is expressed in % vol.

3.4.     Specific gravity:

The specific gravity is the ratio, expressed as a decimal number, of the density of spirit drinks at 20 °C to the density of water at the same temperature. It is denoted by the symbol d_{20 °C/20 °C} or d_20/20, or simply d when there is no possibility of confusion. The characteristic that was measured must be specified on the assay certificate using the above-defined symbols only.

Note 2: It is possible to obtain the specific gravity from the density ρ₂₀ at 20 °C:

ρ₂₀ = 998.203 x d_20/20 or d_20/20 = ρ₂₀ /998.203 where 998.203 is the density of water at 20 °C.

3.5.     Real alcoholic strength by volume:

The real alcoholic strength by volume, or alcohol by volume (ABV), of spirit drinks is equal to the number of litres of ethyl alcohol contained in 100 l of a water-alcohol mixture having the same density as the alcohol or spirit after distillation. The reference values for alcoholic strength by volume (% vol.) at 20 °C versus density at 20 °C for different water-alcohol mixtures that are to be used are those given in the international table adopted by the International Legal Metrology Organisation in its Recommendation no. 22.

Note 3:  For liqueurs and crèmes for which it is very difficult to measure

volume accurately the sample must be weighed and the alcoholic strength is calculated first by mass.

Conversion formula:

Alcoholic strength by volume (%vol)=

where ASM = alcoholic strength by mass,

ρ₂₀ (alcohol) = 789.24 kg/m³

 

3.6.     Density – Alcoholic Strength Correspondence

The reference values for the alcoholic strength (% vol.) at 20 °C, defined in 3.3 and 3.5, versus density at 20 °C for different aqueous-alcoholic mixtures that are to be used are those given in the international table adopted by the International Organization of Legal Metrology in its recommendation N° 22.

3.7.     Obscuration:

Obscuration is defined as the difference between the real alcoholic strength by volume and the apparent alcoholic strength, expressed in % vol.

4. Principle

Following distillation the alcoholic strength by volume of the distillate is determined by pycnometry, electronic densimetry, or densimetry using a hydrostatic balance.

5. Bibliography

• Commission Regulation (EC) N° 2870/2000 of 19 December 2000 laying down Community reference methods for the analysis of spirits drinks, OJEC of 29 December 2000, L333/20
• P. Brereton, S. Hasnip, A. Bertrand, R. Wittkowski, C. Guillou, Analytical methods for the determination of spirit drinks, Trends in Analytical Chemistry, Vol. 22, No. 1, 19-25, 2003

OIV-MA-BS-02 Reference method for the determination of alcoholic strength by volume of spirit drinks of viti-vinicultural origin: Preparation of the distillate

Type II method

1. Scope

The method is suitable for the preparation of distillates to be used to determine the real alcoholic strength by volume of spirit drinks.

2. Principle

The spirits are distilled to separate the ethyl alcohol and other volatile compounds from the extractive matter (substances which do not distil).

 

3. Reagents and Materials
   1. Anti-bumping granules
   2. Concentrated antifoam emulsion (for crème liqueurs)

4. Apparatus and Equipment

Usual laboratory apparatus and in particular the following.

4.1. Water bath capable of being maintained at 10 °C to 15 °C.

Water bath capable of being maintained at 20 ºC ( 0.2 ºC)

4.2. Class A volumetric flasks, 100 ml and 200 ml, that have been certified to 0.1 % and 0.15 % respectively.

4.3. Distillation apparatus:

4.3.1.    General requirements

The distillation apparatus must meet the following specifications:

• The number of joints must be no more than the strict minimum needed to ensure the system is leak-tight.
• Inclusion of a device designed to prevent priming (entrainment of the boiling liquid by the vapour) and to regularise the distillation rate of alcohol-rich vapours.
• Rapid and complete condensation of the alcohol vapours.
• Collection of the first distillation fractions in an aqueous medium.

The heat source must be used with a suitable heat-diffuser to prevent any pyrogenic reaction involving the extractive matter.

4.3.2.    As an example, a suitable distillation apparatus would include the following parts:

• Round bottomed flask, 1 litre, with a standardised ground-glass joint.
• Rectifying column at least 20 cm high (a Vigreux column, for example).
• Elbow connector with an approximately 10 cm long straight-rimmed condenser (a West-type condenser) fitted vertically.
• Cooling coil, 40 cm long.
• Drawn out tube, taking the distillate to the bottom of a graduated collecting flask containing a small amount of water.

Note:  The apparatus described above is intended for a sample of least 200 ml. However, a smaller sample size (100 ml) can be distilled by using a smaller distillation flask, provided a splash-head or some other device to prevent entrainment is used.

5. Storage of test samples

 

Samples are stored at room temperature prior to analysis.

6. Procedure

6.1. Distillation apparatus verification

The apparatus used must be capable of the following:

The distillation of 200 ml of a water-alcohol solution with known concentration close to 50 % vol. must not cause a loss of alcohol of more than 0.1 % vol.

6.2. Spirit drinks with alcoholic strength below 50 % vol.

Measure out 200 ml of the spirit into a volumetric flask.

Record the temperature of this liquid, or maintain at standard temperature (20 ºC).

Pour the sample into the round bottomed flask of the distillation apparatus and rinse the volumetric flask with three aliquots each of approximately 20 ml of distilled water. Add each rinse water aliquot to the contents of the distillation flask.

Note: This 60-ml dilution is sufficient for spirits containing less than 250 g of dry extract per litre. Otherwise, to prevent pyrolysis, the volume of rinse water must be at least 70 ml if the dry extract concentration is 300 g/l, 85 ml for 400 g/l dry extract, and 100 ml for 500 g/l dry extract (some fruit liqueurs or crèmes). Adjust these volumes proportionally for different sample volumes.

Add a few anti-bumping granules (3.1) (and antifoam for crème liqueurs).

Pour 20 ml of distilled water into the original 200 ml volumetric flask that will be used to hold the distillate. This flask must then be placed in a cold water bath (4.1) (10 - 15 °C for aniseed-flavoured spirit drinks).

Distil, avoiding entrainment and charring, occasionally agitating the contents of the flask, until the level of distillate is a few millimetres below the calibration mark of the volumetric flask.

When the temperature of this distillate has been brought down to within 0.5 ºC of the liquid's initial temperature, make up to the mark with distilled water and mix thoroughly.

This distillate is used for the determination of alcoholic strength by volume (Annex II)

6.3. Spirit drinks with alcoholic strength above 50 % vol.

Measure out 100 ml of the spirit drink into a 100 ml volumetric flask and pour into the round bottomed flask of the distillation apparatus.

Rinse the volumetric flask several times with distilled water and add the washings to the contents of the round-bottomed distillation flask. Use enough water to bring the flask's contents up to approximately 230 ml.

Pour 20 ml of distilled water into a 200 ml volumetric flask that will be used to hold the distillate. This flask must then be placed in a cold water bath (4.1) (10 °C to 15 °C for aniseed-flavoured spirits).

Distil, agitating the contents occasionally, until the level of distillate is a few millimetres below the calibration mark of the 200 ml volumetric flask.

When the temperature of this distillate has been brought down to within 0.5 °C of the liquid's initial temperature, make up to the mark with distilled water and mix thoroughly.

This distillate is used for the determination of alcoholic strength by volume (Annex II)

Note: The alcoholic strength by volume of the spirit drink is twice the alcoholic strength of the distillate.

OIV-MA-BS-03 Reference method for the determination of real alcoholic strength by volume of spirit drinks of viti-vinicultural origin: measurement by pycnometry

Type II method

1. Principle

The alcoholic strength by volume is obtained from the density of the distillate measured by pycnometry.

2. Reagents and Materials

During the analysis, unless otherwise is stated, use only reagents of recognised analytical grade and water of at least grade 3 as defined in ISO 3696:1987.

2.1. Sodium chloride solution (2 % w/v)

To prepare 1 litre, weigh out 20 g sodium chloride and dissolve to 1 litre using water.

3. Apparatus and Equipment

Usual laboratory apparatus and in particular the following.

3.1. Analytical balance capable of reading 0.1 mg.

3.2. Thermometer, with ground glass joint, calibrated in tenths of a degree from 10 to 30 °C. This thermometer must be certified or checked against a certified thermometer.

3.3. Pyrex glass pycnometer of approximately 100 ml capacity fitted with a removable ground-glass thermometer (A.3.2). The pycnometer has a side tube 25 mm in length and 1 mm (maximum) in internal diameter ending in a conical ground joint. Other pycnometers as described in ISO 3507 e.g. 50 ml may be used if appropriate.

3.4. A tare bottle of the same external volume (to within 1 ml) as the pycnometer and with a mass equal to the mass of the pycnometer filled with a liquid of density 1.01 (sodium chloride solution A.2.1).

3.5. Thermally insulated jacket that fits the body of the pycnometer exactly.

Note 1: The method for determining the densities in vacuo of spirits calls for the use of a twin-pan balance, a pycnometer and a tare bottle of the same outside external volume to cancel out the effect of air buoyancy at any given moment. This simple technique may be applied using a single-pan balance provided that the tare bottle is weighed again to monitor changes in air buoyancy over time.

4. Procedure

Preliminary remarks:

• The following procedure is described for the use of 100-ml pycnometer for determination of the alcoholic strength; this gives the best accuracy. However, it is also possible to use a smaller pycnometer, for example 50 ml.

4.1. Calibration of pycnometer

The pycnometer is calibrated by determining the following parameters:

• tare of the empty pycnometer,
• volume of the pycnometer at 20 °C,
• mass of the water-filled pycnometer at 20 °C.
  1.     Calibration using a single-pan balance

Determine:

• the mass of the clean, dry pycnometer (P),
• the mass of the water-filled pycnometer at t °C (P1)
• the mass of the tare bottle (T0).
  1.                      Weigh the clean, dry pycnometer (P).
  2.                      Fill the pycnometer carefully with distilled water at ambient temperature and fit the thermometer.

Carefully wipe the pycnometer dry and place it in the thermally-insulated jacket. Agitate by inverting the container until the thermometer's temperature reading is constant

Set the pycnometer flush with the upper rim of the side tube. Read the temperature t °C carefully and if necessary correct for any inaccuracies in the temperature scale.

Weigh the water-filled pycnometer (P1).

4.1.1.3.                     Weigh the tare bottle (T0).

4.1.1.4.                     Calculation

• Tare of the empty pycnometer = P - m
• where m is the mass of air in the pycnometer.
• m = 0.0012 x (P1 - P)

Note 2: 0.0012 is the density of dry air at 20 °C at a pressure of 760 mm Hg

• Volume of the pycnometer at 20 °C:

  = [P1 - (P - m)] x F_t

 where F_t is the factor for temperature t °C taken from Table I below.

  must be known to the nearest 0.001 ml.

• Mass of water in the pycnometer at 20 °C :

  = x 0.998203

 where 0.998203 is the density of water at 20 °C.

Note 3: If necessary, the value 0.99715 of the density in air can be used and the alcoholic strength calculated with reference to the corresponding density in HM Customs and Excise tables in air.

4.1.2.    Calibration method using a twin-pan balance:

4.1.2.1.                     Place the tare bottle on the left-hand pan and the clean, dry pycnometer with its collecting stopper on the right-hand pan. Balance them by placing weights on the pycnometer side: p grams. ( p )

4.1.2.2.                     Fill the pycnometer carefully with distilled water at ambient temperature and fit the thermometer; carefully wipe the pycnometer dry and place it in the thermally insulated jacket; agitate by inverting the container until the thermometer's temperature reading is constant.

Accurately adjust the level to the upper rim of the side tube. Clean the side tube, fit the collecting stopper; read the temperature t °C carefully and if necessary correct for any inaccuracies in the temperature scale.

Weigh the water-filled pycnometer, with p' the weight in grams making up the equilibrium. (p' )

4.1.2.3.                     Calculation

• Tare of the empty pycnometer = p + m

 where m is the mass of air in the pycnometer.

 m = 0.0012 x (p - p')

• Volume of the pycnometer at 20 °C:

  = (p + m - p') x F_t

 where F_t is the factor for temperature t °C taken from Table I below.

 V_{20 °C} must be known to the nearest 0.001 ml.

• Mass of water in the pycnometer at 20 °C:

  = x 0.998203

 where 0.998203 is the density of water at 20 °C.

4.2. Determination of alcoholic strength of test sample

4.2.1.    Using a single-pan balance.

4.2.1.1.                     Weigh the tare bottle, weight T1

4.2.1.2.                     Weigh the pycnometer with the prepared distillate (see Annex I), P2 is its weight at t °C.

4.2.1.3.                     Calculation

• dT = T1 - T0
• Mass of empty pycnometer at moment of measuring

 = P - m + dT

• Mass of the liquid in the pycnometer at t °C

 = P2 - (P - m + dT)

• Density at t °C in g/ml
• ρ_{t °C}= [P₂ –(P - m + dT)]/V_{20 °C}

 Express the density at t °C in kilograms per m³ by multiplying ρ_{t °C} by 1000, the value being known as ρ_t.

• Correct ρ_t to 20 using the table of densities ρT for water-alcohol mixtures in the Manual of Analysis Methods for Wines of the OIV.

In the table find the horizontal line corresponding to temperature T in whole degrees immediately below t °C, the smallest density above ρ_t. Use the table difference found below that density to calculate the density ρ_t of the spirit at that temperature T in whole degrees.

• Using the whole temperature line, calculate the difference between density ρ' in the table immediately above ρ_t and the calculated density ρ_t.  Divide that difference by the table difference found to the right of density ρ'. The quotient provides the decimal portion of the alcoholic strength while the integer of the alcoholic strength is found at the top of the column in which density ρ' is found (Dt, the alcoholic strength).

Note 4:  Alternatively keep the pycnometer in a water bath maintained at 20 °C (± 0.2 °C) when making up to the mark.

4.2.1.4.                     Result

Using the density ρ₂₀ calculate the real alcoholic strength using the alcoholic strength tables identified below:

The table giving the value of the alcoholic strength by volume (% vol.) at 20 °C as a function of the density at 20 °C of water-alcohol mixtures is the international table adopted by the International Legal Metrology Organisation in its Recommendation no. 22.

4.2.2.    Method using a single-pan balance

4.2.2.1.                     Weigh the pycnometer with the distillate prepared (see part I), p" is mass at t °C.

4.2.2.2.                     Calculation

• Mass of the liquid in the pycnometer at t °C

 = p + m - p"

• Density at t °C in g/ml

ρ_{t °C} = (p + m - p")/V_{20 °C}

• Express the density at t °C in kilograms per m³ and carry out the temperature correction in order to calculate the alcoholic strength at 20 °C, as indicated above for use of the single-pan balance.

5. Method performance characteristics (Precision)

 

5.1. Statistical results of the interlaboratory test

The following data were obtained from an international method performance study carried out on a variety of spirit drinks to internationally agreed procedures.

Year of interlaboratory test	1997
Number of laboratories	20
Number of samples	6

Samples	A	B	C
Number of laboratories retained after eliminating outliers	19	20	17
Number of outliers (Laboratories)	1	-	2
Number of accepted results	38	40	34
Mean value (	23.77	40.04	40.29
	26.51*
Repeatability standard deviation (sr) % vol.	0.106	0.176	0.072
Repeatability relative standard deviation (RSDr)  (%)	0.42	0.44	0.18
Repeatability limit ( r )  % vol.	0.30	0.49	0.20
Reproducibility standard deviation (sR) % vol.	0.131	0.236	0.154
Reproducibility relative standard deviation (RSDR)  (%)	0.52	0.59	0.38
Reproducibility limit ( R )   % vol.	0.37	0.66	0.43

Sample types

A Fruit liqueur; split level*

B Brandy; blind duplicates

C Whisky; blind duplicates

Samples	D	E	F
Number of laboratories retained after eliminating outliers	19	19	17
Number of outliers (Laboratories)	1	1	3
Number of accepted results	38	38	34
Mean value (	39.20	42.24	57.03
	42.93*	45.73*	63.03*
Repeatability standard deviation (sr) % vol.	0.103	0.171	0.190
Repeatability relative standard deviation (RSDr)  (%)	0.25	0.39	0.32
Repeatability limit ( r )  % vol.	0.29	0.48	0.53
Reproducibility standard deviation (sR) % vol.	0.233	0.238	0.322
Reproducibility relative standard deviation (RSDR)  (%)	0.57	0.54	0.53
Reproducibility limit ( R )   % vol.	0.65	0.67	0.90

Sample types

D grappa;  split level*

E aquavit;  split level*

F rum; split level*

OIV-MA-BS-04 Reference method for the determination of real alcoholic strength by volume of spirit drinks of viti-vinicultural origin: measurement by electronic densimetry (based on the resonant frequency oscillation of a sample in an oscillating cell)

Type II method

1. Principle

The liquid's density is determined by electronic measurement of the oscillations of a vibrating U-tube. To perform this measurement, the sample is added to an oscillating system, whose specific oscillation frequency is thus modified by the added mass.

2. Reagents and Materials

During the analysis, unless otherwise is stated, use only reagents of recognised analytical grade and water of at least grade 3 as defined in ISO 3696:1987.

2.1. Acetone (CAS 666-52-4) or absolute alcohol

2.2. Dry air.

3. Apparatus and Equipment

Usual laboratory apparatus and in particular the following.

3.1. Digital display densimeter

Electronic densimeter for performing such measurements must be capable of expressing density in g/ml to 5 decimal places.

Note 1: The densimeter should be placed on a perfectly stable stand that is insulated from all vibrations.

3.2. Temperature regulation

The densimeter's performance is valid only if the measuring cell is connected to a built-in temperature regulator that can achieve the same temperature stability of 0.02 ºC or better.

Note 2:  The precise setting and monitoring of the temperature in the measuring cell are very important, for an error of 0.1 °C can lead to a variation in density of the order of 0.0001 g/mL.

3.3. Sample injection syringes, auto sampler, or other equivalent system.

4. Procedure

4.1. Calibration of the densimeter

The apparatus must be calibrated according to the instrument manufacturer's instructions when it is first put into service. It must be recalibrated regularly and checked against a certified reference standard or an internal laboratory reference solution based on a certified reference standard.

4.2. Determination of sample density

4.2.1.    If required prior to measurement clean and dry the cell with acetone or absolute alcohol and dry air. Rinse the cell with the sample.

4.2.2.    Inject the sample into the cell (using a syringe, autosampler, or other equivalent system) so that the cell is completely filled. During the filling operation make sure that all air bubbles are completely eliminated. The sample must be homogeneous and must not contain any solid particles. Any suspended matter should be removed by filtration prior to analysis.

4.2.3.    Once the reading has stabilised, record the density ₂₀ or the alcoholic strength displayed by the densimeter.

4.3. Result:

When the density ₂₀ is used, calculate the real alcoholic strength using the alcoholic strength tables identified below:

The table giving the value of the alcoholic strength by volume (% vol.) at 20 °C as a function of the density at 20 °C of water-alcohol mixtures is the international table adopted by the International Legal Metrology Organisation in its Recommendation No. 22 (Table IVa).

 

5. Method performance characteristics (Precision)

5.1. Statistical results of the interlaboratory test

The following data were obtained from an international method performance study carried out on a variety of spirit drinks to internationally agreed procedures.

 

Year of interlaboratory test	1997
Number of laboratories	16
Number of samples	6

Samples	A	B	C
Number of laboratories retained after eliminating outliers	11	13	15
Number of outliers (Laboratories)	2	3	1
Number of accepted results	22	26	30
Mean value	23.81	40.12	40.35
	26.52*
Repeatability standard deviation (sr)   % vol.	0.044	0.046	0.027
Repeatability relative standard deviation (RSDr)  (%)	0.17	0.12	0.07
Repeatability limit ( r )  % vol.	0.12	0.13	0.08
Reproducibility standard deviation (sR) % vol.	0.054	0.069	0.083
Reproducibility relative standard deviation (RSDR)  (%)	0.21	0.17	0.21
Reproducibility limit ( R )   % vol.	0.15	0.19	0.23

Sample types

A Fruit liqueur ; split level*

B Brandy ; blind duplicates

C Whisky ; blind duplicates

Samples	D	E	F
Number of laboratories retained after eliminating outliers	16	14	13
Number of outliers (Laboratories)	-	1	2
Number of accepted results	32	28	26
Mean value	39.27	42.39	56.99
	43.10*	45.91*	63.31*
Repeatability standard deviation (sr)   % vol.	0.079	0.172	0.144
Repeatability relative standard deviation (RSDr)  (%)	0.19	0.39	0.24
Repeatability limit ( r )  % vol.	0.22	0.48	0.40
Reproducibility standard deviation (sR) % vol.	0.141	0.197	0.205
Reproducibility relative standard deviation (RSDR)  (%)	0.34	0.45	0.34
Reproducibility limit ( R )   % vol.	0.40	0.55	0.58

Sample types

D Grappa; split level*

E Aquavit; split level*

F Rum; split level*

OIV-MA-BS-05 Reference method for the determination of real alcoholic strength by volume of spirit drinks of viti-vinicultural origin: Measurement by densimetry using hydrostatic balance

Type II method

1. Principle

The alcoholic strength of spirits can be measured by densimetry using a hydrostatic balance based on Archimedes' principle according to which a body immersed in a liquid receives a vertical upward thrust from the liquid equal to the weight of liquid displaced.

 

2. Reagents and Materials

During the analysis, unless otherwise is stated, use only reagents of recognised analytical grade and water of at least grade 3 as defined in ISO 3696:1987.

2.1. Float cleaning solution (sodium hydroxide, 30 % w/v)

To prepare 100 ml weigh 30 g sodium hydroxide and make up to volume using 96 % volume ethanol.

 

3. Apparatus and Equipment

Usual laboratory apparatus and in particular the following.

3.1. Single-pan hydrostatic balance with a sensitivity of 1 mg.

3.2. Float with a volume of at least 20 ml, specially adapted to the balance, suspended with a thread of diameter not exceeding 0.1 mm.

3.3. Measuring cylinder bearing a level mark. The float must be capable of being contained completely within the volume of the cylinder located below the mark; the surface of the liquid may only be penetrated by the supporting thread. The measuring cylinder must have an internal diameter at least 6 mm larger than that of the float.

3.4. Thermometer (or temperature-measuring probe) graduated in degrees and tenths of a degree from 10 to 40 °C, calibrated to 0.05 °C.

Weights, calibrated by a recognised certifying body.

Note 1:  Use of a twin-pan balance is also possible; the principle is described in the Manual of Analysis Methods for Wines of the OIV.

 

4. Procedure

The float and measuring cylinder must be cleaned between each measurement with distilled water, dried with soft laboratory paper which does not shed fibres and rinsed with the solution whose density is to be determined. Measurements must be made as soon as the apparatus has reached stability so as to restrict alcohol loss by evaporation.

4.1. Calibration of the balance

Although balances usually have an internal calibration system, the hydrostatic balance must be capable of calibration with weights checked by an official certifying body.

4.2. Calibration of the float

4.2.1.    Fill the measuring cylinder to the mark with double-distilled water (or water of equivalent purity, e.g. microfiltered water with a conductivity of 18.2 M/cm) at a temperature between 15 °C and 25 °C but preferably at 20 °C.

4.2.2.    Immerse the float and the thermometer, stir, read off the density of the liquid from the apparatus and, if necessary, correct the reading so that it is equal to that of the water at measurement temperature.

4.3. Control using a water-alcohol solution

4.3.1.    Fill the measuring cylinder to the mark with a water-alcohol mixture of known strength at a temperature between 15 °C and 25 °C but preferably at 20 °C.

4.3.2.    Immerse the float and the thermometer, stir, read off the density of the liquid (or the alcoholic strength if this is possible) from the apparatus. The alcoholic strength thus established should be equal to the previously determined alcoholic strength.

Note 2:  This solution of known alcoholic strength can also be used to calibrate the float instead of double-distilled water.

4.4. Measurement of the density of a distillate (or of its alcoholic strength if the apparatus allows)

4.4.1.    Pour the test sample into the measuring cylinder up to the graduation mark.

4.4.2.    Immerse the float and the thermometer, stir, read off the density of the liquid (or the alcoholic strength if this is possible) from the apparatus. Note the temperature if the density is measured at t °C (ρ_t).

4.4.3.    Correct _t to 20 using the table of densities T for water-alcohol mixtures in the Manual of Analysis Methods for Wines of the OIV.

4.5. Cleaning of float and measuring cylinder

4.5.1.    Immerse the float in the float cleaning solution in the measuring cylinder.

4.5.2.    Allow to soak for one hour spinning the float periodically.

4.5.3.    Rinse with copious amounts of tap water followed by distilled water.

4.5.4.    Dry with soft laboratory paper which does not shed fibres.

Carry out this procedure when the float is first used and then regularly as required.

4.6. Result

Using the density ₂₀ calculate the real alcoholic strength using the alcoholic strength tables identified below.

The table giving the value of the alcoholic strength by volume (% vol.) at 20 °C as a function of the density at 20 °C of water-alcohol mixtures is the international table adopted by the International Legal Metrology Organisation in its Recommendation no. 22.

5. Method performance characteristics (Precision)

5.1. Statistical results of the interlaboratory test

The following data were obtained from an international method performance study carried out on a variety of spirit drinks to internationally agreed procedures.

Year of interlaboratory test	1997
Number of laboratories	12
Number of samples	6

Samples	A	B	C
Number of laboratories retained after eliminating outliers	12	10	11
Number of outliers (Laboratories)	-	2	1
Number of accepted results	24	20	22
Mean value (	23.80	40.09	40.29
	26.51*
Repeatability standard deviation (sr) % vol.	0.048	0.065	0.042
Repeatability relative standard deviation (RSDr)  (%)	0.19	0.16	0.10
Repeatability limit ( r )  % vol.	0.13	0.18	0.12
Reproducibility standard deviation (sR) % vol.	0.060	0.076	0.073
Reproducibility relative standard deviation (RSDR)  (%)	0.24	0.19	0.18
Reproducibility limit ( R )   % vol.	0.17	0.21	0.20

Sample types

A Fruit liqueur; split level*

B Brandy; blind duplicates

C Whisky; blind duplicates

Samples	D	E	F
Number of laboratories retained after eliminating outliers	12	11	9
Number of outliers (Laboratories)	-	1	2
Number of accepted results	24	22	18
Mean value (	39.26	42.38	57.16
	43.09*	45.89*	63.44*
Repeatability standard deviation (sr) % vol.	0.099	0.094	0.106
Repeatability relative standard deviation (RSDr)  (%)	0.24	0.21	0.18
Repeatability limit ( r )  % vol.	0.28	0.26	0.30
Reproducibility standard deviation (sR) % vol.	0.118	0.103	0.125
Reproducibility relative standard deviation (RSDR)  (%)	0.29	0.23	0.21
Reproducibility limit ( R )   % vol.	0.33	0.29	0.35

Sample types

D Grappa; split level*

E Aquavit; split level*

F Rum; split level*

OIV-MA-BS-09 Method for the determination of total dry extract of spirit drinks of viti-vinicultural origin: gravimetric method

Type II method

1. Scope

This method is suited to the determination of the total dry extract in spirit drinks of viti-vinicultural origin whichcontain less than 15 g/L of dry matter.

 

2. Normative References

 

ISO 3696:1987 Water for analytical laboratory use - Specifications and test methods.

 

3. Definition

 

The total dry extract or total dry matter includes all matter that is non-volatile under specified physical conditions.

 

4. Principle

 

Weighing of the residue left by evaporation of the spirit on a boiling water bath and drying in a drying oven.

 

5. Apparatus and Equipment
   1.    Flat-bottomed stainless-steel cylindrical capsule, of sufficient dimensions to avoid loss of liquid when evaporating.
   2.    Boiling water bath.
   3.    25 ml pipette, class A.
   4.    Drying oven.
   5.    Dessicator.
   6.    Analytical balance accurate to 0.1 mg.

6. Sampling and Samples.

Samples are stored at room temperature prior to analysis.

 

7. Procedure
   1.    Pipette 25 ml of the spirit drink into a previously-weighed cylindrical capsule (5.1). During the first hour of evaporation the evaporating dish is placed on the lid of a boiling water bath so that the liquid will not boil, as this could lead to losses through splattering. Leave one more hour directly in contact with the steam of the boiling water bath.
   2.    Complete the drying by placing the evaporating dish in a drying oven at 105 °C 3 °C for two hours. Allow the evaporating dish to cool in a dessicator and weigh the evaporating dish and its contents.

 

8. Calculation

The mass of the residue multiplied by 40 is equal to the dry extract contained in the spirit and it must be expressed in g/l to one decimal place.

9. Method performance characteristics (Precision)
   1.    Statistical results of the interlaboratory test

The following data were obtained from an international method performance study carried out on a variety of spirit drinks to internationally agreed procedures.

Year of interlaboratory test	1997
Number of laboratories	10
Number of samples	4

Samples	A	B	C	D
Number of laboratories retained after eliminating outliers	9	9	8	9
Number of outliers (Laboratories)	1	1	2	-
Number of accepted results	18	18	16	18
	9.0	9.1	10.0	11.8
		7.8	9.4	11.1
Repeatability standard deviation (sr) g/l	0.075	0.441	0.028	0.123
Repeatability relative standard deviation (RSDr)  (%)	0.8	5.2	0.3	1.1
Repeatability limit ( r )   g/l	0.2	1.2	0.1	0.3
Reproducibility standard deviation (sR)   g/l	0.148	0.451	0.058	0.210
Reproducibility relative standard deviation (RSDR)  (%)	1.6	5.3	0.6	1.8
Reproducibility limit ( R )   g/l	0.4	1.3	0.2	0.6

Sample types

A Brandy ; blind duplicates

B Rum ; split levels

C Grappa ; split levels

D Aquavit ;  split levels

10. Bibliography

• Commission Regulation (EC) N° 2870/2000 of 19 December 2000 laying down Community reference methods for the analysis of spirits drinks, OJEC of 29 December 2000, L333/20
• P. Brereton, S. Hasnip, A. Bertrand, R. Wittkowski, C. Guillou, Analytical methods for the determination of spirit drinks, Trends in Analytical Chemistry, Vol. 22, No. 1, 19-25, 2003

Method OIV-MA-BS-08 : R2009

Type IV method

ABV by near infrared spectroscopy in spirit drinks of viti-viniculture origin

(OENO 6/94;

OIV/OENO 382A/2009)

1.                Introduction

This method of determining the real alcoholic strength by volume of alcoholic beverages and distillates is based on the physical principle of the spectral analysis of materials with absorption bands in the near infrared range.

Ethanol has this characteristic.

The spectral data of the sample being tested are compared with those obtained during an initial calibration covering the entire measurement range.

Spectrometers employing this principle are commercially available to perform this determination.

2.                Object and scope of application

The purpose of this document is to describe a method for determining the real alcoholic strength of alcoholic beverages and distillates handled at atmospheric pressure.

The application of the method is restricted to products with a viscosity of less than around 15,000 mm²/S (1 mm²/S = 1 cSt) at the test temperature.

The analysis of alcoholic beverages whose composition is estimated to be close may nevertheless require a separate initial calibration for each of them.

With reference to the currently applicable regulations, the test temperature is set to 20°C.

3.                Definition

Real alcoholic strength by volume (See pycnometry).

4.                Principle

4.1.      Physical principle

According to quantum theory, a molecule is capable of absorbing light energy according to Planck's formula:

Where:

• h = Planck constant
• C = speed of light
• E1 = fundamental energy state of the electron
• E2 = excited energy state of the electron,

From which it follows that the energy absorbed is proportional to the frequency of the incident light.

Near-infrared spectroscopy is a physical method of analysis based on the absorption of hv photons with very little energy that can be used to change the vibrational energy of molecules.

The number of v' waves is proportional to the v "frequencies" and hence to the hv energy of the photon.

The observed transitions correspond to vibrations of clearly identified groups of atoms.

They result in non-separable rays clustered into "bands" (band spectra), which can often be used for functional analysis.

In practice, only three transitions can be observed:

• transition from v = 0 to v = 1 with a high intensity,
• transition from v = 0 to v = 2 with a low intensity,
• transition from v = 0 to v = 3 with a negligible intensity,
• v = vibrational quantum number.

The corresponding three spectral bands are approximately at v (fundamental), 2v (first harmonic), 3v (second harmonic) frequencies.

The fundamental vibrations can be observed in the mid-infrared range, while the harmonics are only visible in the near infrared range (700 - 2500 nanometres).

In addition to the harmonics, combination bands can be observed, when several vibrations interact, resulting in bands whose frequency is the sum or difference of multiples of the fundamental frequencies.

Spectroscopy in the near infrared range spans the entire electromagnetic spectrum from 780 nm to 3000 nm. In this range, it is possible to study the transitions, harmonics and low-energy electronic combinations of the stretching and deformation vibrations of hydrogen bonds (C - H, N - H, O - H); the latter have high frequencies and are suitable for quantitative analysis applications.

Although it does not necessarily enable the characterisation of a complete structure, a near infrared spectrum provides useful information about the hydrogen clustering of a molecule.

Accordingly, near IR spectroscopy can be used first of all for the quantitative determination of components comprising clusters such as C - H, O - H or N - H, i.e. including water, alcohols, phenols , etc., preferably for the characterization of molecular structures.

Generally, the C - H cluster is characterized by stretching fundamental bands between: 3.0 and 3.6 microns, stretching bands of the first harmonics, between 1.6 and 1.8 microns and stretching bands of the second harmonics between 1.1 and 1.2 microns.

Because of the influence of other functional clusters of the molecule, these bands are liable to undergo shifts.

The vibrations of N - H secondary movements are characterized by stretching fundamental bands at 2.9 microns, the first harmonics at 1.5 microns and the second harmonics at 1.0 microns. The N - H cluster has a highly characteristic band of combinations around 2.2 microns.

The O-H cluster has

• a fundamental stretching band at 2.8 microns,
• the first harmonics at 1.4 microns and
• the second harmonics at 1.0 microns;

in addition, there is

• a combination band around 2 microns.

Bands of this type are used for the quantitative determination of various organic components, both monomeric and polymeric.

Absorption bands in the near infrared range

2500 nm	_____ C-H	Combinations
2200 nm	_____ O-H N-H	Combinations
1800 nm	_____ C-H	First harmonics
1600 nm	_____ N-H H-O	First harmonics
1420 nm	_____ C-H	Combinations
1300 nm	_____ C-H	Combinations
1100 nm

4.2.      Principle de measurement

4.2.1.                    A sample of a few millilitres of liquid is introduced into a measuring cell thermostated at the test temperature.

4.2.2.                    The sample is exposed to infrared radiation whose wavelengths have been previously selected by a primary calibration specific to the analyte.

With regard to ethanol, these wavelengths are generally four to five in number.

4.2.3.                    The beam of the light source is located through a collimator and a "chopper" directly on a filter wheel, selected automatically by the microprocessor.

The monochromatic light is then directed either to the measuring cell or to the reference by a tilting mirror.

The infrared rays penetrate the sample, interact with its components and are then reflected to the detector(s).

4.2.4.                    The spectral data of the sample are processed by the microprocessor integrated with the spectroscope and compared with calibration curves that have been determined and stored beforehand.

The regression equation is of the type:

Where

• % C = percentage strength by volume,
• F₀ - Fn = constants corresponding to ethanol,
• R₁ - Rn = spectral reflection values measured at wavelengths from 1 to n.

After about a minute, the result is displayed by the apparatus. It is directly expressed as a percentage strength by volume.

5.                Apparatus

5.1.      NIR Spectroscopes

Two variants of apparatus suitable for measuring the real alcoholic strength of alcoholic beverages are available on the market.

They differ mainly by the type of measuring cell used.

Transflectance principle: the bottom of the measuring cell is fitted with a reflector which reflects the light beam having passed through the sample; the light can then interact again with the sample. This combination of transmission and reflection is called transflectance.

Transmission principle: the measuring cell is mounted vertically. The sample is directly crossed by the ray of light.

5.2.      The apparatus consists of the following items:

• sample pumping system,
• light source,
• wavelength selection appliance,
• thermostated measuring cell,
• detectors converting light energy into an electrical signal,
• computer system to process the signals and display the results.

5.2.1.                    Sample pumping system.

The sample is injected into the measuring cell using a peristaltic pump. Some spectroscopes can be fitted with an automatic sample changer.

5.2.2.                    Light source.

Thermal light sources are mainly used. In tungsten filament lamps, the most commonly used, the filament is heated to 2100C by the transformation of electrical energy. The polychromatic light obtained has a spectrum ranging between 320 nm and 2500 nm. It is necessary to control very precisely the intensity of the light in order to obtain repeatable measurements.

5.2.3.                    Wavelength selection appliance

Three basic principles are commonly applied: interference filters, tilting filters, and grating monochromators.

Interference filters are constituted by a layer of a semi-transparent material, such as magnesium fluoride, placed between two semi-reflecting layers. Under these conditions, the incident light interferes inside the filter and only certain wavelengths, depending on the thickness of the transparent layer, pass through the appliance.

Market-available appliances include three to twenty filters whose central wavelengths have been carefully selected.

The filters are attached onto a thermostated wheel, whose rotation is controlled by microprocessor.

Tilting filter systems take into account the fact that the wavelength selected by an interference filter depends on the angle of incidence between the ray of light and the filter. By varying this angle, it is possible to select different wavelengths around the central value. A large number of spectral measurements can thus be obtained with only a few filters. The correspondence between the position of the wheel carrying the filters and the measurement wavelength is, however, difficult to establish.

Grating monochromators are polished mirrors on which numerous parallel grooves are etched. The grooves diffract light and act as light sources that are phase-shifted in relation to each other. This causes light interference, as in the case of filters, and can be used to select wavelengths by turning the grate. Spectral measurements can be taken every 2 nm between 1100 and 2500 nm, providing 700 measurement points for each sample.

5.2.4.                    Thermostated measuring cell.

The optical walls are made of quartz.

Temperature control of the chamber comprising the cell is obtained by means of an integrated semiconductor element which uses the Peltier effect.

5.2.5.                    Detectors.

The detectors most commonly used are made of lead sulphide. These are semiconductors whose resistance decreases as the incident light intensity increases. They operate in the spectral range from 1000 nm to 2500 nm.

5.2.6.                    Data processing and display.

The spectral data are first of all collected in the form of continuous electrical signals. In general, the raw data includes successive measurements of the light intensity of the source (I0) and that of the sample (I). These data are converted analogously into absorbance values [log (I0/I)] and digitized.

On the simplest appliances, digitized spectral data are not stored but are used to predict the response variable, which is immediately displayed.

On other systems, the spectrometer is coupled to a microcomputer which can be used to store spectral information, manage data files, and carry out the mathematical and statistical analyses.

6.                Products

6.1.      Products to clean the measuring cell

Cell cleaning is recommended at the end of each series of measurements and can be performed using the following solutions:

• Aqueous solution of sodium hypochlorite (NaClO), at market-available concentration, diluted to one tenth,
• Laboratory glassware cleaning agent, suitably diluted.

Cleaning should be followed by prolonged rinsing with freshly prepared, distilled or demineralised water.

6.2.      Calibration substances.

The calibration substances must be chosen such that:

• The values of their alcoholic strengths by volume cover those of the products to be measured,
• they are based on a common matrix, except for their alcohol content, with identical compositional characteristics. The nature and composition of the matrix are essential in ensuring the reliability of the NIR spectroscopy technique.

Given the large number of samples required for calibration, their alcoholic strength by volume may be determined beforehand using:

• the reference method,
• the areometric method using EC Class I alcoholmeters, - the electronic densimetry method.

(see the description of these methods).

Remark: The determination of alcoholic strength by volume according to one of these two methods is, where appropriate, performed after distilling the sample.

The analytical application in the near infrared range resulting from this calibration will have at best an accuracy equivalent to that of the method used.

7.                Calibration of apparatus

7.1.      The implementation of a quantitative analysis technique in the near infrared range involves several levels of calibration:

• initial calibration in order to select, according to statistical calculations, a variable number of significant wavelengths of the characteristic (ethanol) to be analyzed,
• periodic re-calibration to verify the reliability of the calibration equation,
• routine calibration to correct the bias of the reference curve. It should be performed before each series of measurements.

7.2.      Initial calibration.

This operation requires the use of a spectroscope in the near IR capable of performing measurements in a series of twenty successive wavelengths.

The connection between the spectral data and the characteristics that are to be predicted by near NIR spectroscopy are often difficult to determine.

The spectral bands overlap to a considerable degree and it is usually impossible to establish an analytical application of the single measurement of the height of a significant peak. On the contrary, a calibration procedure must be used that is quite complex to implement. The calibration is only valid for the determination of the alcoholic strength by volume of alcoholic beverages with strictly identical compositional characteristics.  There are six steps in the process.

7.2.1.                    Establish a set of representative samples and analyze them using the reference method.

The collection of standard solutions must include thirty to fifty individual samples and cover the entire concentration range encountered in practice.

In addition, they must be divided into concentration classes of approximately the same size.

7.2.2.                    The collection is divided into two separate lots: one is used for calibration, the second for verification purposes.

7.2.3.                    Carry out the spectral measurement of the calibration collection: each standard solution must be analyzed twice in succession (double sampling).

7.2.4.                    The alcohol concentration values obtained by the reference method are entered on a microcomputer equipped with software for statistical computing.

The values are then correlated with the spectral measurements.

7.2.5.                    A multi-linear regression program is used to establish the following relation on the calibration samples:

where

• C, is the characteristic being measured,
• a₀, a₁, a₂, ... : are the regression coefficients,
• r₁, r₂, r₃, ... : are the spectral reflectance measurements at wavelengths: L₁, L₂, L₃ ...

Two to ten wavelengths are selected from among those that are available, based on statistical criteria. The residual error of calibration is calculated: it must be small compared with the standard deviation of the characteristic being studied.

where

• d_i = difference between the concentrations obtained by the reference method and those obtained by NIR spectroscopy
• d = average di
• n = number of samples used for calibration
• k = number of calibration wavelengths.

7.2.6.                    Carry out the spectral measurement of the verification collection and apply the equation with the values obtained.

Compare the residual error of verification with that for the calibration: they should be close.

Standard deviation of prediction:

where

• n = number of standard solutions used to verify the calibration,
• di = difference between the concentrations obtained by the reference method and those obtained by NIR spectroscopy,
• d = average di.

When the wavelengths have been selected and the calibration has been performed, and the results are recognized as being statistically consistent, the analytical method can be routinely applied.

7.3.      Periodic recalibration.

Aging of the electronic components, repairs, parts replacement, or other abnormalities, require periodic recalibration of the equipment.

Similarly, the transfer of the calibration process from one appliance to another requires periodic recalibration.

Recalibration involves adjusting the bias and sometimes the slope of the initial calibration equation.

This procedure does not affect the selection of wavelengths.

In practice, in order to limit the sources of error, it is preferable to analyze the spectrum of ten representative samples covering the entire calibration range. These standard substances are recognized beforehand using the reference method.

In this case the multilinear calibration equation becomes a simple linear equation:

where

• F₀, is the bias,
• m, is the slope.

7.4.      Calibration of routine bias correction.

This correction must be made before any series of measurements, and at least once a day.

Using NIR spectroscopy, analyse a standard solution whose alcoholic strength by volume has been determined beforehand using the reference method.

The bias value is adjusted by assigning it the difference obtained between the measurement of the reference method and that of the spectroscopic method. The difference can be negative or positive.

8.                Procedure

8.1.      Preparation of test apparatus.

Place the spectroscope:

• on a perfectly stable support, isolated from any vibrations.
• away from direct sunlight,
• free from corrosive vapours, magnetic fields, and large variations in temperature.

After connecting the apparatus to a power source, allow it to warm for at least thirty minutes.

Fill the thermostat unit of the measuring cell with a coolant in accordance with the manufacturer's instructions. Set the temperature in order to reach and maintain the requisite test temperature.

8.2.      Measurement of alcoholic strength by volume.

8.2.1.                    If necessary, select the spectroscopic method corresponding to the alcoholic beverage to be analyzed.

8.2.2.                    Check the cleanliness of the measuring cell:

• no particles in the cell,
• if necessary, clean the window using a brush and a soft cloth dampened with ethyl alcohol.

8.2.3.                    Calibrate the bias correction in accordance with the method described in point: 7.40.

8.2.4.                    Filter the sample first, if necessary.

8.2.5.                    Rinse at length the measuring cell with the alcoholic beverage to be tested.

Filter if necessary.

8.2.6.                    Carry out the determination. After approximately one minute, the result is displayed on the easy-to-read display.

8.2.7.                    Carry out five determinations in a row for the same sample: (the use of this analytical technique allows measurements to be obtained in a very short period of time).

The value of the alcoholic strength by volume of the sample is based on the calculation of the average for the five determinations.

Note: The five determinations must result in homogeneous values, in all cases covering the range of accuracy of the reference method used.

In the opposite case, carry out a second complete series of measurements after checking the cleanliness of the measuring cell and if necessary re-calibrating the bias of the calibration curve.

8.2.8.                    Check the relevance of the measurement accuracy; determinations performed in series should include the periodic analysis of a standard solution recognized by the reference method.

The cycle is to be respected, under the conditions described above, involves the analysis of a standard after five determinations.

8.2.9.                    Clean and rinse the measuring cell at length, at the end of the analysis.

CALIBRATION OF AN INFRALYSER FOR THE DETERMINATION OF THE ALCOHOLIC STRENGTH BY VOLUME ON BRANDIES OR LIQUEUR WINES

1.                Selection of filters

As an example, the filter selections listed below can be used to measure the alcoholic strength by volume in the following alcoholic beverages:

1.1.      Wine brandies aged in wooden casks

1.1.1.                    First combination.

• 2310 nm
• 1778 nm
• 2100 nm
• 1680 nm

1.1.2.                    Second combination.

• 2310 nm
• 2230 nm
• 1769 nm
• 1940 nm
• 1680 nm

1.2.      "Pastis" aniseed-flavoured alcoholic beverage.

• 2270 nm
• 2230 nm
• 1769 nm
• 1940 nm
• 1680 nm

Use of an infralyser for WINE BRANDIES

Select the following filters: 4, 13, 14 and 20.

Four levels of alcoholic strength by volume are calibrated on raw brandies:

1 : 37.5 to 43 % vol.

2 : 42.5 to 47.5 % vol.

3 : 57.5 to 62 % vol.

4 : 67.5 to 72.5 % vol.

For each level, and by increments of 0.5% vol., 10 to 11 determinations of the alcoholic strength by volume based on the pycnometric method are used to calibrate the infralyser

2.                Bibliography

1. Calibration of an infralyser for the determination of the alcoholic strength by volume on brandies or liqueur wines.
2. Station viticole, Cognac National Interprofessionnel Bureau