Characterization and Measurement of Flavor Compounds - American

5 A New Analytical Method for Volatile Aldehydes 1

Downloaded by CALIFORNIA INST OF TECHNOLOGY on November 30, 2016 | http://pubs.acs.org Publication Date: November 8, 1985 | doi: 10.1021/bk-1985-0289.ch005

Tateki Hayashi , Clayton A. Reece, and Takayuki Shibamoto Department of Environmental Toxicology, University of California, Davis, C A 95616

Trace quantities of formaldehyde and methyl glyoxal in aqueous and food samples were determined by a newly developed method. Formaldehyde and methyl glyoxal were reacted with cysteamine in aqueous medium or food sample to give thiazolidine and 2-acetylthiazolidine, respectively, at pH 6 and 8. Thiazolidine derivatives formed from formaldehyde and methyl glyoxal were extracted with dichloromethane or chloroform and subsequently analyzed by a gas chromatograph equipped with a fused s i l i c a capillary column and a thermionic detector. Seventeen commercial food items were analyzed for formaldehyde and methyl glyoxal. The quantities of formaldehyde and methyl glyoxal varied from 0 to 17 ppm and from 0 to 620 ppm, respectively.

Certain v o l a t i l e aldehydes such as formaldehyde and methyl glyoxal have always presented some d i f f i c u l t i e s i n the determination of t h e i r levels i n foods and beverages- Formaldehyde i s d i f f i c u l t to extract from an aqueous solution with an organic solvent because i t i s very water soluble or exists as a polymer i n an aqueous média» Methyl glyoxal i s also hard to recover from food samples because i t exists as a copolymer with some amines such as amino acids and proteins. Formaldehyde i s widely used i n many manufacturing processes and i t s production i n the U.S. reached 5.6 b i l l i o n pounds i n 1980 (JJ · Exposure to formaldehyde has caused dermatitis and pulmonary i r r i t a t i o n i n workers, and recent evidence based on animal studies has implicated formaldehyde as a p o t e n t i a l carcinogen (2). Methyl glyoxal has been found i n many foods, such as bread (3), boiled potatoes (4_), roast turkey (J3), and tobacco smoke (6_) · It i s a well known fact that sugar caramelization produces numerous carbonyls including formaldehyde and methyl glyoxal (_7) · Among

1

Current address: Nagoya University, Department of Food Science and Technology, Nagoya, Japan 464 0097-6156/85/0289-0061 $06.00/0 © 1985 American Chemical Society

Bills and Mussinan; Characterization and Measurement of Flavor Compounds ACS Symposium Series; American Chemical Society: Washington, DC, 1985.


62

C H A R A C T E R I Z A T I O N A N D M E A S U R E M E N T OF FLAVOR C O M P O U N D S

those products, methyl glyoxal i s one of the most highly reactive compounds and readily undergoes secondary reactions to form some heterocyclic compounds [&)· A browning model system consisting of cysteamine and D-glucose produced numerous t h i a z o l i d i n e derivatives (9_) · It was proposed that D-glucose decomposed into short-chain carbonyls such as formaldehyde, acetaldehyde, glyoxal, and methyl glyoxal and subsequently reacted with cysteamine to give the corresponding t h i a z o l i d i n e derivatives. This suggests that cysteamine reacts r e a d i l y with carbonyl compounds to y i e l d t h i a z o l i d i n e derivatives i n an aqueous solution. Thiazoles, dehydrated products of t h i a z o l i d i n e , have been found i n various foods (10 ), but t h i a z o l i d i n e s have never been i s o l a t e d from any food samples. The formation of t h i a z o l i d i n e from formaldehyde and cysteamine was reported 50 years ago (11) · Some a l k y l t h i a z o l i d i n e s were prepared from cysteamine and aldehydes to synthesize 2-alkyl-N-nitrosothiazol i d i n e s for the mutagenicity study of nitrosamines (12)· Since direct analyses for formaldehyde and methyl glyoxal are d i f f i c u l t with gas chromatography (GC) or any other methods, we attempted to determine levels of formaldehyde and methyl glyoxal i n various food samples using t h e i r derivatives t h i a z o l i d i n e and 2a c e t y l t h i a z o l i d i n e , respectively. The proposed mechanism of t h i a z o l i d i n e formation from cysteamine and corresponding aldehydes i s shown i n Figure 1·

-NH

2

I

.C— R "SH

Figure 1.

NH

Ok

Proposed formation mechanism of t h i a z o l i d i n e s .

Literature Review Formaldehyde. Formaldehyde i s one of the most common aldehydes i n foods and the environment. Formaldehyde i s also an important a i r pollutant i n a variety of i n d u s t r i a l and domestic atmospheres (13)· Formaldehyde i s present i n various forms i n an aqueous medium. The simplest form i s free (monomeric) formaldehyde which i s known to be t o x i c . Other forms of formaldehyde, such as a polymer or a copolymer with other compounds, may produce toxic free formaldehyde under certain conditions. These forms of formaldehyde are categorized as follows:


5.

HAYASHI ET A L .

1. 2. 3.


4.

63

Volatile Aldehydes

formaldehyde adsorbed onto p a r t i c l e s . para (polymeric) formaldehyde. formaldehyde that has combined with other compounds to produce i d e n t i f i a b l e substances such as hexamine, which may be i n solution or absorbed onto p a r t i c l e s . formaldehyde that has combined with, for example, proteins which may then remain i n solution or be adsorbed onto other particles.

In spite of the increasing interest i n monitoring the levels of formaldehyde i n foods and the environment, there i s no s a t i s f a c t o r y , simple method available for the determination of trace quantities of formaldehyde. Many laboratories have attempted to develop a method for determining trace quantities of formaldehyde i n various samples. The conventional methods used most widely are shown i n Table I.

Table I.

Conventional methods used for formaldehyde analysis

Method

Determinand

Range of concentration (mg/1)

A. B.

Manual Manual

Free Total

0-2 0-10

C. D.

Automated Automated

Free Total

0-25 0-25 0-5

E.

Automated

Methanol

0-20

F.

Manual

Methanol

0-50

Basis

Spectrophotometry D i s t i l l a t i o n and spectrophotometry Spectrophotometry Direct hydrolysis and spectrophotometry D i s t i l l a t i o n and spectrophotometry Oxidation and spectrophotometry Gas chromatogrphy with FID

Free formaldehyde i s reacted with acetylacetone i n the presence of an excess of an ammonium s a l t to form the yellow fluorescent compound, 3,5-diacetyl-1,4-dihydrolutidine and subsequently determined spectrophotometrically i n methods A-E (14)· In these methods, the t e s t sample must be colorless and free from other carbonyl compounds. Some other derivatives have been used to analyze formaldehyde. For example, formaldehyde was reacted with sodium 4,5-dihydroxy-2,7-naphthalene disulfonate i n s u l f u r i c acid solution to y i e l d a purple color (580 nm) and then subjected to colorimetric analysis. A purple-colored pararosaniline derivative was used to analyze formaldehyde i n a i r (15)· A i r sample was passed through an aqueous solution which contained 0.4% of 3-methyl-2benzothiazolone hydrazone hydrochloride and then a dye produced was determined at 635 or 670 nm (16). Molecular sieve (1.6 mm pettet) was used to trap formaldehyde i n a i r samples. The formaldehyde



64


trapped on the molecular sieve was rinsed o f f with deionized water and subsequently determined c o l o r i m e t r i c a l l y as a pararosaniline derivative. The recovery of ppb levels was reported using t h i s method (17). A major drawback of these colorimetrie methods i s that many other compounds can i n t e r f e r e with the analysis. If a test solution i s contaminated with a compound having an absorption around 570-580 nm, s i g n i f i c a n t interference occurs. Formaldehyde i n a i r was sampled with s i l i c a gel coated with 2,4-dinitrophenylhydrazine (2,4-DNPH) and the r e s u l t i n g hydrazone was extracted with a c e t o n i t r i l e and determined by reverse-phase HPLC with UV detection at 340 nm. This method was validated over the range of 2.5-93.3 ug formaldehyde (18). An a i r sample was bubbled into an aqueous solution of 2,4-DNPH and the hydrazones that formed were determined by high performance l i q u i d chromatography (HPLC). Fourteen aldehydes including formaldehyde were analyzed i n an a i r sample using t h i s method ( 19 ) · Formaldehyde i n fresh shrimp was analyzed by HPLC as 2,4-DNPH derivative. Characteristics of t h i s method included an estimated detection l i m i t of 0.05 mg of formaldehyde/kg of shrimp, an average recovery of 72.3% at the 10 mg/kg l e v e l , and a t o t a l analysis time of 2 h. (20). Complete GC separation of the 2,4-DNPH derivatives of ten a l i p h a t i c aldehydes, eight a l i p h a t i c ketones and four aromatic aldehydes was obtained with a 20 m χ 0.25 mm i . d . glass c a p i l l a r y column coated with SF-96, with the exception of the derivatives of n-valeraldehyde and isobutyl methyl ketone, whose peaks overlapped, and the o- and mtolualdehyde derivatives, which were poorly separated (21). High resolution glass c a p i l l a r y GC separate syn-anti isomers of 2,4-DNPH derivative of v o l a t i l e aldehydes (22) · Sample c o l l e c t i o n from a automobile exhaust and d e r i v a t i z a t i o n were performed d i r e c t l y i n a midget impinger containing an a c e t o n i t r i l e solution of 2,4-DNPH and c a t a l y s t . This method allowed direct i n j e c t i o n of an aliquot of the sample into HPLC. The detection l i m i t for formaldehyde was 20 ppb with an analysis time as short as 10 min (23 ). Ambient a i r samples were collected on molecular sieve 13X absorbents at 2-h i n t e r v a l s , and the trapped formaldehyde was then determined by the mass fragmentograms of m/z 29 and m/z 30. The detection l i m i t of t h i s method was s u f f i c i e n t to quantify the low ppb levels of ambient formaldehyde i n r u r a l a i r (24)· Formaldehyde i n a i r reacted with Nbenzylethanolamine-coated Chromosorb 102 sorbent to produce a derivative of formaldehyde, 3-benzyloxazolidine. The oxazolidine recovered from the sorbent was determined using a GC equipped with a 25m fused s i l i c a c a p i l l a r y column coated with Carbowax 20M. The detectable range of t h i s method was from 0.55 to 4.71 mg/m (25). 3

Methyl glyoxal. Methyl glyoxal (pyruvaldehyde, 2-ketopropionic aldehyde, or 2-oxopropanal), i s a hygroscopic, yellow, mobile l i q u i d with a pungent, stinging odor, was f i r s t obtained by warming isonitrosoacetone with d i l u t e s u l f u r i c acid i n 1887 (26). As mentioned above, methyl glyoxal i s one of the sugar caramelization or decomposition products and i s often found i n heat-treated food. Methyl glyoxal and a number of other carbonyl compounds were i s o l a t e d from sucrose melted at 150°C, the temperature of baking bread crusts (27). One cup of instant coffee (lg/100 ml of water) and one cup of coffee prepared from ground coffee beans (8g/100 mL of water) contained 100-150 ug and 470-730 ug of methyl glyoxal,



5. HAYASHI ET A L .

Volatile Aldehydes

65

respectively (28)· Methyl glyoxal has been used i n minute traces i n imitation coffee, maple, honey, caramel and rum flavor (29). Direct analysis of methyl glyoxal by gas chromatography i s possible; however, i t may be present as a copolymer with other compounds i n foods so aht recovery e f f i c i e n c y of methyl glyoxal by organic solvent extraction has not been established. Methyl glyoxal and other dicarbonyl compounds i n cigarette smoke were analyzed with GC or HPLC as quinoxaline derivatives a f t e r reacting with ophenylenediamine. The levels of methyl glyoxal i n the smoke from commercial cigarettes were 33-70 ug/cigarett and 19-40 ug/cigarette for n o n - f i l t e r cigarettes and f i l t e r cigarettes, respectively (6)· This method was also used to determine l e v e l s of methyl glyoxal i n coffee samples (28)· There i s no simple method available for methyl glyoxal analysis at the present time. Experimental Materials. Cysteamine hydrochloride, formaldehyde (37% i n water), methyl glyoxal (40% i n water), and N-methylacetamide were purchased from the A l d r i c h Chemical Co., Milwaukee, WI. The extraction solvents (dichloromethane and chloroform) were obtained commercially and used without further treatment. Standard f a t t y aldehydes were obtained from r e l i a b l e commercial sources. I n s t r u m e n t a l A n a l y s i s * A Hewlett-packard Model 5880 A GC, equipped with thermionic s p e c i f i c detector and a 50 m χ 0.23 mm i . d . fused s i l i c a c a p i l l a r y column coated with Carbowax 20M, was used for quantitative analysis of t h i a z o l i d i n e and 2-acetylthiazolidine derived from formaldehyde and methyl glyoxal, respectively. GC peak areas were calculated with a HP 5880 A series GC integrator. The oven temperature was programmed from 70 t o 180°C at 2° C/min. A Finnigan Model 3200 combination GC/MS equipped with an INCOS MS data system was used for mass spectral i d e n t i f i c a t i o n of t h i a z o l i d i n e derivatives. GC analysis of v o l a t i l e aldehyde standards. A mixture of formaldehyde, acetaldehyde, propionaldehyde, isobutyl aldehyde, isovaleraldehyde, methyl glyoxal, and f u r f u r a l (0.1 mg each) were added t o 20 ml of cysteamine solution (6g/1 l i t e r of deionized water). The pH of the solution was adjusted to 8 with 6 Ν NaOH solution. The reaction proceeded promptly to formm t h i a z o l i d i n e derivatives. The reaction mixture was then extracted with 2 ml of dichloromethane, and an aliquot of the extract was injected i n the GC. A gas chromatoram of the extract i s shown i n Figure 2. Components were i d e n t i f i e d with GC/MS. A t h i a z o l i d i n e derivative i s e a s i l y i d e n t i f i e d using single ion. monitoring with m/z 88 ( t h i a z o l i d i n e r i n g - H). Table II shows mass spectra and GC retention data of t h i a z o l i d i n e derivatives. The gas chromatogram of the extract from the reaction mixture indicated that methyl glyoxal produced three products, 2-acetylthiazoline (peak #7), 2a c e t y l t h i a z o l i d i n e (peak #8), and 2-formyl-2-methylthiazolidine (peak #9).


66

C H A R A C T E R I Z A T I O N A N D M E A S U R E M E N T O F FLAVOR C O M P O U N D S

1


100

34

5

6

Φ _>

50H

ο

10

L

—ι— 20

10

30

40

Min.

Figure 2. Gas chromatogram of standard t h i a z o l i d i n e s . Peak 1 = 2 methylthiazolidine, 2 = t h i a z o l i d i n e , 3 = 2 - e t h y l t h i a z o l i d i n e , 4 = 2 - i s o p r o p y l t h i a z o l i d i n e , 5 = N-methylacetamide ( i n t e r n a i standard), 6 = 2 - i s o b u t y l t h i a z o l i d i n e , 7 = 2 - a c e t y l t h i a z o l i n e , 8 = 2-acetylt h i a z o l i d i n e , 9 = 2-forayl-2-methylthiazolidine, 10 = 2 - ( 2 - f u r y l ) thiazolidine. Table I I . The products of aldehydes and cysteamine and t h e i r spectral data

Peak # i n Figure 3

1 2 3 4 5 6 7 8 9 10

Products

2-Methylthiazolidine Thiazolidine 2-Ethylthiazolidine 2-Isopropylthiazolidine N-Methylacetamide 2-Isobutylthiazolidine 2-Acetylthiazoline 2-Acetylthiazolidine 2-Formy1-2-methy1thiazolidine 2-(2-furyl)thiazolidine

MS data

+

M M M M

+

+

+

M M M

+

+ +

= 103 (48, 88 (76), 56 (100)

89 (100), 88 (43), 59 (38) = 117 (21), 88 (100), 70 (21)

131 (5), 88 (100) = 145 (8), 88 (100), 56 (37)

129 (100), 101 (11), 60 (76) = 131 (4), 88 (100), 61 (27)

+

M = 13 (4), 101 (32), 60 (100) (tentative)


5.

HAYASHI ET A L .

67

Volatile Aldehydes

It would be i d e a l i f 2-acetylthiazolidine was formed exclusively from methyl glyoxal. The optimum reaction condition f o r 2-acetylthiazolidine, therefore, was determined by the following experiments :


Reaction temperature. The same molar r a t i o of methyl glyoxal and cysteamine was reacted at 0, 25, and 100°C. The results are shown i n Table I I I . The reaction at room temperature (20°C) gave the best results for 2-acetylthiazolidine formation.

Table I I I . Relative r a t i o of products from the reaction of methyl glyoxal and cysteamine at various temperature

Temperature (° C)

Products r a t i o (%) 2-Acetyl2-Formy1-2-methyl thiazolidine thiazolidine

2-Acetylthiazoline

0 25 100

81.0 8.5 2.0

19.0 89.0 40.0

0 2.5 58.0

The e f f e c t of molar r a t i o of methyl glyoxal and cysteamine. This was examined at room temperature and the results are shown i n Table IV. When the molar r a t i o of cysteamine and methyl glyoxal was 1000 at pH 8, 2-acetylthiazolidine was produced exclusively. On the other hand, formation of 2-acetylthiazolidine remained constant pH 6. Therefore, cysteamine was reacted with samples of interest at 25°C and i n a quantity to exceed 1000 f o l d the estimated amount of methyl f l y o x a l i n the following experiments, when experiment was cunducted at pH 8.

Table IV.

E f f e c t of reactants r a t i o on 2-acetylthiazolidine formation from cysteamine and methylglyoxal

Cysteamine Methylglyoxal (Molar ratio)

1 10 100 1000

Y i e l d of products (%) 2-Acetylthiazoline pH 6 pH 8

2.5 2.0 2.2 0.0

1.6 1.2 0.0 0.0

2-Acetylthiazolidine pH 6 pH 8

96.8 98.0 97.8 98.2

98.8 47.5 63.0 100.0

2-Formy1-2-methy1thiazolidine pH 6 pH 8

0.7 0.0 0.0 1.8


0.2 51.3 37.0 0.0


68

CHARACTERIZATION AND MEASUREMENT OF FLAVOR COMPOUNDS

E f f e c t of pH. The e f f e c t of pH on t h i a z o l i d i n e and 2a c e t y l t h i a z o l i d i n e was examined using solutions (250 ml) containing cysteamine (1.5 g)/formaldehyde (0.198 mg) and cysteamine (0.75 g)/methyl glyoxal (o.57 mg), respectively. The optimum recovery e f f i c i e n c y of t h i a z o l i d i n e and 2-acetylthiazolidine was obtained around pH 8. The test solutions were, therefore, adjusted to pH 8 r o i u t i n e l y . The recovery of t h i a z o l i d i n e reached i t s maximum at pH 8 and leveled off (Figure 3). Assuming that the reaction proceeds at 100% e f f i c i e n c y , 0.1975 mg of formaldehyde (MW = 30) should produce 0.5859 mg of t h i a z o l i d i n e (MW = 89). The maximum formaldehyde recovery calculated as t h i a z o l i d i n e from the c a l i b r a t i o n curve was 118% at pH 8. The reason(s) for t h i s excess recovery i s not yet understood. These data are averages of at least three r e p l i c a t i o n s . The correction can, therefore, be made e a s i l y for an actual analysis. In another series of experiments, methyl glyoxal was recovered s i g n i f i c a n t l y less from a solution of pH 6 than from a solution of pH 8.

10

11

PH

Figure 3.

E f f e c t of pH on t h i a z o l i d i n e recovery.

Preparation of c a l i b r a t i o n curve f o r formaldehyde a n a l y s i s . The c a l i b r a t i o n curve f o r t h i a z o l i d i n e (formaldehyde d e r i v a t i v e ) was prepared with N-methylacetamide as an i n t e r n a l standard.


5. HAYASHI ET AL.

Volatile Aldehydes

69


N-Methylacetamide (0.5 mg) was added to each standard chloroform solution of t h i a z o l i d i n e (0.125-1.00 mg). Gas chromatographic peak area r a t i o of t h i a z o l i d i n e and the standard were plotted against quantity of t h i a z o l i d i n e . A t y p i c a l c a l i b r a t i o n curve prepared f o r t h i a z o l i d i n e determination i s shown i n Figure 4.

Amount of Thiazolidine (mg)

Figure 4.

A gas chromatographic c a l i b r a t i o n curve f o r t h i a z o l i d i n e .

Preparation of c a l i b r a t i o n curve for methyl glyoxal analysis. The c a l i b r a t i o n curve for methyl glyoxas was prepared using Nmethylacetamide as an i n t e r n a l standard. N-Methylacetamide was added to each standard reaction mixture of methyl glyoxal (1.0-7.5 mg) and 0.75 g of cysteamine i n 70 ml of dichloromethane at pH 6. Gas chromatographic peak r a t i o of 2-acetylthiazolidine and the standard were plotted against quantity of methyl glyoxal i n the o r i g i n a l solution. Solvent Choice. When an aqueous solution of cysteamine (1g/250 ml, pH 8) was extracted with dichloromethane using a l i q u i d - l i q u i d continuous extractor for three hours, a certain amount of t h i a z o l i d i n e was i s o l a t e d and i d e n t i f i e d by GC/MS. The presence of t h i a z o l i d i n e i n a dichloromethane extract was observed even though the cysteamine solution was washed with ethyl acetate p r i o r to use. The t h i a z o l i d i n e also was i s o l a t e d when double d i s t i l l e d water was used instead of deionized water to prepare cysteamine solutions. Thiazolidine quantity decreased s i g n i f i c a n t l y when chloroform or ethyl acetate was used as an extraction solvent.


70

C H A R A C T E R I Z A T I O N A N D M E A S U R E M E N T O F FLAVOR C O M P O U N D S

These results indicated that dichloromethane contains certain quantities of formaldehyde as a contaminant. Formaldehyde contamination i n solvents was confirmed by the following experiment. Various concentrations of cysteamine solution were extracted using d i f f e r e n t quantities of dichloromethane. The conditions and results of t h i s experiment are show i n Table V.


Table V.

Effect of solvent volume and cysteamine quantity on t h i a z o l i d i n e recovery from solvents

Solvent

Volume (ml)

Dichloromethane Dichloromethane Dichloromethane Chloroform Chloroform

300 100 100 100 100

Amount of cysteamine added (g)

0.60 0.70 0.25 2.1 0.75

Thiazolidine recovered (mg)

14.0 5.8 4.3 0.033 0.033

When the quantity of dichloromethane was reduced from 300 ml to 100 ml, the amount of t h i a z o l i d i n e recovered decreased from 14 mg to 4.3-5.8 mg. When chloroform was used as a solvent, the amount of t h i a z o l i d i n e recovered remained constant (0.0033 mg) . Reagent grades of dichloromethane and chloroform obtained from various commercial sources were analyzed f o r formaldehyde i n order to choose the optimum solvent f o r further experiments. The results were shown i n Table VI.

Table VI.

Quantities of formaldehyde i n commercial dichloromethane and chloroform

Solvent

Source

Dichloromethane Dichloromethane Dichloromethane Dichloromethane Dichloromethane Chloroform Chloroform Chloroform Chloroform

/3

A

1 A Β C D 2

A

1 *2 3 Β A

Quantity of formaldehyde (ppm)

14.0 9.1 2.3 12.0 9.4 0.064 0.15 0.11 0.77

D i f f e r e n t batch



5.

HAYASHI ET A L .

Volatile Aldehydes

71

A l l dichloromethane examined showed 2-14 ppm of formaldehyde contamination. Several clean up methods were applied to remove formaldehyde such as washing with sodium b i s u l f i t e , treatment with active charcoal of Porapak Q porous polymer without success. Trace levels of formaldehyde i n solvents may be impossible to remove. Therefore, chloroform was used as the solvent for formaldehyde analysis i n further experiments. The amount of contaminant obtained from a blank solvent was always subtracted from the values of actual r e s u l t s . Dichloromethane was, however, used for methyl glyoxal analysis. The extraction e f f i c i e n c y of chloroform and d i c h l o r omethane was almost i d e n t i c a l . Dichloromethane was easier to use for a l i q u i d - l i q u i d continuous extraction than chloroform because of i t s lower b o i l i n g point. Sample preparations 1) D-Glucose the system consisting of methyl glyoxal and Dglucose was examined to determine interference of D-glucose i n t h i a z o l i d i n e and 2-acetylthiazolidine formation. The solutions containing formaldehyde, methyl glyoxal and D-glucose were treated with cysteamine. The reaction conditions are shown i n Table VII along with the recovery e f f i c i e n c i e s . 2) sauce.

Soy sauce

cysteamine (0.75g) was added to 20 ml of soy

3) Soy bean paste soy bean paste (20 g) was dissolved i n 200 ml of deionized water and 0.75 g of cysteamine was added. 4) Brewed coffee regular or decaffeinated coffee (80 g) was added to 1 l i t e r of b o i l i n g water. After 10 min., the coffee was f i l t e r e d and 2 50 ml of f i l t r a t e was reacted with 0.75 g of cysteamine. 5) Instant coffee, cocoa, instant tea, maple syrup, and nonfat dry milk instant coffee (3 g), cocoa (5 g), maple syrup (10 g), and nonfat dry milk (8 g) were dissolved i n 250 ml each of hot deionized water and each solution was reacted with 0.75 g of cysteamine. 6) Coke, wine, beer, orange juice, tomato juice, root beer, and apple juice Cysteamine (0.75 g) was added d i r e c t l y to 2 50 ml of each of these samples. A l l of the above solutions were s t i r r e d f o r 30 min with a magnetic s t i r r e r . The pH of the solutions was adjusted to 6 or 8 with 6N NaOH immediately after addition of cysteamine. The reaction mixtures were extracted with 70 ml of dichloromethane or chloroform for 6 h using a l i q u i d - l i q u i d continuous extractor. The extracts were dried over anhydrous sodium sulfate f o r 12 h. After the removal of sodium sulfate, 0.5 mg of N-methylacetamide was added to each solution as an i n t e r n a l standard. The extracts were quantitatively analyzed for t h i a z o l i d i n e and 2-acetylthaizolidine by GC.


72



Results and Discussion Preliminary experiments* The results of preliminary experiments showed that v o l a t i l e aldehydes reacted with cysteamine readily to give the corresponding t h i a z o l i d i n e derivative. Thiazolidine derivatives are much more stable than underivatized formaldehyde or methyl glyoxal. Solvent contaminants did not i n t e r f e r e with the GC analysis of t h i a z o l i d i n e . Moreover, because t h i a z o l i d i n e contains nitrogen, a highly sensitive and s e l e c t i v e thermionic detector i s applicable f o r analysis. When a food sample i s treated with cysteamine, some food constituents such as carbohydrate may i n t e r f e r e with t h i a z o l i d i n e or 2-acetylthiazolidine formation. DGlucose was chosen to represent a possible interference caused by food constituents. Recovery of formaldehyde and methyl glyoxal was reduced about 10% and 5% i n the presence of 10% D-glucose, respectively (Table VII). In the 10% D-glucose solution, the quantity of D-glucose was over 10,000 times that of formaldehyde of methyl glyoxal. Food constituents, such as D-glucose, apparently do not i n t e r f e r e s i g n i f i c a n t l y with t h i a z o l i d i n e or 2a c e t y l t h i a z o l i d i n e formation.

Table VII.

Recovery e f f i c i e n c y of t h i a z o l i d i n e and 2-acetylthiazo l i d i n e i n the presence of ^-glucose,

Aldehyde

Amount used (mg)

Concentration of glucose (%)

Methylglyoxal

0.57

0

Methylglyoxal

0.57

5

Methylglyoxal

0.57

10

Formaldehyde Formaldehyde

0.2 0.2

0 10

Product

Y i e l d of product (%)

2-Acety1thiazolidine 2-Acety1thiazolidine 2-Acety1thiazolidine Thiazolidine Thiazolidine

100 96 95 100 90

Analysis of formaldehyde and methyl glyoxal i n food samples. The chloroform extract of cysteamine-treated decaffeinated coffee showed obvious existence of aldehydes (Figure 5). In contrast to the gas chromatogram of cysteamine-untreated decaffeinated coffee (Figure 6), that of the treated sample showed new peaks; 1 (2methylthiazolidine, 2 ( t h i a z o l i d i n e ) , 3 (2-acetylthiazolidine), and 4 ( f u r f u r y l t h i a z o l i d i n e ) . The peak of the i n t e r n a l standard (S) did not i n t e r f e r e with any t h i a z o l i d i n e derivatives. The number of peaks appearing i n coffee extracts seemed to be too small but t h i s was due to high d i l u t i o n of the samples. The number of peaks appearing i n the coffee extracts seemed to be too small, compared to t h e i r number i n a previous study using concentrated samples (30, 31) instead of the solvent-diluted samples of the present study. The r e l a t i v e l y high d i l u t i o n of these samples would explain the smaller number of peaks. Among the coffee


5.

HAYASHI ET AL.

73

Volatile Aldehydes


1001

5(H

Lu

uni 10

20

30

Min.

Figure 5. A gas chromâtοgram of the chloroform extract of cysteamine-treated decaffeinated coffee. Peak 1 = 2-methylthiazo l i d i n e , 2 - t h i a z o l i d i n e , S = i n t e r n a l standard, 3 = 2-acetyl thiazolidine, 4 = 2-furfurylthiazolidine.

Figure 6. A gas chromatogram of the chloroform extract of cysteamine-untreated decaffeinated coffee.


74



v o l a t i l e s , nitrogen-containing compound such as pyrazines and pyrroles, may be, detected by the thermionic detector and appear as peaks on the chromatogram. The results of formaldehyde and methyl glyoxal analysis i n commercial foods are shown i n Table VIII. Formaldehyde was i d e n t i f i e d i n the levels of 3.7-17 ppm i n coffee obtained from various commercial sources. It was found at higher levels i n instant coffees than i n brewed coffee. This suggests that formaldehyde may escape from coffee during brewing. Formaldehyde has been reported i n coffee v o l a t i l e s by several researchers (31)· There are however, no reports on quantitative analysis of formaldehyde i n coffee p r i o r to the present study. Methyl glyoxal has never been reported i n soy sauce or soy bean paste p r i o r to this study. Certain aldehydes (acetaldehyde, npropanal, 2-methylpropanal, and 3-methylbutanal) were found i n soy sauce previously (32 ) · A gas chromatogram of the extract from cysteamine-treated soy sauce and untreated soy sauce are shown i n Figures 7 and 8.

Table VIII.

Formaldehyde and methylglyoxal contents i n foods

Content Formaldehyde pH 6

Food

Group I Brewed coffee Decaffeinated brewed coffee Instant coffee A Instant coffee Β Cocoa Instant tea Nonfat dry milk Soy Sauce A Soy Sauce Β Soy bean paste (miso) Group II Coke A Coke Β Root beer Beer Wine (white) Apple juice Orange juice Tomato juice Maple syrup

4.9 3.7 10.0 17.0 3.0 3.0 1.8 1.2 0.88 3.5

0.35 ND 0.44 0.08 0.096 0.12 0.15 0 1.5

(ppm) Methylglyoxal pH 6 pH 8

25.0 47.0 23.0 ND 1.2 2.4 1.4 7.6 3.0 0.7

0.23 0.24 0.76 0.084 0.11 0.26 0.04 0.064 2.5

ND:Data not available.


540 620 430 430 0 0 0 66 87 5.0

0.4 ND 2.3 0.57 0.28 0.31 0.39 0.11 12.0

5.

HAYASHI ET AL.

Volatile Aldehydes

75


The additional peaks i n Figure 7 represent t h i a z o l i d i n e derivatives. The content of methyl glyoxal i n soy bean paste was considerably lower than that of soy sauce even though they were prepared by s i m i l a r procedures. This may be due to differences i n fermentation time. Coffee contained the largest quantity of methyl glyoxal among the food samples tested. The values of instant coffee were less

lOO-i

50-

I— 10

20

30

Figure 7. A gas chromatogram of the chloroform e x t r a c t of cysteamine-treated soy sauce. Peak 1 = 2-methylthiazolidine, 2 = t h i a z o l i d i n e , 3 = 2 - i s o p r o p y l t h i a z o l i d i n e , 4 = i n t e r n a l standard (N-methylacetamide, 5 = i s o b u t y l t h i a z o l i d i n e , 6 = unknown t h i a z o l i d i n e derivate, 7 = 2-acetylthiazolidine.

than that of brewed coffee. This i s consistent with the r e s u l t s reported previously (28)· In the present study, one cup of instant coffee (1 g/100 ml) and brewed coffee (8 g/100 ml) contained 238 ug and 900-1030 ug of methyl glyoxal, respectively, when they were treated with cysteamine at pH 8. On the other hand, methyl glyoxal recovery was reduced s i g n i f i c a n t l y when the coffee was treated with cysteamine at pH 6. The same phenomenon was observed i n the case of soy sauce. It i s not clear why more methyl glyoxal was recovered at pH 8. Methyl glyoxal may e x i s t as a polymer or may form a complex,


76



100-1

LiliL ι— 10

20

30

Figure 8· A gas chromatogram of the chloroform extract of cysteamine-untreated soy sauce.

combining with an amino group of a large molecule such as a protein i n a lower pH s o l u t i o n . At a higher pH such as 8, a c e r t a i n amount of methyl glyoxal may be released from a polymer or a complex. I t was expected that foods containing a caramelized sugar such as coke and maple syrup would have more methyl glyoxal (33). However, the amount of methyl glyoxal detected i n these products was less than that of coffee or soy sauce. The samples examined i n t h i s study can be c l a s s i f i e d i n t o two groups. One i s foods consumed without a d d i t i o n a l water (Group I) and the other i s foods consumed with addition of a c e r t a i n amount of water (Group I I ) . I t i s important to know the amount of formaldehyde and methyl glyoxal intake when a food i s consumed. Table IX shows the calculated values of methyl glyoxal intake f o r each food item when i t i s consumed. Even a f t e r d i l u t i o n , coffee provided the most methyl glyoxal intake. I t i s d i f f i c u l t to estimate formaldehyde or methyl glyoxal intake from soy sauce because i t s use varies widely. Since formaldehyde and methyl glyoxal have been proven to be genotoxic and are found i n various foods and beverages i n somewhat s i g n i f i c a n t q u a n t i t i e s , i t i s important to study further t h e i r r i s k to human health.


5.

HAYASHI ET AL.

11

Volatile Aldehydes

Table IX· Calculated values of formaldehyde (FA) and methylglyoxal (MG) intake f o r each food when i t i s consumed

Amount consumed


Food

Amount of intake (ug) FA MG pH 8 pH 6 pH 6

Group I Brewed coffee 3g/180ml Decaffeinated coffee 3g/180ml Instant coffee A 1g/180ml Instant coffee Β 1g/180ml Cocoa 4g/180ml Instant tea 0.3g/180ml Nonfat dry milk 22.7g/240ml

14.8 11.2 10.1 17.1 12.1 0.9 30.6

75.6 140.4 22.7 ND 4.9 0.7 31.2

1620 1854 428 428 0 0 0

Group II Coke A Coke Β Root beer Beer Wine (white) Apple juice Orange juice Tomato juice

123.9 ND 155.8 28.4 9.6 36.0 53.1 0

81.4 85.0 269.0 29.7 11.0 78.0 14.2 11.3

141.6 ND 814.2 201.8 28.0 31.9 138.1 19.5

354ml/can 354ml/can 354ml/can 355ml/can 100ml/glass 300ml/glass 354ml/can 177ml/can

NDtData not available.

Acknowledgments The f i n a n c i a l support of t h i s study by the University of C a l i f o r n i a Cancer Research Coordinating Committee (#3-504017-19900) i s g r a t e f u l l y acknowledged.

Literature Cited 1. 2.

3. 4. 5. 6. 7. 8.

Chem. Eng. News 1981, 59, 35. Karns, W. D. "Long-Tern Inhalation Toxicity and Carcinogen icity Studies of formaldehyde in Rats and Mice." The Third Chemical Industry Institute of Toxicology Conference on Toxicology, Raleigh, NC, 1980. Wiseblatt, L.; Kbhn, F. E. Cereal Chem. 1960, 37, 55-56. Kajita, T.; Senda, M. Nippon Nogei Kagaku Kaishi. 1972, 46, 137-145. Hrdlicka, J.; Kuca, J.; Poult. Sci. 1965, 44, 27-31. Moree-Testa, P.; Saint-Jalm, T. J. Chromatogr. 1981, 217, 197-208. Hodge, J. E. Symp. Foods: Chem. Physiol. Flavors, Proc. 1967, p. 472. Ohnishi, S.; Shibamoto, T. J. Agric. Food Chem. 1984, 32, 987-992. Bills and Mussinan; Characterization and Measurement of Flavor Compounds ACS Symposium Series; American Chemical Society: Washington, DC, 1985.

78


9. 10. 11. 12.


13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

31. 32. 33.

Sakaguchi, Μ·; Shibamoto, T. J. Agric. Food Chem. 1978, 26, 1179-1183. Maga, J. A. Crit. Rev. Food Sci. Nutri. 1975, 153-176. Ratner, S.; Clarke, H. T. J. Am. Chem. Soc. 1937, 59, 200207. Umano, K.; Shibamoto, T.; Fernando, S. Y.; Wei, C.-I. Food Chem. Toxicol. 1984, 22, 253-259. Matthews, T. G.; Howell, T. C. Anal. Chem. 1982, 54, 14951498. Nash, T. Biochem. J. 1953, 55, 416. Miksch, R. R.; Anthon, D. W.; Hollowell, C. D.; Revzan, K.; Glanville, J. Anal. Chem. 1981, 53, 2118-2123. Sawicki, E.; Hauser, T. R.; Stanley, T. W.; Elbert, W. Anal. Chem. 1961, 33, 93-96. Georghiou, P. E.; Harlick, L.; Winsor, L.; Snow, D. Anal. Chem. 1983, 55, 567-570. Beasley, R. K.; Hoffmann, C. E.; Rueppel, M. L.; Worley, J. W. Anal. Chem. 1980, 52, 1110-1114. Kuwata, K; Uebori, M.; Yamasaki, Y. J. Chromatogr. Sci. 1979, 17, 264-268. Radford, T.; Dalsis, D. E. J. Agric. Food Chem. 1982, 30, 600602. Hoshika, Y; Takata, Y. J. Chromatogr. 1976, 120, 379-389. Uralets, V. P.; Rijks, J. Α.; Leclercq, P. A. J. Chromatogr. 1980, 194, 135-144. Lipari, F.; Swarin, S. J. J. Chromatogr. 1982, 247, 297-306. Yokouch, T.; Fuji, T.; Ambe, Y.; Fuwa, K. J. Chromatogr. 1979, 180, 133-138. Kennedy, E. R.; Hill, R. H. Jr. Anal. Chem. 1982, 54, 17391742. Pechmann, V. Chem. Ber. 1887, 20, 3213. Hodge, J. E. J. Agric. Food Chem. 1953, 1, 928-943. Kasai, H.; Rumeno, K.; Yamaizumi, Z.; Nishimura, S.; Nagao, M.; Fujita, Y.; Sugimura, T.; Nukaya, H.; Kosuge, T. Gann. 1982, 73, 681-683. Arctander, S. "Perfume and Flavor Chemicals"; Steffen Archtander: Elizabeth, N. J., 1969; p. 2 786. Shibamoto, T.; Harada, K.; Mihara, S.; Nishimura, O.; Yamaguchi, K.; Aitoku, Α.; Fukada, T. Application of HPIC for evaluation of coffee flavor quality. In the quality of food beverages. G. Charalambous, Ed. Academic Press, New York, 312. Gianturco, Μ. Α.; Giammarino, A. S.; Friedel, Nature 1966, 210, 1358. Nunomura, N.; Sakai, M.; Asano, Y.; Yokotsuka, T. Agric. Biol. Chem. 1976, 40, 485. Hodge, J. E. J. Agric. Food Chem. 1953, 1, 928.

RECEIVED June 24, 1985


Characterization and Measurement of Flavor Compounds - American

Recommend Documents