Quantitation of Nine Lactones in Dairy Cream by Stable Isotope

Nov 7, 2017 - In contrast, all γ-lactones were not much affected by the pasteurization process or were even reduced (γ-nonalactone). Literature data...
11 downloads 9 Views 551KB Size
Download PDF
Article pubs.acs.org/JAFC

Quantitation of Nine Lactones in Dairy Cream by Stable Isotope Dilution Assays Based on Novel Syntheses of Carbon-13-Labeled γ‑Lactones and Deuterium-Labeled δ‑Lactones in Combination with Comprehensive Two-Dimensional Gas Chromatography with Timeof-Flight Mass Spectrometry Jessica Schütt and Peter Schieberle* Deutsche Forschungsanstalt für Lebensmittelchemie Lise-Meitner-Straße 34, 85354 Freising, Germany S Supporting Information *

ABSTRACT: Lactones are well-known aroma compounds in, e.g., fruits and fermented foods as well as in dairy products, such as cream or milk powders. The latter are often used in confectionary products, e.g., milk chocolate. Lactones are suggested to contribute to the distinct aroma of dairy products and have also been reported in milk chocolate. However, data on their contribution to the overall aroma of this type of chocolate are scarce. As a result of their pH-dependent instability and their low volatility, a reliable quantitation of lactones is a challenge. Thus, to allow for a quantitation of nine lactones in one single comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry in electron ionization mode run, new synthetic routes were developed for five carbon-13-labeled γ-lactones and four deuterium-labeled δ-lactones, with the isotope label in the ring to be used in stable isotope dilution assays. The concentrations of the nine lactones were then analyzed in raw and pasteurized cream as well as in a heat-treated raw cream. δ-Dodecalactone and δ-decalactone showed the highest concentrations in both the raw and pasteurized cream. In the latter, δ-dodecalactone reached a 2.5-fold higher concentration compared to the raw cream. Subsequent heat treatments in a lab scale showed a further increase by factors of 13 and 19, respectively, suggesting a high potential of lactone precursors in cream. The results serve as a basis for further studies on lactone formation in other thermally processed products, such as milk chocolate. KEYWORDS: GC × GC−TOF−MS, stable isotope dilution assay (SIDA), [13C2]-labeled γ-lactones, [2H3−4]-labeled δ-lactones, cream

■

INTRODUCTION Lactones showing peach- or coconut-like aroma attributes are well-known odorants in foods, such as fruits or fermented products, as well as in dairy products. About 742 million tons of milk and dairy products are produced annually worldwide. The typical “milky” aroma of dairy products is often suggested to be caused by lactones,1−6 while older studies considered lactones as off-flavor compounds.7−10 The formation of lactones during heat treatment of butterfat is thought to occur from bound δhydroxy acids as precursors,5 and Kinsella et al.6 proposed a mechanism for δ-decalactone formation from milk fat triglycerides (Figure 1). Shiratsuchi et al.11 analyzed the aroma compounds in spray-dried skim milk powder and identified 10 lactones among the 196 volatiles characterized. They suggested a contribution of δ-decalactone, δ-undecalactone, γ-undecalactone, and γ-dodecalactone to the sweet and milky odor of skim milk powder.12 Kobayashi et al.13 analyzed heat-treated dairy products and found 10 lactones in an ultrahigh-temperature milk and 6 of them also in high-heat skim milk powder. However, although lactones are mentioned in many publications on dairy products,2,11−15 a reliable method for their quantitation in one single run by comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry in electron ionization mode (GC × GC−TOF−MS/EI) is not yet available. A stable © 2017 American Chemical Society

isotope dilution assay (SIDA) is a well-established means for reliable quantitation of volatiles in food matrices. However, isotopically labeled homologues of lactones are scarcely available. In the last few decades, several syntheses for isotopically labeled lactones have been published; varying levels of success have been reported; and often the label was introduced in the side chain.16−19 However, labeling the alkyl chain often described in the literature is not appropriate for mass spectrometry with electron ionization (MS/EI) measurements (Figure 2), because the analyte and internal standard show the same base peak (m/z 85) of the lactone ring, while the molecular ions m/z 128 (Figure 2A) and m/z 130 (Figure 2B) are too weak to be used for quantitation. Thus, such labeled homologues can only be used in mass spectrometry with chemical ionization (MS/CI). Therefore, the aim of this study was to develop new SIDA for lactone quantitation on the basis of the synthesis of isotopically labeled γ- and δ-lactones, which can be used in MS/EI as well as MS/CI measurements. The method was applied on raw and thermally processed cream (i) to confirm the presence of lactone precursors in raw cream Received: Revised: Accepted: Published: 10534

September 21, 2017 October 31, 2017 November 6, 2017 November 7, 2017 DOI: 10.1021/acs.jafc.7b04407 J. Agric. Food Chem. 2017, 65, 10534−10541

Article

Journal of Agricultural and Food Chemistry

Figure 1. Proposed mechanism6 of δ-lactone formation from milk fat triglycerides.

Figure 2. Mass spectra (MS/EI) of (A) γ-decalactone and (B) [2H2]-γ-decalactone bearing the deuterium atoms in the side chain. freezing overnight, the raw product was filtered off and the solvent was evaporated (yield of 50%). [13C2]-Ethyl 2-Bromoacetate. [13C2]-Bromoacetic acid (1.12 g, 7.94 mmol) and ethanol (2.2 g, 47.8 mmol, containing 0.13% sulfuric acid) were stirred for 2 h at 82 °C. The organic phase was washed 3 times with water (50 mL) and dried over anhydrous sodium sulfate. After solvent evaporation, the slightly brown [13C2]-ethyl bromoacetate was obtained (yield of 46%). [13C2]-Ethyl 2-Iodoacetate. For the bromine−iodine exchange, [13C2]-ethyl 2-bromoacetate (1.238 g, 7.33 mmol), sodium iodide (2.225 g, 14.83 mmol), and acetone (50 mL) were refluxed at 80 °C for 2 h under nitrogen.23 After cooling to room temperature, the solution was filtered, the solvent was evaporated, and the residue was taken up in diethyl ether. The solution was washed with brine and dried over anhydrous sodium sulfate; diethyl ether was evaporated; and [13C2]-ethyl 2-iodoacetate was obtained in a yield of 38%. [13C2]-γ-Decalactone. [13C2]-Ethyl 2-iodoacetate (0.2048 g, 0.95 mmol), 1-octene (0.4583 g, 4.092 mmol), and copper (0.3203 g, 5.04 mmol) were mixed under an atmosphere of nitrogen and stirred at 132 °C for 8 h.24 The residue was dissolved in diethyl ether and dried over anhydrous sodium sulfate. After solvent evaporation, the raw product was purified by column chromatography (Table 1). [13C2]-γDecalactone was obtained in a purity of 99% in fractions Vb and VIa. MS/EI, m/z (%): 87 (100), 41 (13), 58 (10), and 43 (8). MS/CI, m/z (%): 173 (100) and 155 (67). The retention index and odor quality were identical with the data of the unlabeled compound (Table 2). [13C2]-γ-Undecalactone. By application of the same approach as described for [13C2]-γ-decalactone, [13C2]-γ-undecalactone was synthesized by stirring [13C2]-ethyl 2-iodoacetate (0.2051 g, 0.948 mmol), 1-nonene (0.6032 g, 4.79 mmol), and copper (0.3928 g, 6.18 mmol) at 147 °C for 8 h in a nitrogen atmosphere. Isolation/

and (ii) as a basis for use in further studies on the impact of dairy products influencing the distinct aroma of confectionary products, such as milk chocolate.

■

MATERIALS AND METHODS

Cream Samples. Untreated raw cream (RC) and pasteurized cream (PC), with a fat content of 34.7%, were provided from a local dairy. Care was taken that the cream was from the production line. Heat treatment of raw cream was performed at 80 °C for 18 h in sealed glass tubes in the lab. Chemicals and Reagents. γ-Octalactone, γ-nonalactone, δnonalactone, γ-decalactone, δ-decalactone, γ-undecalactone, γ-dodecalactone, and δ-dodecalactone were obtained from Aldrich (SigmaAldrich-Chemie, Taufkirchen, Germany). δ-Octalactone was purchased from Alfa Aesar (Karlsruhe, Germany). [2H4]-3-Bromopropionic acid was purchased from CDN Isotopes (Quebec, Canada). Other chemicals were from either Sigma-Aldrich (Taufkirchen, Germany) or VWR (Darmstadt, Germany). Diethyl ether and npentane were freshly distilled using a column (120 × 5 cm) packed with Raschig rings. To purify and activate silica gel 60 (0.063−0.2 mm) a previously published method20 was used. Syntheses. [13C2]-γ-Octalactone to [13C2]-γ-Dodecalactone. A four-step synthetic route was developed to synthesize all five [13C2]-γlactones from one labeled intermediate starting from [13C2]-acetic acid. [13C2]-Bromoacetic Acid. Trifluoroacetic anhydride (5.513 g, 26.3 mmol) was carefully added dropwise to magnetically stirred [13C2]acetic acid (0.9834 g, 15.9 mmol) at 0 °C. After the addition of phosphorus tribromide (0.06 g, 0.22 mmol), the solution was refluxed at 60 °C and bromine (0.83 mL, 32.4 mmol) was added dropwise. The reaction was finished after the brown color disappeared.21,22 After 10535

DOI: 10.1021/acs.jafc.7b04407 J. Agric. Food Chem. 2017, 65, 10534−10541

Article

Journal of Agricultural and Food Chemistry

[13C2]-γ-Octalactone was obtained in a purity of 99%. MS/EI, m/z (%): 87 (100), 58 (12), 41 (10), and 57 (6). MS/CI, m/z (%): 145 (100), 127 (50), and 98 (7). [13C2]-γ-Nonalactone. By application of the same approach used for the synthesis of [13C2]-γ-octalactone, [13C2]-γ-nonalactone was synthesized by heating [13C2]-ethyl 2-iodoacetate (0.2048 g, 0.948 mmol), 1-heptene (0.4583 g, 4.67 mmol), and copper (0.3203 g, 5.04 mmol). MS/EI, m/z (%): 87 (100), 41 (10), 58 (9), and 43 (8). MS/CI, m/z (%): 159 (100), 141 (62), and 123 (5). [2H3−4]-δ-Octal-, [2H3−4]-δ-Nona-, [2H3−4]-δ-Deca-, and [2H3−4]δ-Dodecalactone. To synthesize the four deuterium-labeled δlactones, a three-step synthetic route starting from [2H4]-ethyl 3bromopropionic acid was developed. During synthesis, the deuterium label was partially lost but mainly a mixture of the [2H4]- and [2H3]isotopologues was generated. [2H4]-Ethyl 3-Bromopropionate. [2H4]-3-Bromopropionic acid (1.0175 g, 6.48 mmol) and ethanol (7.0182 g, 153 mmol) containing sulfuric acid (0.102 g in 100 mL of ethanol) were stirred for 2 h at 82 °C. After washing with water (50 mL), the organic phase was dried over sodium sulfate. The solvent was evaporated and [2H4]-ethyl 3bromopropionate was obtained (yield of 98%). [2H4]-Ethyl 3-Iodopropionate. [2H4]-ethyl 3-bromopropionate (1.1714 g, 6.33 mmol), potassium iodide (2.2063 g, 13.29 mmol), and acetone (50 mL) were refluxed at 80 °C for 2 h under nitrogen.23 After cooling overnight, potassium bromide was removed by filtration. Then, acetone was evaporated, and the raw product was diluted in diethyl ether, washed with an aqueous saturated sodium chloride solution, and dried over anhydrous sodium sulfate. The solvent was evaporated, and the target compound was obtained in a yield of 79%. [2H3−4]-δ-Octalactone. [2H4]-Ethyl 3-iodopropionate (0.2069 g, 0.892 mmol), 1-pentene (1.6852 g, 24.1 mmol), and copper (0.2662 g, 4.19 mmol) were mixed in a glass-lined autoclave and heated to 132 °C for 24 h under nitrogen. After dissolving the residue in diethyl ether, the solution was dried over sodium sulfate. To purify [2H3−4]-δoctalactone formed, the residue was subjected to column chromatography (Table 1). MS/EI, m/z (%): 103 (100), 74 (71), 75 (69), and 102 (67). MS/CI, m/z (%): 147 (100), 146 (65), 129 (34), and 128 (30). Retention indices and odor qualities of all four lactones were identical with those of the respective unlabeled lactones (Table 2) [2H3−4]-δ-Nonalactone. By application of the same approach used for the synthesis of [2H3−4]-δ-octalactone, [2H3−4]-δ-nonalactone was prepared by heating [2H4]-ethyl 3-iodopropionate (0.2133 g, 0.919 mmol), 1-hexene (1.4465 g, 17.2 mmol), and copper (0.3639 g, 5.73 mmol) in an atmosphere of nitrogen. MS/EI, m/z (%): 103 (100), 75 (73), 44 (54), and 102 (52). MS/CI, m/z (%): 161 (100), 160 (60), 143 (41), and 142 (30). [2H3−4]-δ-Decalactone. Using the same method as described above, [2H4]-ethyl 3-iodopropionate (0.3072 g, 1.324 mmol), 1-heptene (1.8774 g, 19.2 mmol), and copper (0.4224 g, 6.65 mmol) were heated in an autoclave for 24 h to obtain [2H3−4]-δ-decalactone. MS/EI, m/z (%): 103 (100), 102 (60), 75 (59), and 54 (54). MS/CI, m/z (%): 175 (100), 174 (50), 157 (46), and 156 (30). [2H3−4]-δ-Dodecalactone. [2H3−4]-δ-Dodecalactone was prepared from [2H4]-ethyl 3-iodopropionate (0.2033 g, 0.876 mmol), 1-nonene (1.4928 g, 11.9 mmol), and copper (0.2953 g, 4.65 mmol). The conditions used were the same as given above. MS/EI, m/z (%): 103 (100), 75 (64), 74 (56), and 44 (55). MS/CI, m/z (%): 203 (100), 202 (92), 185 (43), and 184 (39). Column Chromatography. A water-cooled (10 °C) glass column (20 × 1 cm inner diameter) was filled with purified and activated silica gel 60 (0.063−0.2 mm, 10 g)20 and was conditioned with 50 mL of npentane. After application of the raw product, elution was performed with a constant solvent flow of 3 mL/min using the elution order shown in Table 1. Fractions V and VI were split in 2 × 25 mL. The fractions obtained were dried over anhydrous sodium sulfate, filtered, and analyzed by gas chromatography−olfactometry (GC−O)/MS. Fractions containing the target compound were combined, and the solvent was evaporated under reduced pressure.

Table 1. Elution Order and Fractions Used for the Purification of γ- and δ-Lactones by Column Chromatography on Silica solvent fraction

volume (mL)

pentane (%)

diethyl ether (%)

I II III IV Va Vb VIa VIb VII VIII IX X

50 50 50 50 25 25 25 25 50 50 50 50

100 95 90 85 80 80 75 75 70 65 50 0

0 5 10 15 20 20 25 25 30 35 50 100

Table 2. Odor Qualitiy (OQ), Molecular Weight (MW), and Retention Index (RI) of Unlabeled and Labeled [13C2]-γLactones and [2H3−4]-δ-Lactones RI onb lactone γ-octalactone [13C2]-γ-octalactone γ-nonalactone [13C2]-γ-nonalactone γ-decalactone [13C2]-γ-decalactone γ-undecalactone [13C2]-γ-undecalactone γ-dodecalactone [13C2]-γ-dodecalactone δ-octalactone [2H3−4]-δ-octalactone δ-nonalactone [2H3−4]-δ-nonalactone δ-decalactone [2H3−4]-δ-decalactone δ-dodecalactone [2H3−4]-δ-dodecalactone

OQa coconut-like coconut-like peach-like and coconut-like peach-like peach-like coconut-like coconut-like coconut-like peach-like and coconut-like

MW

FFAP

DB-5

142 144 156 158 170 172 184 186 198 200 142 146 156 160 170 174 198 202

1916

1257

2026

1365

2141

1471

2259

1600

2372

1684

1976

1284

2065

1350

2191

1497

2426

1709

a Odor quality perceived at the sniffing port. bRetention index determined in comparison to a homologous series of n-alkanes.

purification of the target compound by column chromatography was performed as described above. MS/EI, m/z (%): 87 (100), 41 (17), 58 (10), and 55 (10). MS/CI, m/z (%): 187 (100), 169 (61), and 151 (21). [13C2]-γ-Dodecalactone. The procedure used for [13C2]-γ-decalactone was also used to synthesize [13C2]-γ-dodecalactone. Under nitrogen, [13C2]-ethyl 2-iodoacetate (0.2073 g, 0.959 mmol), 1-decene (0.6032 g, 4.31 mmol), and copper (0.4120 g, 6.48 mmol) were stirred at 172 °C for 8 h. MS/EI, m/z (%): 87 (100), 41 (21), 55 (12), and 43 (12). MS/CI, m/z (%): 201 (100), 183 (58), and 165 (23). [13C2]-γ-Octalactone. As a result of the low boiling point of 1hexene (63 °C), the method described above was modified as follows: [13C2]-ethyl 2-iodoacetate (0.2067 g, 0.957 mmol), 1-hexene (0.7878 g, 9.38 mmol), and copper (0.3222 g, 5.07 mmol) were mixed under nitrogen in a glass-lined autoclave and heated to 132 °C for 8 h. The residue was dissolved in diethyl ether and dried over anhydrous sodium sulfate, and column chromatography was used to purify the raw product (Table 1). 10536

DOI: 10.1021/acs.jafc.7b04407 J. Agric. Food Chem. 2017, 65, 10534−10541

Article

Journal of Agricultural and Food Chemistry

DB-5 (2 m × 0.15 mm inner diameter, 0.30 μm film thickness, Agilent, Santa Clara, CA, U.S.A.) was used. For the first oven, the starting temperature of 40 °C was held for 2 min, then raised by 6 °C/min to 230 °C, and held for 7 min. The second oven with a starting temperature of 60 °C was held for 3 min, then raised by 6 °C/min to 250 °C, and held for 10 min. Data analysis was performed using GC Image (GC Image, Lincoln, NE, U.S.A.).

GC−O. The odor quality of the lactones was determined on a HRGC 5160 Mega Series gas chromatograph (Carlo Erba Instruments, Hofheim, Germany). The gas flow was divided at the end of the capillary column using a deactivated Y-type splitter made of glass (Chrompack, Frankfurt, Germany) that passed 50% of the eluate into the sniffing port (200 °C) and 50% to the flame ionization detector (FID) (250 °C) using two uncoated fused silica capillaries (25 cm × 0.20 mm inner diameter). For identification experiments, two fused silica capillary columns of different polarity were used: a J&W Scientific DB-FFAP, 30 m × 0.32 mm inner diameter, 0.25 μm film thickness (Agilent Technologies, Inc., Santa Clara, CA, U.S.A.), and a J&W Scientific DB-5, 60 m × 0.32 mm inner diameter, 0.25 μm film thickness (Agilent Technologies, Inc., Santa Clara, CA, U.S.A.). The linear retention indices of the labeled standards were calculated using n-alkanes (C6−C26 for DB-FFAP and C6−C18 for DB-5) as described previously.25 Using helium as the carrier gas, separation was performed after cold on-column sample application at 40 °C and by means of the following temperature program: 40 °C held for 2 min, then raised at 6 °C/min to 190 °C, and finally raised at 8 °C/min to 230 °C. Isolation and Quantitation of Cream Volatiles. Cream samples (2 g) were thoroughly mixed with diethyl ether (50 mL) and spiked with defined amounts of all nine internal standards. The concentrations of the labeled compounds lay between 0.1- and 5fold of the amount of the analyzed volatile compounds, as determined in preliminary experiments. The suspension was stirred at room temperature for 2 h and was then subjected to solvent-assisted flavor evaporation (SAFE) for volatile isolation.26 To avoid co-elutions on the GC column, the SAFE distillate was separated into two fractions of acidic and neutral/basic compounds by extraction with an aqueous solution of sodium bicarbonate (0.5 M, total of 70 mL).25 Both extracts were concentrated by means of a Vigreux column (60 × 1 cm) to 1 mL and finally reduced to 0.2 mL25 and analyzed by GC−O followed by comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry (GC × GC−TOF−MS). For the determination of the response factors (RFs; Table 3), mixtures of known concentrations of labeled standard and analyte were analyzed using the fragments for the unlabeled and labeled lactones.

■

RESULTS AND DISCUSSION Syntheses. For the development of SIDAs, nine isotopically labeled lactones were synthesized to allow for quantitation of all desired lactones in MS/EI. The selection of lactones was based on previous data reported for lactones in dairy products. Because lactones labeled in the alkyl side chain are not appropriate for MS/EI (Figure 2), a new synthetic approach was developed to introduce either deuterium or carbon-13 labeling in the lactone ring. For the synthesis of the labeled γlactones, [13C2]-acetic acid (1 in Figure 3) was used as the starting material. After bromination (2 in Figure 3), esterification (3 in Figure 3), and iodination, [13C2]-ethyl 3iodoacetate (4 in Figure 3) was obtained. This ester could then be used as a starting material for all considered γ-lactones by reacting it with the corresponding 1-alkenes to generate the desired side chain. As an example, the mass spectrum of unlabeled γ-octalactone is compared to the spectrum of isotopically labeled [13C2]-γ-octalactone in Figure 4. The base peak (MS/EI) showed a clear shift by two mass units. The ring fragment m/z 85 in unlabeled lactone (Figure 4A) was shifted to m/z 87 in the carbon-13-labeled lactone (Figure 4B). MS/CI of [13C2]-γ-octalactone revealed a base ion for the molecular ion m/z 145, confirming the introduction of two carbon-13 atoms. In addition, the same retention index as compared to unlabeled lactone as well as the same odor quality (Table 2) confirmed the successful synthesis. The mass spectra of the further four isotopically labeled lactones are given in the Supporting Information. The synthesis of δ-lactones followed a similar principle but using deuterium labeling instead of carbon-13 labeling. First, deuterated [2H4]-ethyl 3-iodoacetate (3 in Figure 5) was synthesized starting from [2H4]-bromopropionic acid (1 in Figure 5). The iodo ester then served as the common intermediate for the synthesis of all four lactones, again using 1-alkenes to generate the desired side chain. During synthesis, the deutetrium label was partially lost. As exemplarily shown for δ-octalactone, the base peak of unlabeled lactone m/z 99 (Figure 6A) showed a shift by 3−4 units with m/z 103 and 102 as the two main fragments in the labeled lactone (Figure 6B). MS/CI finally confirmed the incorporation of mainly three- and four-labeled deuterium atoms in the lactone. Further data on the retention index and odor quality (Table 2) corroborated the successful synthesis of the labeled isotopologues. The mass spectra of the three further labeled δ-lactones are shown in the Supporting Information, all showing a labeling with mainly three to four deuterium atoms in the ring. Mass Spectrometric Response Curve. Five mixtures containing the nine labeled lactones and the respective unlabeled lactones in varying concentrations were added to fat-free milk and worked up, as described below for the cream samples. Using the fragments shown in Table 3, regression lines were calculated for each lactone. The curve obtained for γoctalactone shown in Figure 7 indicated very good linearity. Factors calculated for the further eight lactones are summarized in Table 3.

Table 3. Mass Traces of the Nine Lactones (Analyte) and Respective Labeled Isotopologues (Internal Standard) Used in the Quantitation by SIDAsa mass trace (m/z)b lactone γ-octalactone γ-nonalactone γ-decalactone γ-undecalactone γ-dodecalactone δ-octalactone δ-nonalactone δ-decalactone δ-dodecalactone

internal standard 13

C2 C2 13 C2 13 C2 13 C2 2 H3−4 2 H3−4 2 H3−4 2 H3−4 13

analyte

internal standard

RFc

R2

85 85 85 85 85 99 99 99 99

87 87 87 87 87 102 + 103 102 + 103 102 + 103 102 + 103

1.0264 1.0528 1.0847 0.9411 1.1857 1.2506 0.9336 0.9031 0.8463

0.9998 0.9994 0.9997 1 0.9996 1 0.9999 0.9999 0.9995

a

RFs were calculated from defined mixtures of analytes and internal standards. bMass trace obtained by MS/EI. cMS response factor. Quantitation by GC × GC−TOF−MS. For quantitation, a LECO Pegasus 4D time-of-flight mass spectrometer (TOF−MS) equipped with an Agilent 7890A gas chromatograph as the first GC oven and with a two-stage modulator for cryogenic temperatures using a modulation time of 4 s was used (LECO Corporation, St Joseph, MI, U.S.A.). To generate mass spectra in the electron ionization mode (MS/EI), the spectrometer was used with an electron energy of 70 eV. In the first dimension, a J&W Scientific DB-FFAP column (25 m × 0.25 mm inner diameter, 0.25 μm film thickness, Agilent, Santa Clara, CA, U.S.A.) was used, and in the second dimension, a J&W Scientific 10537

DOI: 10.1021/acs.jafc.7b04407 J. Agric. Food Chem. 2017, 65, 10534−10541

Article

Journal of Agricultural and Food Chemistry

Figure 3. Synthetic route used in the preparation of [13C2]-γ-octa-, [13C2]-γ-nona, [13C2]-γ-deca-, [13C2]-γ-undeca-, and [13C2]-γ-dodecalactone.

Figure 4. Mass spectra (MS/EI) of (A) γ-octalactone and (B) [13C2]-γ-octalactone.

than in the raw cream. In contrast, all γ-lactones were not much affected by the pasteurization process or were even reduced (γnonalactone). Literature data on lactone concentrations are mainly available for butter or milk. Peterson and Reineccius3 reported δ-octalactone and δ-decalactone with 72.8 and 1193 μg/kg, respectively, in sweet cream salted butter. Schieberle et al.4 quantitated δ-decalactone in five different types of butter, with concentrations between 2150 and 5000 μg/kg, and Czerny and Schieberle27 found concentrations of 41 and 300 μg/kg for δ-octalactone and δ-decalactone in ultrahigh-temperature processed milk. Because a thermal treatment obviously influences the concentrations of lactones in dairy products, the differences in the concentrations between the studies might be explained by a formation from precursors during heat treatment of the dairy products analyzed. Furthermore, it has to be taken into account that lactones are most likely located in the fat phase of the dairy product.

Quantitation of Lactones in Cream. In general, pasteurization is applied to obtain longer best before dates for a food. However, any heat treatment is known to influence the set of volatile aroma compounds by either degradation reactions or formation reactions from precursors present in the raw material. To obtain insight into the influence of heat treatment on the lactone concentrations in raw cream, the newly synthesized labeled internal standards were used for quantitation. All lactones were already present in the raw cream, with the highest concentrations for δ-dodecalactone (1030 μg/kg), γ-dodecalactone (707 μg/kg), and δ-decalactone (226 μg/kg) (Table 4). The other lactones were present in lower concentrations. In the pasteurized cream from the same batch, the highest concentrations were measured for δdodecalactone (2620 μg/kg), γ-dodecalactone (839 μg/kg), and δ-decalactone (551 μg/kg) (Table 4). While the concentrations of three even-numbered δ-lactones were doubled in the pasteurized cream, δ-nonalactone was lower 10538

DOI: 10.1021/acs.jafc.7b04407 J. Agric. Food Chem. 2017, 65, 10534−10541

Article

Journal of Agricultural and Food Chemistry

Figure 5. Synthetic route used in the preparation of [2H3−4]-δ-octalactone to [2H3−4]-δ-dodecalactone (lactones C8−C12).

Figure 6. Mass spectra (MS/EI) of (A) δ-octalactone and (B) [2H3−4]-δ-octalactone.

Table 4. Concentrations (μg/kg) of Nine Lactones in Raw Cream (RC) and Pasteurized Cream (PC) RC

Figure 7. Response curve used for the determination of the response factor for γ-octalactone. a

Lactones are thought to be released from triglyceride-bound hydroxy fatty acids during heat treatment.5 A previously postulated mechanism for δ-lactone formation from milk fat triglycerides (Figure 1) suggests the hydrolysis of triglycerides

PC

lactone

concentrationa

RSD (%)

concentrationa

RSD (%)

γ-octalactone γ-nonalactone γ-decalactone γ-undecalactone γ-dodecalactone δ-octalactone δ-nonalactone δ-decalactone δ-dodecalactone

17.7 58.6 100 115 707 26.5 17.9 226 1030

1.97 15.5 15.0 8.50 1.19 2.18 3.88 6.82 9.17

11.9 50.9 48.0 120 839 68.8 11.7 551 2620

4.20 2.43 3.21 7.10 11.2 8.55 4.34 10.4 11.2

Mean value of at least three work-ups.

during heat treatment, causing the release of δ-hydroxy acids and their subsequent conversion to the corresponding 10539

DOI: 10.1021/acs.jafc.7b04407 J. Agric. Food Chem. 2017, 65, 10534−10541

Journal of Agricultural and Food Chemistry

■

lactones.6 This mechanism, although not yet confirmed, may explain that only the even-numbered δ-lactones showed clear increases in their concentrations in the pasteurized cream compared to raw cream. Obviously, the corresponding γhydroxy fatty acids are not present in the milk fat because the γlactones were not increased during the pasteurization. Heat Treatment of Cream. To obtain deeper insight into the precursor potential for lactone formation from milk fat, raw cream was heated in the lab scale at 80 °C for 18 h. Interestingly, the heat treatment did not much affect the concentrations of all γ-lactones, except γ-dodecalactone, which increased by a factor of more than 6 to reach 4480 μg/kg (Table 5). In contrast, all δ-lactones increased significantly,

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jafc.7b04407. Mass spectra (A, MS/CI; B, MS/EI) of [13C2]-γoctalactone (Figure S1), mass spectra (A, MS/CI; B MS/EI) of [13C2]-γ-nonalactone (Figure S2), mass spectra (A, MS/CI; B, MS/EI) of [13C2]-γ-decalactone (Figure S3), mass spectra (A, MS/CI; B, MS/EI) of [13C2]-γ-undecalactone (Figure S4), mass spectra (A, MS/CI; B, MS/EI) of [13C2]-γ-dodecalactone (Figure S5), mass spectra (A, MS/CI; B, MS/EI) of [2H3−4]-δoctalactone (Figure S6), mass spectra (A, MS/CI; B, MS/EI) of [2H3−4]-δ-nonalactone (Figure S7), mass spectra (A, MS/CI; B, MS/EI) of [2H3−4]-δ-decalactone (Figure S8), and mass spectra (A, MS/CI; B, MS/EI) of [2H3−4]-δ-dodecalactone (Figure S9) (PDF)

difference

lactone

RC

HT-RC

μg/kg

%

γ-octalactone γ-nonalactone γ-decalactone γ-undecalactone γ-dodecalactone δ-octalactone δ-nonalactone δ-decalactone δ-dodecalactone

17.7 58.6 100 115 707 26.5 17.9 226 1030

16.1 92.0 86.4 107 4480 513 22.1 4380 13500

1.60 33.4 13.6 8.00 3773 487 4.2 4154 12470

−9.04 −56.9 −13.6 −6.96 534 1840 23.5 1840 1210

ASSOCIATED CONTENT

S Supporting Information *

Table 5. Differences in Lactone Concentrations (μg/kg) in Raw Cream before and after Heat Treatment (HT-RC) concentrationa

Article

■

AUTHOR INFORMATION

Corresponding Author

*Fax: +49-8161-71-2970. E-mail: [email protected]. ORCID

Peter Schieberle: 0000-0003-4153-2727 Notes

The authors declare no competing financial interest.

■

a

Mean value of at least three work-ups, with a standard deviation of