4136 J. Agric. Food Chem., Vol. 56, No. 11, 2008
Bravo et al.
with ethyl acetate (300 mL), followed by ethyl acetate/methanol (1:1,
500 mL) affording the target compound as a crude material. Further
fractionation by semipreparative RP-HPLC allowed the isolation and
purification of the target compound as well as 2-(2′-hydroxyethyl)-3-
methyl quinoxaline (1,4-DDP quinoxaline) and 2-(2′,3′-dihydroxypro-
pyl) quinoxaline (3-DP quinoxaline). Separation was performed
isocratically using acetonitrile/water (4:96, v/v). The target quinoxaline
eluted at tR ) 20-22 min. Acetonitrile was eliminated by evaporation
instrument. The R-dicarbonyls methylglyoxal, 2,3-butanedione, 1,4-
DDP, 1-DP, and 1,4-DDH were detected as their corresponding
quinoxaline derivatives at the tR given in parentheses: 2-methylqui-
noxaline (Aldrich) (10.2 min), 2,3-dimethylquinoxaline (Aldrich) (12.3
min), 1,4-DDP quinoxaline (synthetic material) (19.7 min), 1-DP
quinoxaline (synthetic material) (21.4 min), and 1,4-DDH quinoxaline
(synthetic material) (23.9 min). Calibration curves were prepared by
treating standard solutions containing known amounts of each pure,
synthetic reference compound as described previously for beer samples.
The peak areas were determined at the target ions m/z 144, 158, 170,
245, and 272, respectively, for the mentioned quinoxalines and m/z
158 for the IS quinoxaline.
Quantification of Water Soluble Quinoxalines. After extraction
with chloroform, 10 mL of the remaining aqueous phase was spiked
with the IS (0.1006 M, 60 µL) and passed through a preconditioned
LC-18 solid phase extraction cartridge (6 mL, Supelclean, LC18,
Supelco). The cartridge was washed with distilled water (10 mL), and
quinoxalines were eluted with methanol (HPLC grade, 4 mL) and dried
under a stream of nitrogen. The solid was resuspended in BSTFA and
TMCS, 99:1 (Supelco) (500 µL) and acetonitrile (500 µL) and heated
(1 h, 60 °C), and then 1 µL of the solution was injected into the GC-
MS. The R-dicarbonyls 3-DP, 1-DH, and 3-DH were detected as their
corresponding quinoxaline derivatives at the tR given in parentheses:
3-DP quinoxaline (synthetic material) (23.3 min), 1-DH quinoxaline
(synthetic material) (24.7 min), and 3-DH quinoxaline (synthetic
material) (26.8 min). The peak areas were determined at the target ions
m/z 245, 246, and 245, respectively, for the mentioned quinoxalines
and m/z 158 for the IS. To verify the performance of the developed
HRGC-MS method for R-dicarbonyls, parameters such as limit of
detection (LOD), linearity, and recovery were evaluated. LOD limits
were calculated following the IUPAC approach (13), which consists
of analyzing blank beer samples (n ) 10), calculating the standard
deviation, and expressing the results as the blank average value plus
2.26 (the t value for 95% probability and 9 degrees of freedom) times
the standard deviation. The IS was added to each blank sample to correct
for the injection and extraction errors. The recoveries of R-dicarbonyls
from beer were assessed by spiking the corresponding pure quinoxaline
(25, 60, 9, 15, 120, 9.8, 20, and 25 µM for 1,4-DDP, 1,4-DDH, 1-DP,
3-DP, 3-DH, 1-DH, diacetyl, and methylglyoxal, respectively). Three
replicates were prepared for the recovery study.
Quantification of Strecker Aldehydes by SPME-HRGC-MS. The
aldehydes 2-methylpropanal, 2-methylbutanal, and 3-methylbutanal
(Aldrich) were quantified by the method developed by Vesely et al.
(14) with slight modifications. Briefly, a 65 µm PDMS/DVB SPME
fiber (Supelco) was exposed to a PFBHA solution (6 g/L) for 5 min at
40 °C. The loaded fiber was then exposed (30 min at 50 °C) to the
headspace of 20 mL beer samples placed in 40 mL vials containing
the IS (3-methyl-2-butenal, 200 µL, 2 mg/L). Derivatized aldehydes
were desorbed from the fiber in the HRGC-MS injection port.
Semipreparative RP-HPLC. A Waters HPLC System LC Module
I Plus was used, consisting of a 600 pump, Wisp 717 autosampler,
UV-vis 486 detector operating at 320 nm, and an automatic fraction
collector from Waters. Separations were performed on a stainless steel
column (6 µm, 60 Å, 19 mm × 300 mm i.d., C18 Prep Nova Pak HR,
Waters Corp.) at a flow rate of 10 mL/min. Chromatographic data were
acquired using Millennium 2010 Chromatography Manager (Waters
Corp, v. 2.15).
HRGC-MS. HRGC-MS was performed using a Hewlett-Packard
HP6890 GC (Agilent Technologies, Palo Alto, CA) instrument by using
a capillary HP-5MS column (30 m × 0.25 mm i.d., 0.25 µm, Agilent
Technologies) and a split/splitless inlet at 220 °C. The quinoxaline
samples were applied in the split mode (5:1), and the oven temperature
was 70 °C. The temperature of the oven was raised at 6 °C/min to 260
°C, then raised at 25 °C/min to 280 °C and held for 2 min. For Strecker
aldehyde analysis, the PDMS/DVB fiber was desorbed in the injection
port by being heated in the pulsed splitless mode for 1 min at 250 °C
(9.9 psi for 0.5 min). The oven temperature program was as follows:
70 °C (1 min), then 6 °C/min to 200 °C, and 25 °C/min to 265 °C (5
min). For both methods, MS analysis was performed with a Hewlett-
Packard 5973A mass selective detector in the electron impact mode
(70 eV), and the mass range was between m/z 15 and 550.
1
in vacuo and then freeze-dried (white powder, 3.28 mmol): H NMR
(400 MHz, in D2O) δ 2.60 (s, 3H, -CH3), 3.71 [dd, 1H, 2J ) 11.96, 3J
) 7.08, -CHaHb(OH)], 3.80 [dd, 1H, 2J ) 11.96, 3J ) 4.12,
3
3
-CHaHb(OH)], 5.05 [dd, 1H, J ) 7.08, J ) 4.12, -CH(OH)-], 7.42
[m, 2H], 7.64 [m, 2H]; HRGC-MS (MS/EI, trimethylsilylated quinoxa-
line) m/z 348 (25, [M]+), 333 (17), 258 (26), 245 (100), 169 (7), 147
(18), 73 (55).
2-(2′-Hydroxyethyl)-3-methyl Quinoxaline (1,4-DDP Quinoxaline).
1,4-DDP quinoxaline was isolated from the same fraction containing
1-DP quinoxaline by semipreparative RP-HPLC. Separation was
performed isocratically using a mixture of acetonitrile and water (4:
96, v/v). The target quinoxaline eluted at tR ) 65-67 min. The
acetonitrile was eliminated by evaporation in vacuo and then freeze-
1
dried (white powder, 0.23 mmol): H NMR (300 MHz, in DMSO-d6)
δ 2.72 (s, 3H, sCdCsCH3), 3.14 [t, 2H, 3J ) 6.90, sCd
3
CsCH2sCH2(OH)], 3.90 [dt, 2H, 3J ) 6.90, JH-OH ) 5.49, -CH2-
CH2(OH)], 4.73 [t, 1H, 3JH-OH ) 5.49, -CH2-CH2(OH)], 7.72 [m, 2H],
7.96 [m, 2H]; HRGC-MS (MS/EI, trimethylsilylated quinoxaline) m/z
260 (34, M+), 245 (60), 170 (100), 143 (19), 73(63).
2-(2′,3′-Dihydroxypropyl) Quinoxaline (3-DP Quinoxaline). 3-DP
quinoxaline was isolated from the same fraction containing 1-DP
quinoxaline by semipreparative RP-HPLC. The separation was per-
formed isocratically using a mixture of acetonitrile and water (4:96,
v/v). The target quinoxaline eluted at tR ) 17-19 min. The acetonitrile
was eliminated by evaporation in vacuo and then freeze-dried (white
2
powder, 0.07 mmol): (400 MHz, in D2O) δ 3.05 [dd, 1H, J ) 14.1
3
2
3
Hz, J ) 8.8, -CHaHb-CH(OH)-], 3.16 [dd, 1H, J ) 14.1 Hz, J )
4.4, -CHaHb-CH(OH)-], δ 3.58 [dd, 1H, 2J ) 11.7 Hz, 3J ) 4.0,
2
3
-CHaHb(OH)], 3.69 [dd, 1H, J ) 11.7 Hz, J ) 6.2, -CHaHb(OH)],
4.16 [m, 1H, -CH(OH)], 3.64 [dd, 1H, 2J ) 13.4, 3J ) 6.8, -CH(OH)-
CHaHb(OH)], 7.76 [m, 2H], 7.90 [m, 2H], 8.70 [s, 1H, 2 × NdCHs];
HRGC-MS (MS/EI, trimethylsilylated quinoxaline) m/z 348 (8, [M]+.),
333 (29·), 258 (100), 245 (2), 169 (48), 147 (29), 144 (35), 73 (67).
2-Propyl-3-methyl Quinoxaline (IS). 2,3-Hexanedione (44 mmol) and
1,2-DAB (46 mmol) were stirred in methanol (20 mL) for 1 h at 60
°C. The reaction mixture was extracted with chloroform (3 × 10 mL),
the combined organic layers were washed with 0.1 M HCl (3 × 10
mL) and dried over anhydrous MgSO4. The chloroform was evaporated
until dryness in vacuo. Quinoxaline was obtained as a reddish brown
1
2
liquid: H NMR (400 MHz, chloroform-d3) δ 1.06 ppm (t, 3H, J )
2
2
7.3), 1.85 (tc, 2H, J ) 7.3 and 6.5), 2.75 (s, 3H), 2.96 (t, 2H, J )
6.5), 7.65 (m, 2H), 7.98 (m, 2H).
Quantification of r-Dicarbonyls by HRGC-MS. A 2.2 mL aliquot
of 460 mM 1,2-DAB (Sigma, St. Louis, MO) in methanol was added
to 222 mL of beer kept in a bottle and under a CO2 stream at ambient
temperature. The bottle was immediately recapped with crown caps
using a bench capper. The beer pH increased to 4.5 after 1,2-DAB
addition. No further pH adjustment was carried out. Two derivatization
procedures were carried out, the reaction time being the difference
between them. In the derivatization procedure 1 (DE1), the reaction
time with 1,2-DAB varied with the aging period. In this case, bottles
were stored at 28 °C up to 3 months. In the derivatization procedure 2
(DE2) 1,2-DAB was allowed to react in beer at 28 °C for 24 h. After
the storage period, 15 µL of 0.1006 M 2-propyl-3-methyl quinoxaline
(synthetic material, see Syntheses of Reference Material) in methanol
was added to a 25 mL portion of each beer sample as an IS, and the
beer was extracted with chloroform (3 × 10 mL) to obtain the
chloroform and aqueous soluble quinoxalines.
Quantification of Chloroform Soluble Quinoxalines. The com-
bined chloroform extracts were dried over anhydrous MgSO4 and
evaporated in vacuo until dry. The solid obtained was resuspended in
BSTFA and TMCS, 99:1 (Supelco, Bellefonte, PA) (400 µL) and heated
(1 h at 60 °C), and then 1 µL of the solution was injected into the GC