A rapid method for determination of ethanol in alcoholic beverages using capillary gas chromatography

Volume 11 Issue 2 Article 3

A rapid method for determination of ethanol in alcoholic beverages using A rapid method for determination of ethanol in alcoholic beverages using

capillary gas chromatography capillary gas chromatography

Follow this and additional works at: https://www.jfda-online.com/journal

Recommended Citation Recommended Citation

Wang, M.-L.; Choong, Y.-M.; Su, N.-W.; and Lee, M.-S. (2003) "A rapid method for determination of ethanol in alcoholic

beverages using capillary gas chromatography,"

Journal of Food and Drug Analysis

: Vol. 11 : Iss. 2 , Article 3.

Available at: https://doi.org/10.38212/2224-6614.2710

This Original Article is brought to you for free and open access by Journal of Food and Drug Analysis. It has been accepted for

inclusion in Journal of Food and Drug Analysis by an authorized editor of Journal of Food and Drug Analysis.

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003, Pages

133

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003, Pages 133-140

A Rapid Method for Determination of Ethanol in Alcoholic

Beverages Using Capillary Gas Chromatography

MEI-LING WANG

, YOUK-MENG CHOONG

, NAN-WEI SU

AND MIN-HSIUNG LEE

Department of Food Health, Chia-Nan University of Pharmacy and Science. 60, Al-jen Road. Sect. 1,

Pau-an, Jen-teh Hsiang, Tainan Hsien 717, Taiwan, R.O.C.

Department of Food Science and Technology, Tajen Institute of Technology, 20, Wei-shinn Rd., Yan-puu hsiang,

Pintung Hsien 907, Taiwan, R.O.C.

Department of Agricultural Chemistry, National Taiwan University, 1,sec. 4, Roosevelt Road., Taipei 106, Taiwan, R.O.C .

(Received: May 16, 2002; Accepted: September 30, 2002)

ABSTRACT

A simple and rapid method was developed to determine ethanol content in alcoholic beverages using megapore polar column (CP-

Wax 58 CB, 30 m

0.53 mm) with direct injection gas chromatography. Contrary to packed GLC method, distillation and/or stepwise

dilution of samples were not necessary by the method developed in this study. Ethanol in sample was injected directly into GC for

analysis, after adding suitable amount of internal standard, acetonitrile solution. Using this method, less than 8 min was required to

obtain result since sample preparation, and the limit of quantitation (LOQ) was about 0.5 µg/mL. Recovery studies were performed

using 0.5 mL of red wine and whisky. Each was spiked with ethanol at 50 and 100 mg, respectively. The recoveries were found in the

range of 99~104% and 99~101%, respectively. The coefficients of variation were less than 3.4%. Comparisons of the AOAC method

(AOAC 969.12 and 920.57) with current method showed no significant difference. These results suggested that precision of direct

injection GC method was higher than that of AOAC methods. Several commercial alcohol beverages, including distilled and non-

distilled spirits, were analyzed by the current method. The ethanol content of distilled and non-distilled spirit were found as: 165.2 ±

4.9~415.7 ± 17.6 and 28.2 ± 0.8~141.2 ± 4.9 mg/mL, respectively. Using this GC method, we could change the concentration in gravi-

metric percentage (%, w/v) to volumetric percentage (%, v/v) by the equation (%, w/v) = 0.814 (%, v/v) with linear coefficient R

higher than 0.999.

Key words: alcoholic beverages, ethanol, direct injection gas chromatography, quantitative analyses.

INTRODUCTION

Ethanol content is very important for the mouth-feel

and flavor of alcoholic beverages. Ethanol contents of

wine, liqueur, and beer range from 7

21%(v/v),

50%(v/v), and 3

6%(v/v), respectively

(1)

. In general,

ethanol contents serve as the quality index and taxation

factor for alcoholic beverages

(2)

. After entering WTO,

alcoholic beverages in Taiwan are taxed according to the

ethanol contents, like the taxation system in United

States

(3)

. The higher the ethanol content in an alcoholic

beverage, the higher the tax. A simple, accurate, and quan-

titative analysis method for the determination of different

levels (0

50%, v/v) of ethanol content is needed as the

standard method of quality control for beverage manufac-

turers and as a guide for government-related sanitary

agencies.

Currently, ethanol determination in alcoholic

beverages can be performed in several ways: (1) boiling

point depression of the ethanol solution relative to water,

(1)

(2) densimetric analysis

(4)

, (3) refractive index method

(1,4)

(4) oxidation of the distillate

(1,4-5)

, (5) dichromate oxidation

spectrophotometry

(1,4,6)

, (6) enzymatic method

(7-9)

, (7)

biosensor

(10)

, (8) potentiometry

(11)

, (9) gas chromatography

(GC)

(1,4,12-21)

, (10) capillary electrophoresis

(2)

, (11) high

performance liquid chromatography (HPLC)

(22-23)

, (12)

modular Raman spectrometry

(24)

, (13) near-infrared (NIR)

spectroscopy

(25)

, (14) beer analyzer

(4)

, and (15) flow

injection analysis

(26-27)

. Due to complicated pretreatment

procedures (e.g. sample distillation and accurate weighing

process) and large sample volume required, the two most

popular methods, picnometry and densimetric analysis, are

not applicable for samples with small amount. As for

oxidation of the distillate and dichromate oxidation spec-

trophotometry, more than 5 mL of sample volume is

required for analysis. Besides, the reagents used are highly

toxic and classified. Low stability, low reproducibility and

low accuracy are the disadvantages for enzymatic method,

biosensor, and potentiometry

(9)

. Raman spectrometry and

capillary electrophoresis, are not popular due to the

expensive instruments required. Since sample distillation

and accurate weighing process are still required as the pre-

treatment procedures, and only spectrophotometry is used

for analysis, HPLC method obtains a comparatively low

sensitivity

(22)

. Recently developed NIR spectroscopy and

beer analyzer are time consuming in establishing calibration

curves and have low accuracy as they can be interfered by

* Author for correspondence. Tel:886-8-7624002 ext 352;

Fax:886-8-7621972; E-mail: [email protected]

other alcohols in alcoholic beverages. In conclusion, GC

method is the most appropriate and rapid method for deter-

mination of ethanol contents in alcoholic beverages with

complicated alcohol contents and small sample amount.

Bouthilet et al. had developed a packed GC method to

analyze ethanol contents in alcoholic beverages

(14)

However, the method required at least 100 mL of sample

and distillation as the pretreatment process. The packed

GLC

(4,12-13,15-16)

and capillary GC

(17-21)

, which required

only small sample amount, were then developed. The GC

methods currently available for determining ethanol

contents in alcoholic beverages are US official methods,

AOAC 968.09, 984.14, and 986.12

(4)

, and Taiwan official

method, CNS N6181

(15)

. These official GC methods adapt

different sample preparation procedures when dealing with

different sort of samples, and these samples require distilla-

tion or dilution as the pretreatment process. Since packed

column is used, lower resolution and interference by other

alcohols in alcoholic beverages still exist

(12,21)

. Besides,

durability of packed column and reproducibility of the

retention time are relatively low

(12)

. When capillary GC is

used in determining ethanol contents in alcoholic

beverages, time-consuming pretreatment procedures, e.g.

solid-phase extraction

(17)

, head space balancing

(18)

, solvent

extraction

(19)

and distillation

(20-21)

, are still required. The

capillary GC method is one of the most important modern

analytical techniques because of its advantage of high reso-

lution and sensitivity. Owing to years of research using

GC, we found that insertion of a ball of glass wool into a

liner of GC injector could effectively prevent the non-

volatile compounds form getting into the analytical column

and thus, moderate the interference from the contaminants.

We find that commercially available megapore capillary GC

column is superior in water resistance. Even when an

aqueous solution sample is injected directly into the

column, the resolution and reproducibility of retention time

are almost the same as a new column

(28-30)

. In this study,

we will prepare and inject the alcoholic beverages with

adequate amount of internal standard solution directly into

a GC analyzer without any pretreatment process,. Together

with the application of appropriate column and GC condi-

tions, we plan to develop a simple, rapid and accurate

method in determining ethanol contents in alcoholic

beverages. Meanwhile, the AOAC methods will be

compared and assessed in recovery and standard deviations

of intraday/ interday analysis in order to evaluate the

accuracy and precision of the study method.

MATERIALS AND METHODS

I. Materials

Twenty-six alcoholic beverages, including 12 distilled

spirits (whisky, brandy, kaoliang wine, rice wine, Mizhiu

Tou, and medicine wine) and 14 non-distilled spirits (millet

wine, grape wine, fruit wine, Shaohsing wine, and beer),

were purchased from supermarkets in Tainan, Taiwan. LC

grade (purity > 99.5%) ethanol, acetonitrile, 1-propanol,

isopropanol, acetone, 1-butanol and t-butanol were obtained

from ALPS (Taiwan).

II.

Methods

(I) Preparation of standard solution and internal standard

solution

Ten grams of ethanol and acetonitrile were dispensed

into 1000 mL volumetric bottle and then distilled water was

added to the 1000-mL mark. This was the 1% (w/v)

ethanol standard solution or internal (acetonitrile) standard

solution.

(II)

Relative response factor (RRF) of ethanol to acetoni-

trile

Ethanol, 1% (w/v), was mixed with 1% (w/v) acetoni-

trile in various ratios (ethanol:acetonitrile = 15:1, 10:1, 5:1,

2:1, 1:1, 1:2, 1:5, 1:10, and 1:15). A linear regression line

was generated with the GC peak area-under-curve (AUC)

ratio of ethanol to acetonitrile (Y-axis) against the concen-

tration ratio of ethanol to acetonitrile (X-axis). Relative

response factor (RRF) is the slope of the regression line, as

in the Equation 1: RRF = (A

) ÷ (A

); in which,

= ethanol AUC, A

= acetonitrile AUC, W

= ethanol

weight (mg), W

= acetonitrile weight (mg).

(III)

Quantitative ethanol determination

(1) Direct injected GC method

Beverage sample solution (0.5 mL) was dispensed into

an l-mL caped sample vial, and then 5 mL of 1% internal

standard solution (equivalent to 50 mg) was added. After

mixing, 0.1 µL of the sample solution was injected directly

into a GC with syringe. Ethanol content was calculated

according to the Equation 2: Ethanol (mg/mL) = (A

)

/RRF)

1/V; in which, V = sample volume (mL).

(2) Dichromate oxidation method

(1, 4)

Beverage sample solution (1~5 mL) was steam distil-

lated to obtain alcoholic eluate (> 50 mL), and then

oxidized with acidified dichromate. The excessive

potassium dichromate was then titrated with ferric oxide.

The ethanol content in beverage sample could be obtained

by calculating the volume difference of potassium dichro-

mate consumption between sample solution and control

solution.

(3) Distillation-hydrometric method

(1, 4)

Alcoholic volatile compounds in beverage samples

were separated by distillation, and the gravity of the distil-

late was measured by hydrometer. The ethanol content was

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003

134

then converted.

(IV)

The limit of quantitation (LOQ) of ethanol by GC-FID

Ethanol standard solution (10 mg/mL) was initially

diluted with distilled water to make concentrations of 50,

25, 10, 5, 2, ,1, 0.5, 0.1, and 0.05 µg/mL, and then the

internal standard solution was added to each solution.

Solution was injected directly into a GC, which was

equipped with an FID detector. The FID signal was set at

range = 1, and attenuation = 1. The coefficient of variation

(CV, %) of ethanol recovery was set at 15%. This was the

LOQ of ethanol

(28, 31)

. Each analysis was carried out in

triplicate.

(V)

Recovery

5 and 10 mL of 1% (w/v) ethanol standard solution

(equivalent to 50 mg, and 100 mg, respectively) was added

into 0.5 mL of red wine and whisky (in 20-mL vial),

respectively; and then 5 mL of 1% (w/v) internal standard

solution was added. After gentle mixing, 0.1 µL of sample

solution was injected into a GC with syringe for the deter-

mination of the ethanol content. Each analysis was carried

out in triplicate. Blank sample was analyzed at the same

time.

(VI)

Validation of the analysis method

Ethanol standard solution (500 mg/mL) was diluted

with distilled water to make concentrations of in 500, 250,

100, 50, 20, and 10 mg/mL. Each solution (0.5 mL) was

dispensed into a 20-mL caped vial. After gentle mixing

with 5 mL of 1% (w/v) internal standard solution (equiva-

lent to 50 mg), 0.1 µL of the mixture solution was injected

into a GC with syringe. Each concentration of ethanol

standard solution was measured in triplicate in 1 day

(intraday) or in 3 successive days (interday). Standard

deviation (SD) and coefficient of variation (CV, %) were

measured to evaluate precision of the method. The relative

error of the mean (REM) was measured to evaluate

accuracy of the method. REM was calculated according to

the equation: REM (%) = [(measured value - true value) ÷

(true value)

100%].

(VII)

GC conditions

This study used Trace GC 2000 (TermoQuest, Milan,

Italy), which was equipped with computer integrator

software (Chrom-Card version 1.06 for Trace GC,

TermoQuest, Milan, Italy) and an FID detector. The flow

rates of H

and air were set at 30 and 300 mL/min, respec-

tively. The temperature of the FID detector and the

injection port was set at 285ûC, and 225û C, respectively.

Nitrogen (N

) in the flow rate of 2 mL/min was used as the

carrier gas. The CP-Wax 58 CB separation column (30 m

0.53 mm, Chrompack, Netherlands) was used. Oven tem-

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003

135

perature was set initially at 45ûC for 2 min, and then

increased to the final temperature of 245ûC in 1 min at the

rate of 45ûC/min. Injection volume was limited to 0.1 µL.

Splittless injection mode was selected.

RESULTS AND DISCUSSION

I. Analysis of GC Conditions

During the selection of GC column, we had compared

the polar and mid-polar megapore capillary columns in the

determination of ethanol contents in alcoholic beverages.

Because there was no sample pretreatment procedures

required, and the sample could be injected directly into a

GC for analysis, polar CP-Wax 58 CB column was the most

appropriate column for the determination of ethanol

contents in alcoholic beverages. When applying the GC

conditions described in Materials and Methods, the

retention time of ethanol standard solution was 2.73 min

(Figure 1A, 1B, and 1C). The resolution of megapore

capillary column (Rs = 5.8) was better than the packed

GLC method (Rs = 1.4~1.9), as described in AOAC

(4)

and

CNS

(15)

. The GC peaks of the other minute components in

alcoholic beverages could be separated effectively with

megapore capillary GC (data not shown).

In the selection of internal standard, distilled spirits

(e.g. whisky) and non-distilled spirits (e.g. red wine) were

47.4

37.6

27.81

(mVolt)

18.01

8.22

-1.58

0.0 1.613

(A)

2.75

3.35

3.226 4.839 6.452 8.065(min)

408.2

322.2

236.2

(mVolt)

150.2

64.1

-21.9

0.0 1.613

(B)

2.76

3.35

3.87

4.56

5.6825

5.69

7.77.7239

3.226 4.839 6.452 8.065(min)

575.2

455.1

335.0

(mVolt)

214.9

94.8

-25.3

0.0 1.613

(C)

2.73

3.31

3.46

3.83

4.54

5.56 5.886

6.26

6.74

7.7

6.94

3.226 4.839 6.452 8.065(min)

Figure 1. Gas chromatograms of (A) ethanol and acetonitrile

authentic compound; (B) ethanol in non-distilled spirit (red wine) and

Peaks: 1 = ethanol, 2 = acetonitrile (IS).

added with minute quantity of 1-propanol, 2-propanol, ace-

tonitrile, acetone, 1-butanol, and t-butanol. The ethanol

contents were then analyzed according to the above-

mentioned method. The results had shown that retention

times of the 6 standard solutions were 4.43, 4.37, 3.32,

4.06, 5.96, and 5.72 min, respectively (data not showed).

When GC chromatograms of distilled spirits were

compared with non-distilled spirits, no overlapping of ace-

tonitrile GC peaks was observed. Meanwhile, the GC

peaks of acetonitrile and ethanol were closer to each other

than the other compounds. Acetonitrile was then selected

as the internal standard. Due to the possibility of its

existence in some sort of alcoholic beverages

(1,12)

, 1-

propanol, acetone, t-butanol, or 1-butanol, which acted as

the internal standards GLC methods (AOAC

(4)

, CNS

(15)

and other researches

(12, 16)

), are not appropriate as the

internal standard for the determination of ethanol contents

in alcoholic beverages. Acetonitrile is the most appropriate

internal standard for the determination of ethanol contents

in alcoholic beverages, because no acetonitrile could

possibly exist in alcoholic beverages.

For selection of the GC conditions, we initially

selected low temperature, 45ûC for 2 min, and then

increased rapidly (at the rate of 45ûC/min) to 245ûC in 6

min. Ethanol and acetonitrile were eluted at 80ûC and

100ûC, respectively. Through this process, sample compo-

nents will be eluted very rapidly and it takes 7-8 min to

complete a sample analysis.

II.

Relative Response Factor of Ethanol to Iinternal

Standard

In this study, we selected acetonitrile as the internal

standard for quantitative determination of ethanol contents

in alcoholic beverages. In order to obtain an accurate quan-

tification, the RRF value of ethanol against acetonitrile

needed to be specified, then the ethanol contents could be

calculated according to Equation 2. When the AUC ratio of

ethanol to acetonitrile (Y-axis) was plotted against the con-

centration ratio of ethanol to acetonitrile (X-axis), a linear

regression equation (Y = 0.952 X) was generated (R



0.999). The slope of the linear regression line, 0.952, is the

RRF of ethanol to the internal standard. The linearity of

the adjusted regression line is in the range of 0-500 mg/mL.

III.

The Effect of Internal Standard Volume to the Accuracy

of Ethanol Content Quantification for Various Alcoholic

Beverages

Ethanol contents in commercial alcoholic beverages

vary a lot, even for the same sort of beverages, e.g. 3~6%

(w/v) for beers, 7~21% (w/v) for fermented wines, and

20~50% (w/v) for distilled spirits.

(1)

Therefore, different

pretreatment procedures are required when using the

current official packed GLC methods for the determination

of ethanol contents in various alcoholic beverages. In this

study, a rapid capillary GC method, which requires smaller

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003

136

sample amount and simple pretreatment procedures, is

developed and can be adapted in different sort of alcoholic

beverages with ethanol contents between 0~500 mg/mL.

Ethanol standard solutions in the concentrations of 5,

10, 20, 30, 40, and 50% (w/v) were prepared, and then 0.5

mL of each solution was dispensed into a 20-mL caped

vial. After adding and gentle mixing with 0.5, 1.0, 2.0, 2.5,

3.0, 4.0, 5.0, 6.0, 7.0, and 8.0 mL of 1% (w/v) internal

standard solutions, respectively 0.1 µL of each solution was

then injected directly into a GC for analysis. In order to

achieve better accuracy, samples with high contents of

ethanol needed larger amount of 1% (w/v) internal standard

solution (Figure 2). For different beverage samples with

ethanol contents of 5, 10, 20, 30, 40, and 50% (w/v), the

ratios of 1% (w/v) internal standard solution required were

2, 4, 6, 8, 10, and 12, respectively. In order to obtain

results with high consistency, 10x amount of the internal

standard solution was added when ethanol contents in

alcoholic beverages were determined in this study.

In conclusion, we have developed a method which

requires only 0.5 mL, even 0.1 mL (data not shown), of

beverage sample amount mixed with adequate amount

(10x) of 1% internal standard solution. This sample

solution is suitable for injecting directly into a GC for

determination of ethanol content. No pretreatment

procedure is required. This study method, is more simpli-

fied than official methods AOAC

(4)

and CNS

(15)

, which

require dilution or distillation as the pretreatment proce-

dures. This study method is also easier in operation

compared with methods developed by Collins

et al.

(2)

(CE

method) and by Antonelli

(12)

(packed GLC method), which

require ultra-filtration (0.45 µm filter) and dilution as pre-

100

0246810121416

Volume ratio of IS to ethanol

Recovery (%)

5%(w/v)

30%(w/v)

10%(w/v)

40%(w/v)

20%(w/v)

50%(w/v)

Figure 2. Effect of the volume ratio of 1% internal standard (IS)

solution to various concentration of standard ethanol solution

(1)

the quantitative determination

(2)

of ethanol content.

(1) The volume ratio of IS solution to standard ethanol solution, v/v,

0.5 mL of standard ethanol solution was used in each analysis.

(2) Recovery(%) = (level of ethanol detected)/( ethanol level of the

standard solution)

100%.

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003

137

treatment procedures.

IV.

The Limit of Quantitation (LOQ) of Ethanol by GC-FID

In this study, 1% (w/v) ethanol standard solution was

successively diluted, and then mixed with adequate amount

of internal standard. The mixture solutions were then

injected directly into a GC, equipped with an FID detector.

The FID signal was set at range = 1, and attenuation = 1.

The coefficient of variation (CV, %) of ethanol recovery

was set at 15%. This was the LOQ of ethanol.

(28, 31)

The

results showed that the LOQ of ethanol was around 0.5

µg/mL (Table 1).

Validation of the Analysis Method

The recoveries of ethanol standard solutions, which

spiked into a non-distilled spirit (red wine) and a distilled

spirit (whisky), were found in Table 2. The results show

that the recoveries of ethanol with 50 and 100 mg fortifica-

tion level into 0.5 mL of red wine and whisky were in the

range of 99~104%, and 99~101%, respectively, with coeffi-

cient of variable (CV) less than 3.4%. When 6 ethanol

samples with known concentrations (10~500 mg/mL) were

analyzed using the current method, the coefficient of

variant (CV, %) for intraday (1 day) and interday (succes-

sive 3-day) analyses was 1.0~4.7%, and 1.5~3.6%, respec-

tively. These results indicate that the precision of the study

method is very high. In addition, the REM of intraday and

interday analyses was -3.5~4.2% and -1.0~3.0%, respec-

tively. It indicates that the study method possesses very

good accuracy (Table 3). Using the study method, after

beverage samples are mixed with adequate amount of

internal standard solution, they are ready for injecting

directly into a GC for the determination of ethanol contents.

It takes 7~8 min to complete an analysis. The GC chro-

matograms were illustrated in Figure 1A, 1B, and 1C. This

study method is one of the most simple, rapid, and accurate

methods published for quantitative determination of ethanol

contents in alcoholic beverages. It also can be proposed as

the reference method for routine analysis.

VI.

Comparison of Study Method with Dichromate

Oxidation Method and Distillation-Hydrometric Method

The study method was compared with dichromate

oxidation method (AOAC 969.12) and distillation-hydro-

metric method (AOAC 920.57) for the determination of

ethanol contents in red wine, rice wine, and whisky (Table

4). No statistically significant difference between these 3

methods (p < 0.05, n = 3) was observed, which indicates

Table 1. Limit of quantitation, LOQ, of ethanol using gas chromato-

graphy with FID detector by direct injection method

Ethanol content (mg/mL)

Recovery (%)

CV (%)

50.0 105.4 ± 0 2.4 2.3

25.0 97.5

± 0 4.4 4.5

10.0 98.4

± 0 6.6 6.7

05.0 104.7

± 0 9.2 8.8

02.0 106.7

± 0 5.4 5.1

01.0 095.7

± 0 8.9 9.3

00.5 105.3

± 13.3 12.6

00.1 107.2

± 22.3 20.8

00.05 138.4 ± 41.1 29.7

a: FID range = 1, Attenuation = 1.

b: Average of triplicate analyses.

c: Coefficient of variation (cv %).

Table 2. Recoveries of the ethanol in spiked commercial alcoholic beverage by direct injection gas chromatographic method

Sample Blank

(mg) (A) Amount added (mg)(B) Amount found

(mg) (C) Recovery(%)

CV(%)

Red wine 90.5 ± 2.7 50 142.7 ± 4.0 104.0 2.8

100 189.1

± 3.9 98.6 2.1

Whisky 317.9

± 11.6 50 367.3 ± 10.3 99.0 2.8

100 418.5 ± 14.4 100.6 3.4

a: Ethanol in 0.5 mL alcoholic beverage.

b: Average of triplicate analyses.

c: Recovery(%) = (C–A) / B

¥ 100%.

d: Coefficient of variation (cv %).

Table 3. Precision and accuracy of intraday and interday validation for concentration range from 10 to 500 mg/mL

Concentration Intraday (n = 3)

Interday (n = 9)

(mg/mL Precision Accuracy Precision Accuracy

Mean ± SD (CV %) REM(%)

10 9.9

± 0.1 (1.0) -1.0 10.2 ± 0.2 (2.0) 2.0

20 19.7

± 0.4 (2.0) -1.5 19.8 ± 0.3 (1.5) -1.0

50 52.1 ± 1.9 (3.6) 4.2 50.7 ± 0.8 (1.6) 1.4

100 96.5

± 3.1 (3.2) -3.5 102.4 ± 2.7 (2.6) 2.4

250 259.3

± 8.5 (3.3) 3.7 255.6 ± 7.1 (2.8) 2.2

500 487.8 ± 23.1 (4.7) -2.4 515.2 ± 18.7 (3.6) 3.0

a: n = 3, Repeated injection three times on the same day.

b: n = 9, Repeated injection three times each day and a successive three-day.

c: REM = relative error of the mean.

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003

138

the study method we proposed obtains very high accuracy.

The ethanol contents determined by the study method were:

107.4 ± 2.9 mg/mL in red wine, 158.2 ± 6.0 mg/mL in rice

wine, and 326.3 ± 7.8 mg/mL in whisky (CV = 2.4~3.8%).

The ethanol contents determined by dichromate oxidation

method were: 110.6 ± 14.2 mg/mL in red wine, 151.4 ±

16.8 mg/mL in rice wine, and 311.3 ± 37.7 mg/mL in

whisky (CV = 11.1~14.2%). The ethanol contents deter-

mined by distillation-hydrometric method were: 114.1 ±

13.8 mg/mL in red wine, 153.2 ± 14.7 mg/mL in rice wine,

and 307.6 ± 40.5 mg/mL in whisky (CV = 9.6~13.2%).

These results also indicate that the study method has better

precision than the AOAC methods. Both AOAC methods

require dilution and distillation as the pretreatment proce-

dures. High deviation may be found if the connection tubes

are not tightly secured, condensation efficiency is low, or

the sample contains high contents of volatile compounds

(e.g. acetic acid or sulfur dioxide) in distillation process

(1)

Also, highly toxic reagents used in dichromate oxidation

method make the operation and waste disposition trouble-

some and dangerous.

In the application point of view, the simple, rapid, and

direct-injected capillary megapore column GC method

developed in this study has shown more advantages and

values compared with the AOAC

(4)

and CNS official

methods

(15)

. Currently, different beverage samples require

unique sample preparation procedures when the official

AOAC and CNS packed GLC methods are applied, and dis-

tillation or dilution is required as the pretreatment

procedure. Since the resolution is low for packed GLC

methods

(12)

, the accuracy for quantitative analysis and

reproducibility of ethanol contents will be interfered for

those beverages with high and complicated contents of

volatile compounds (e.g. fruit and medicine wines)

(21)

. In

addition, packed GLC column has shorter shelf life and

lower durability. Besides, currently adapted capillary GC

methods require time-consuming sample pretreatment pro-

cedures, such as solid-phase extraction

(17)

, headspace

balancing

(18)

, solvent extraction

(19)

, and distillation

(20-21)

The study method we developed is a simple and rapid

method, without any sample pretreatment procedure

required. After beverage samples mixed with adequate

amount of the internal standard solution, the samples are

ready to be injected into a GC for determination of ethanol

contents.

VII.

The Ethanol Contents of Commercial Alcoholic

Beverages

The ethanol contents of commercial alcoholic

beverages are usually labeled in volume percent (%, v/v).

Sometimes, it is convenient to be labeled in weight-volume

percent (%, w/v). In this study, the ethanol standard

solutions, 10%, 20%, 30%, and 40%, were prepared both in

weight-volume percent (%, w/v), and in volume percent (%,

v/v). After beverage sample were mixed with adequate

amount of internal standard and injected into a GC for the

determination of ethanol content, the relative response

factor (RRF) was then calculated according to the above

mentioned equation. The linear regression line of ethanol

weight-volume percent (%, w/v, Y-axis) against ethanol

volume percent (%, v/v, X-axis) was: Y = 0.814 X, with R

= 0.999 (Figure 5).

Table 5. Ethanol levels determined in alcoholic beverages.

Sample Ethanol Content

Distilled spirit mg/mL %, v/v

Whisky 1 327.6

± 11.8 40.3 ± 1.4

Whisky 2 317.8

± 12.4 39.0 ± 1.5

Whisky 3 331.3

± 13.2 40.7 ± 1.6

White liquor 1 328.9

± 09.5 40.4 ± 1.2

White liquor 2 413.6

± 19.7 50.8 ± 1.8

Kaoliang1 327.5

± 12.4 40.2 ± 1.5

Kaoliang2 415.7

± 17.6 51.1 ± 1.7

XO 313.0

± 13.6 38.5 ± 1.7

VSOP 289.7

± 08.2 35.6 ± 1.0

Rice wine 165.2

± 04.9 20.3 ± 0.6

Mizhiu Tou 324.3

± 10.3 39.8 ± 1.3

Medecine wine 197.7

±05.3 24.3 ± 0.7

Non-distilled spirit

Millet wine1 77.2

± 02.9 9.5 ± 0.4

Millet wine2 81.3

± 01.4 9.9 ± 0.5

Red wine 1 90.1

± 03.7 11.1 ± 0.5

Red wine 2 106.5

± 04.5 13.1 ± 0.5

Red rose 1 128.4

± 05.5 15.8 ± 0.7

Red rose 2 117.1

± 03.7 14.4 ± 0.5

Fruit wine1 91.2

± 02.6 11.2 ± 0.3

Fruit wine2 112.5

± 03.4 13.8 ± 0.5

White wine 1 100.4

± 03.1 12.3 ± 0.4

White wine 2 103.9

± 03.7 12.8 ± 0.5

Shao-Hsing wine 1 141.2

± 04.9 17.3 ± 0.6

Shao-Hsing wine 2 138.6

± 05.5 17.0 ± 0.7

Beer 1 28.2

± 00.8 3.5 ± 0.1

Beer 2 29.8 ± 00.6 3.7 ± 0.1

a: Average of triplicate analyses.

Table 4. Comparison of AOAC methods and proposed method for the determining of ethanol content in alcoholic beverages.

Sample Proposed method

Dichromate oxidation method

Distisllation hydrometric method

Ethanol (mg/mL) CV (%)

Red wine 107.4 ± 2.9 2.7 110.6 ± 15.7 14.2 114.1 ± 13.8 12.1

Rice wine 158.2

± 6.0 3.8 151.4 ± 16.8 11.1 153.2 ± 14.7 9.6

Whisky 326.3 ± 7.8 2.4 311.3 ± 37.7 12.1 307.6 ± 40.5 13.2

a: Direct injection GC method in this study.

b: AOAC 969.12 method.

c: AOAC 920.57 method.

d: Average of triplicate analyses, coefficient of variation (cv%).

The ethanol contents of the 26 commercial alcoholic

beverages (12 distilled spirits and 14 non-distilled spirits)

were analyzed with the study method. As shown in Table

5, the ethanol contents of 12 distilled spirits were:

317.8~331.3 mg/mL (39.0~40.7%, v/v) in whisky,

328.9~413.6 mg/mL (40.4~50.8%, v/v) in white liquor,

327.5~415.7 mg/mL (40.2~51.1%, v/v) in kaoliang wine,

313.0 mg/mL (38.5%, v/v) in XO, 289.7 mg/mL (35.6%,

v/v) in VSOP, 165.2 mg/mL (20.3%, v/v) in rice wine,

324.3 mg/mL (39.8%, v/v) in Mizhiu Tou, and 197.7

mg/mL (24.3%, v/v) in medicine wine. The ethanol

contents of the 14 non-distilled spirits were: 77.2~81.3

mg/mL (9.5~9.9%, v/v) in millet wine, 90.1~128.4 mg/mL

(11.7~15.8%, v/v) in grape wine, 91.2~112.5 mg/mL

(11.2~13.8%, v/v) in fruit wine, 100.4~103.9 mg/mL

(12.3~12.8%, v/v) in white wine, 138.6~141.2 mg/mL

(17.0~17.3%, v/v) in Shaohsing wine, and 28.2~29.8

mg/mL (3.5~3.7%, v/v) in beer. The ethanol contents of

these alcoholic beverages were consistent with their labels,

without any fraud found.

CONCLUSIONS

In this study, it takes only 7~8 min to complete a

sample analysis for the determination of ethanol content in

a beverage sample. A sample solution (0.5 mL) is mixed

with adequate amount (5 mL) of 1% (w/v) internal standard

solution (acetonitrile, equivalent to 50 mg), and injected

into a capillary GC. The study method we developed can

be applied to alcoholic beverages with different alcoholic

contents, and with the advantages of simple sample pre-

treatment procedures, rapidity and accuracy, and maybe a

routine analysis method in substitution of current AOAC

and CNS methods.

ACKNOWLEDGEMENT

This research was supported by Project 90-FH-10 from

Chia-Nan University of Pharmacy and Science. The authors

would like to thank Dr. Hui-Cheng Chen for his translation

work.

REFERENCES

1. Anonymous. 1992. Chapter 4-Alcohols. In Wine

Analysis. Research Institute for Wines, Taiwan Tobacco

& Wine Monopoly Bureau. pp. 1-7.

2. Collins, T. S., Miller, C. A., Altria, K. D. Waterhouse,

A. L. 1997. Development of a rapid method for the

analysis of ethanol in wines using capillary elec-

trophoresis. Am. J. Enol. Vitic. 48: 280-284.

3. US Internal Revenue Service, Wine, Part 240 of Title

26, IRS Publ. No. 146, Code of Federal Regulations.

US Gov. Printing Office, Washington, DC 1970.

4. AOAC Official Method of Analysis. 1990. Wines. In

Official Methods of Analysis of AOAC International.

15th edition, pp. 739-750.

5. Caputi, A. Jr. and Wright, D. 1969. Collaborative study

of the determination of ethanol in wines by chemical

oxidation. J. Assoc. Off. Anal. Chem. 52: 85-88.

6. Caputi, A. Jr., Ueda, M. and Brown, T. 1968.

Spectrophotometric determination of ethanol in wines.

Am. J. Enol. Vitic. 19: 160-165.

7. Jones, A. W. 1995. Measuring ethanol in saliva with the

QED enzymatic test device: comparison of results with

blood and breath alcohol concentration. J. Anal.

Toxicol. 19: 169-174.

8. McCloskey, L. P., and Replogle, L. L. 1974. Evaluation

of an enzymatic method for estimating ethanol in wines

using an enzyme kit. Am. J. Enol. Vitic. 25: 194-197.

9. Mason, M. 1983. Report of the technical subcommittee

on ethanol determination via immobilized enzyme. J.

Am. Soc. Brew. Chem. 41: 89-90.

10. Mason, M. 1983. Ethanol determination in wine with an

immobilized enzyme electrode. Am. J. Enol. Vitic. 34:

173-175.

11. Kokovkin, V. V. and Smolyakov, B. S. 1995. Ethanol

determination in aqueous solutions and wine stocks by

potentiometry with ion selective electrodes. J. Anal.

Chem. 50: 569-573.

12. Antonelli, A. 1994. Ethanol determination by packed

GLC: a quick method with small sample amount and

high sensitivity. Vitic. Enol. Sci. 49: 165-167.

13. Caputi, A. Jr. and Mooney, D. P. 1983. Gas chromatog-

raphy determination of ethanol in wine: Collaborative

study. J. Assoc. Off. Anal. Chem. 66: 1152-1157.

14. Bouthilet, R. J., Caputi, A. Jr. and Ueda, M. 1961.

Analysis of ethanol in wine by gas liquid partition chro-

matography. J. Assoc. Off. Anal. Chem. 44: 410-414.

15. CNS, 1983, Method of test for beverage-Determination

of alcohol content. Chinese National Standard, Taiwan.

Code No. 10292 (N6181).

16. Stackler, R. and Christensen, E. N. 1974. Quantitative

determination of ethanol in wine by gas chromatogra-

phy. Am. J. Enol. Vitic. 25: 202-207.

17. Kumazawa, T., Seno, H., Lee X. P., Ishii, A., Suzuki,

O., and Sato, K. 1996. Detection of ethanol in human

body fluids by headspace solid phase micro extraction

(SPME) capillary gas chromatography. Chromatogr. 43:

393-397.

18. McCarver-May, D. G. and Durisin, L. 1997. An

accurate, automated, simultaneous gas chromatographic

headspace measurement of whole blood ethanol and

acetylaldehyde for human in vivo studies. J. Anal.

Toxic. 21: 134-141.

19. Moore, S. J., Porter, A. C., Ezell, A. E., Hume, A. S.

and Ho, I. K. 1993. The rapid determination of ethanol

in postmortem brain samples by solvent extraction and

gas chromatography. J. Food Drug Anal. 3: 245-249.

20. Masuda, S., Kikuchi, K. and Harayama, K. 2000.

Determination of lightstruck character in beer by gas

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003

139

Journal of Food and Drug Analysis, Vol. 11, No. 2, 2003

140

chromatography-mass spectroscopy. J. Am. Brew.

Chem. 58: 152-154.

21. Naviglio, D., Romano, R. and Attanasio, G. 2001.

Rapid determination of ethanol content in spirits and in

beer by high resolution gas chromatography. Italian

Food Bever. Technol. 24: 19-21.

22. Kupina, S. A. 1984. A simultaneous quantitation of

glycerol, acetic acid and ethanol in grape juice by high

performance liquid chromatography. Am. J. Enol. Vitic.

35: 59-62.

23. Martin, E., Iadaresta, V., Giacometti, J. C., and Vogel, J.

1986. Ethanol determination by HPLC in alcoholic

beverages. Mitteilungen aus dem Gebiete der

Lebensmitteluntersuchung und Hygiene. 77: 528-534.

24. Sanford, C. L. and Mantooth, B. A. 2001. Determin-

ation of ethanol in alcohol samples using a modular

Raman spectrometer. J. Chem. Edu. 78: 1221.

25. Van den Berg, F. W. J., van Osenbruggen, W. A. and

Smilde A. K. 1997. Process analytical chemistry in the

distillation industry using near-infrared spectroscopy.

Proc. Contr. Qual. 9: 51-57.

26. Mattos, I. L., Sartini, R. P., Zagatto, E. A. G., Reis, B.

F. and Gine, M. F. 1998. Spectrophotometric flow

injection determination of ethanol in distilled spirits and

wines involving permeation through a silicon tubular

membrane. Anal. Sci. 14: 1005-1008.

27. Wagner, K., Bilitewski, U. and Schmid R. D. 1992.

Flow injection analysis of wine - accomplishments and

needs. Microchem. J. 45: 114-120.

28. Choong, Y. M., Ku, K. L., Wang, M. L. and Lee, M. H.

1995. Simple and rapid method for the determination of

sorbic acid and benzoic acid in foods. J. Chin. Agric.

Chem. Soc. 33: 247-261.

29. Lin, H. J. and Choong, Y. M. 1999. Simple method for

the simultaneous determination of various preservatives

in liquid foods. J. Food Drug Anal. 7: 291-304.

30. Lin, H. J., Wang, M. L., Chen, C. W., Hwang, B. S.,

Lee, M. H. and Choong, Y. M. 2000. A gas chromato-

graphic method for determination of nicotinamide,

paraben esters and caffeine in commercial health drink,

tonic drink and cold formula. J. Food Drug Anal. 8:

180-186.

31. Loong, G. L., Winefordner, J. D. 1983. Limit of

detection: a closer look at the IUPAC definition. Anal.

Chem. 55: 712-724A.