A novel device for the headspace trapping of volatiles: Studies with a model solution

JOURNAL

OF FOOD

COMPOSITION

AND ANALYSIS

5, l39- 145 ( 1992)

A Novel Device for the Headspace Trapping of Volatiles: Studies with a Model Solution KAZUOISHIHARA*

ANDNOBUOHONMA

Niigata Women’s College, 471, Ebigase, Niigata 950, Japan Received August 18, 199 1, and in revised form January 10, 1992 The use of a novel headspace volatile trapping apparatus consisting of a heated sample solution container, a heated upper part of the sample vessel, and a heated Tenax TA (2,6-diphenyl-pphenyleneoxide polymer) tube for the collection of headspace volatiles from a model solution is described. Vapor from the sample solution was prevented from condensing on the interior surface of the sample vessel and the Tenax TA tube by maintaining the temperature of the heating parts at 45°C the same as that of the sample solution. Removal of the water collected on Tenax TA prior to GC analysis was thus not required. Headspace volatiles were collected on the 200 mg of Tenax TA by passing high-purity nitrogen through the sample solution at a rate of 50 ml/min for I h. The model sample solution contained ethyl acetate, eight different alcohols, and 2,3xylenol. Data on the effect of sodium chloride concentration of the sample solution on the collection of headspace volatiles is presented. The coefficients of variation of the CC peak areas of the 0 1992 Academic PESS. I~C. compounds except for ethanol ranged from I .!I to I%.

INTRODUCTION Headspace volatile assays are the analytical methods of choice for food aroma components because these assays measure those compounds detected by organoleptic tests. Since the concentrations of the compounds giving the aroma in the sample are low and the volume of sample taken by a syringe is limited, the sample usually must be concentrated prior to analysis. Various concentrating methods using Tenax GC (2,6diphenyl-g-phenyleneoxide polymer) have been proposed (Aishima, 1982; Barnes et al., 198 1; Boyko et al., 1978; Ioffe and Vitenberg, 1984; Kuo et al., 1977; Leahy and Reineccius, 1984; Olafsdottir et al., 1985; Shimoda et al., 1984; Tsugita et al., 1979). This paper describes a novel device and conditions for reproducible and accurate collection of headspace volatiles from a model solution using Tenax TA (Gasukuro Kogyo Inc., Tokyo, Japan). MATERIALS

AND METHODS

Sample Solution The sample test solution was prepared by dissolving the compounds shown in Table I in distilled water purified by an automatic water distillation apparatus (Adovantec Toyo Co., Tokyo, Japan). 2,3-Xylenol was the internal standard. Collection of Headspace Volatiles The headspace volatiles trapping apparatus is shown in Fig. 1. One hundred milliliters of the sample solution containing 0 to 30 g of sodium chloride was placed in a 250* To whom correspondence and reprint requests should be addressed 139

0889-1575/92 $5.00 Copyright 6 1992 by Academic Press. Inc. All rights of reproductmn I” any form reserved

140

ISHIHARA

AND TABLE

VOLATILE

HONMA I

COMPOUNDS

SOLUTION

Compound 1 2 3 4 5 6 7 8 9 10

Ethyl acetate Ethanol 1-Propanol 2-Methyl-1-propanol l-Butanol 3-Hethyl- l- butanol 1-Hexanol Benzyl alcohol 2-Phenylethanol 2,3-Xylenol

49 47 57 56 51 57 53 60 53 250

FIG. 1. Headspace volatiles trapping apparatus. 1, reducing valve; 2, needle valve; 3, automatic temperature regulator: 4, connector for different diameters; 5, gas washing bottle; 6, ball filter; 7, sample solution; 8, water bath made of plastic sheet; 9, magnetic stirring bar; 10, magnetic stirrer; 1 I, stainless side port needle; 12, GC injection rubber plug; 13, trapping column; 14, quartz wool; 15, Tenax TA; 16, glass heating tube; 17, rubber tubing; 18, flow meter; 19, constant temperature water circulator; 20, heating cable; 21, thermoregulator; 22, air circulator; 23, rubber stopper; 24, thermometer.

HEADSPACE

VOLATILES

TRAPPING

TECHNIQUES

141

ml gas washing bottle. The Tenax TA tube (4 mm i.d. X 17 cm) containing 200 mg of Tenax TA (60-80 mesh) was connected to the sample vessel. The sample vessel was placed in a water bath of 45°C. The Tenax TA tube was covered with a glass heating tube in which constant temperature water was circulated. The sample vessel was covered with a lid provided with a heating cable. In order to more rapidly heat the upper part of the sample vessel, warm air was blown through a hole (No. 23 in Fig. 1) with a hair dryer. For collection of headspace volatiles, a high-purity nitrogen (purity: 99.999%) was introduced into the stirred sample solution at a rate of 50 ml/ min for 0.5, 1, and 1.5 h. The temperature of three parts-the sample solution, the upper part of the sample vessel, and the Tenax TA tube-could be separately controlled. Gas Chromatography

(CC)

GC analysis was performed on a Shimadzu GC-4CPF gas chromatograph (Kyoto, Japan) equipped with a flame ionization detector and a Shimadzu fused silica capillary column (CBP20-M25-025) 0.2 mm i.d. X 25 m length, coated with CBP20 corresponding to polyethylene glycol 20 M. Both inlet and detector temperatures were 250°C and the column oven was programmed from 60 to 200°C at 4”C/min. Nitrogen was used as carrier gas at a flow rate of 1 ml/min with a split ratio of 1:50. The headspace volatiles collected in the Tenax TA tube were injected into the chromatograph in the stream of nitrogen at a flow rate 65.7 ml/min for 0.5 or 2 min using the thermal desorption injector (Shimadzu FLS-1) equipped with a furnace heated at 240°C. The area of each peak was measured with a Shimadzu Chromatopac C-RIB integrator. The collection of headspace volatiles and GC analysis was repeated four times in a run of the experiment. RESULTS

Temperatures.for

AND

DISCUSSION

Collecting Headspace Volatiles

Initially the sample solution was kept at 45°C and the upper part of the sample vessel and the Tenax TA tube were kept at room temperature (ca. 20°C) 45, and 50°C. Coefficient of variation and mean area of each peak are shown in Table 2. A remarkably large quantity of vapor from the sample solution was condensed on the interior surface of both the upper part of the sample vessel and the Tenax TA tube held at room temperature. The peaks of benzyl alcohol and 2-phenylethanol were developed when the upper part of the sample vessel and the Tenax TA tube were held at 45 or 50°C. The coefficient of variation was decreased by keeping the temperature of these three heating parts at 45°C and thus little vapor from the sample solution condensed on the interior surface of the Tenax TA tube. These latter conditions were regarded as optimal. Water from the sample solution had adverse effects on GC and GC-MS analysis, and thus it is generally removed by passing nitrogen or helium gas through the Tenax TA tube before it is connected to the GC column (Aishima, 1982; Boyko et al., 1978; Kuo et al., 1977; Osajima, 1988; Shimoda et al., 1984; Tsugita et al., 1979). Boyko et al. (1978) reported that a 20-min water removal step at 55°C with nitrogen purging at 30 ml/min caused loss of low-boiling compounds. The quantity of water in the Tenax TA tube under the optimum temperature condition in this experiment was so small that water removal was not necessary. Some vapor condensed on the interior

OF STATISTICAL

A

FOR THE REPRODUCIBILITY DIFFERENT

7.76 0.10 1.25 4.11 2.14 5.M) 5.14 0.00 0.00 0.20

P,d x104 10 40 15 14 11 7.7 8.4 16

CV”

Sa:45PC,Ub:Ca.20DC,TC:Ca.209=.

RESULTS

10.39 0.43 0.46 3.97 2.71 6.62 6.74 0.03 0.02 1.37

x, x lo4 9.1 40 5.5 18 7.9 4.6 5.5 68 94 52

cv

S:45”C,U:45”C,T:45”C.

B

C

VOLATILES

10.31 0.87 0.45 4.12 2.92 7.21 7.28 0.04 0.01 0.94

R, x lo4

9.8 107 11 16 9.1 9.6 6.9 29 115 15

cv

(HSV)

S:459=,WO”C,T:45”C.

OF GC ANALYSIS FOR HEADSPACE TEMPERATURE CONDITIONS

UNDER

D

FOUR

6.54 1.05 0.31 2.38 2.06 6.47 6.34 0.05 0.02 1.06

x, x lo4

18 94 23 16 21 29 23 78 131 62

cv

S:45”C,U:50”C,T:5O”C.

COLLECTED

a.b.cThe keys of S, U and T indicate the temperature of sample solution, upper part of sample vessel and Tenax TA tube, respectively. dAverage GC peak area. “Coefficient of variation for 4 runs. HSV collection conditions except the temperature: Tenax TA, 200 mg; the collection time, 0.5 h at 50 ml&n high-purity NZ flow rate; NaCl, 10 g/100 mL; HSV trasfer time from Tenax TA to GC at 240x, 0.5 min at 65.7 mL/min Na flow rate.

Ethyl acetate Ethanol 1 -Propanol 2-Methyl-1-propanol 1 -Butanol 3-Methyl-1-butanol 1-liexanol Benzyl alcohol 2-Phenylethanol 2,3-Xylenol

Compound

COMPARISON

TABLE 2

P

g

$ u

5; 2 F F

HEADSPACE

VOLATILES

TRAPPING

143

TECHNIQUES

surface of the Tenax TA tube, though the CV was small when the Tenax TA tube and the upper part of the sample vessel were held at 45 and 50°C respectively. Relatively small peak areas were produced by raising the temperature of both the upper part of the sample vessel and the Tenax TA tube to 50°C though no vapor condensed on these parts. The fact that compounds collected in the Tenax TA tube were liberated at 50°C may explain the relatively small peak areas described above. The behavior of low- and high-boiling compounds could be investigated by separately controlling the temperature of these three components. Collection Time of Headspace Volatiles Collection times of 0.5, 1, and 1.5 h were compared at a 50 ml/min high-purity nitrogen flow rate as shown in Table 3. The l-h trap was considered the optimal collection time because CVs were relatively small and quantities of high-boiling compounds relatively large. The 0.5-h trap was insufficient since the peak areas of compounds that are higher-boiling than 3-methyl- 1-butanol increased with collection time. High-boiling compounds were not equilibrated even at 1 h. The peak areas of compounds lower-boiling than 2-methyl- 1-propanol decreased with increase in collection to 0.5 h or more. This may have been due to the fact that the low-boiling compounds on Tenax TA were liberated during this time. Leahy and Reineccius (1984), using a Hewlett-Packard Purge and Trap sample collection system, found that although alcohols and esters broke through the Tenax trap as purge time increased from 0.5 to 1 h, the quantities of these compounds were small. Low-boiling compounds are thus TABLE 3 COMPARISON OF STATISTICAL RESULTS FOR THE REPRODUCIBILITY OF GC ANALYSIS VOLATILES (HSV) COLLECTED FOR 0.5,” 1, AND I .5 h AT A 50 ml/min HIGH-PURITY

A

B

lh Compound Ethyl acetate Ethanol 1 -Propanol 2-Hethyl-1-propanol

I-Butanol 3-Methyl-l-butanol 1-Hexanol Benzyl alcohol 2-Phenylethnnol

2,3-Xylenol BThe data of 0.5 h collection time bAverage GC peak area. “Coefficient

R,b x lo4

1.5 CV”

6.78 0.85 0.44 3.42 3.71 13.49 13.65 0.18

FOR HEADSPACE N2 FLOW RATE

24

x, x 104

cv

2.38

28 114 16 12 8.6 8.1

0.29

100

16 12

0.40 2.86

2.96

15

h

10 13 14

17.83 21.71 0.28

0.10

23

0.15

99

3.83

11

3.37

82

refer

to

B in

of variation

Table for time:

11

69

2.

4 runs.

BSV collection conditions except the collection Tenax TA, 200 mg; the temperature condition, S:45”C,U:45”C1T:45”C, the keys refer to Table 2; NaCl, 10 g/l00 ML; HSV trasfer time from Tenax TA to GC at 24O”c, 0.5 tin at 65.7mL/min N2 flow

rate.

144

ISHIHARA

AND

HONMA

liberated along with the water by passing nitrogen or helium gas through the Tenax TA tube to remove water from the sample solution. Sodium chloride from the sample solution started crystallizing in the ball filter at 1.5 h, causing the flow rate of the highpurity nitrogen to fluctuate. These results indicate that it is difficult to quantitatively collect in a fixed time low-, medium-, and high-boiling headspace volatile compounds from food. Eflects

?f the Addition of Sodium Chloride on the Collection qfHeadspace Volatiles

The results of the investigation on the effects of sodium chloride concentration from 0 to 30 g/ 100 ml on the CV of peak area are shown in Table 4. The amount of collected compounds, except for ethyl acetate and ethanol, increased with the concentration of sodium chloride. For 2-phenylethanol and 2,3-xylenol, there was a 2. I- and 2.4-fold increase, respectively, when the concentration of sodium chloride increased from 20 to 30%. The amounts of the other compounds increased approximately 1.7 to 1.9 times. The small peak areas of ethanol and propanol may have been the reason for the low vapor-pressure due to the association of these compounds with water and the low adsorption of these polar compounds on the hydrophobic surface of Tenax TA (Kuo et al., 1977). The small areas of benzyl alcohol and 2-phenylethanol could also be due to the association of these compounds with water (Tsugita et al., 1979). The optimal conditions for collecting headspace volatiles from the sample solution on the 200 mg of Tenax TA were thus as follows: (1) the sample solution container, the upper part of the sample vessel, and the Tenax TA tube were held at 45°C; (2) headspace volatiles were collected by passing high-purity nitrogen through the sample solution at a rate of 50 ml/min for 1 h; and (3) sodium chloride was added to the sample solution at 30%. CVs of the GC peak areas of compounds except for ethanol

TABLE

4

COMPARISON OF STATISTICAL RESULTS FOR THE REPRODUCIBILITY VOLATILES (HSV) COLLECTED FROM THREE DIFFERENT NaCl

Ethyl acetate Ethanol 1-Propanol 2-Hethyl-I-propanol I-Butanol 3-Methyl-1-butanol 1-Hexanol Benzyl alcohol 2-Phenylethanol 2,3-Xylem1

C

NaCl 0 g/100

I

NaCl 20 g/100

X.8 x 104

CVb

x, x 10’

cv

x, x 104

cv

5.68 0.09 0.54 5.16 5.12 22.03 28.93 0.42 0.51 15.16

13 62 17 17 17 12 7.3 13 14 12

5.11 0.28 0.94 8.79 9.20 40.83 52.56 0.74

15 81 1.9 5.0 4.7 2.8 3.6 6.6 11 6.1

5.12 0.17 0.26 1.83 1.77 5.52 5.17 0.07 0.05 1.67

14 101 14 13 14 11 8.7 ii: 15

sL

6C peak area. bCoefficient of variation for 4 runs. conditions except NaCl concentration: Tenax TA. 206 sg; the temperature condition, S:45=C,U:459J,T:45YZCt the keys refer to Table

NaCl 30 g/106

1.06

36.57

aAverage

HSV collection

the collection

HSV transfer

time,

time

FOR HEADSPACE SOLUTIONS

B

A

Conpound

OF GC ANALYSIS CONCENTRATION

1 h at 50 lWdn

high-purity

from Tenax TA to 6C at 2409=,

NP flow

2 sin

2;

rate;

at 65.7 sLJWdn Na flow

rate.

nL

HEADSPACE

VOLATILES

TRAPPING

TECHNIQUES

145

ranged from 1.9 to 15% under the optimal conditions. Further studies on the use of this system for the assay of foods are in progress. REFERENCES AISHIMA, T. (1982). Comparison of headspace and distillation techniques for soy sauce aroma in relation to analysis by silica capillary gas chromatography and sensory evaluation. Agric. Biol. Chem. 46, 27592767. BARNES,R. D., LAW, L. M., AND MACLEOD, A. J. (198 1). Comparison of some porous polymers as adsorbents for collection of odour samples and the application of the technique to an environmental malodour. Analyst. 106(April), 412-418. BOYKO, A. L., MORGAN, M. E., AND LIBBEY, L. M. (1978). Porous polymer trapping for GC/MS analysis of vegetable flavors. In Analysis of Foods and Beverages, Headspace Techniques, (G. Charalambous, Ed.), pp. 57-79. Academic Press, New York. IOFFE, B. V., AND VITENBERG, A. G. (1984). Analysis of water and aqueous solutions. In Head-Space Analysis and Related Methods in Gas Chromatography, Chap. 3, pp. 100-l 17. Wiley, New York. [English trans. by Mamantov, Ilya A.] Kuo, P. P. K., CHIAN, E. S. K., DEWALLE, F. B., AND KIM, J. H. (1977). Gas stripping, sorption, and thermal desorption procedures for preconcentrating volatile polar water-soluble organics from water samples for analysis by gas chromatography. Anal. Chem. 49, 1023-1029. LEAHY, M. M., AND REINECCIUS,G. A. (1984). Comparison of methods for the isolation of volatile compounds from aqueous model systems. In Analysis Gf Volatiles, Methods and Applications, (P. Schreier, Ed.), pp. 19-47. de Gruyter, Berlin, New York. OLAFSDOTTIR,G., STEINKE, J. A., AND LINDSAY, R. C. ( 1985). Quantitative performance of a simple TenaxGC adsorption method for use in the analysis of aroma volatiles. J. Food Sci. 50, 143 l-1436. OSAJIMA, Y. (1988). Evaluation of flavor. In Latest Food Flavor Technology (T. Shinada, Ed.), Chap. 5, pp. 302-325. Industrial Technology Association, Tokyo, Japan. [in Japanese]. SHIMODA, M., WADA, K., AND OSAJIMA, Y. (1984). Effect of temperature of headspace trapping apparatus on quantification of coffee volatiles. Nippon Shokuhin Kogyo Gakkaishi 31, 805-809. TSUGITA, T., IMAI, T., DOI, Y., KURATA, T., AND KATO, H. (1979). GC and GC-MS analysis of headspace volatiles by Tenax GC trapping techniques. Agric. Biol. Chem. 43, 135 1-l 354.