MW“

‘
1

W

 

HIM!

WWHWHHIUWHWW:

i
l

   

219 M

  

THS

 

 

 

 

 

 

 

mists
l

2004
gasi;fiaa

 

. LlBRARY
M'Chlgan State

University This is to certify that the
thesis entitled

 

 

 

Development of a Method for the Detection and Quantification
of 2-Methoxy-3-Isobutylpyrazine in Grape Juice and \Mne

presented by

Daniel Joseph Wampfler

has been accepted towards fulfillment
of the requirements for the

MS. degree in Horticulture

 

 

 

 

 

 

 

ajor Ptof e'ssor’s Signature

__EE§;£igLeium9;§—___

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDuepssop. 15

 

DEVELOPMENT OF A METHOD FOR THE DETECTION AND QUANTIFICATION
OF 2-METHOXY-3-ISOBUTYLPYRAZINE IN GRAPE JUICE AND WINE
By

Daniel Joseph Wampfler

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Horticulture

2003

ABSTRACT

DEVELOPMENT OF A METHOD FOR THE DETECTION AND QUANTIFICATION
OF 2-METHOXY-3-ISOBUTYLPYRAZINE IN GRAPE JUICE AND WINE
By

Daniel Joseph Wampfler

The expanded utilization of “Bordeaux” red wine cultivars in the cool climate of
the Great Lakes Region is limited by the persistence of a group of naturally occurring
herbaceous, bell-pepper aroma and flavor compounds in the wine. This group is
collectively called methoxypyrazines (MP). The specific compound 2-methoxy-3-
isobutylpyrazine (IBMP) is the MP in greatest quantity and human perception threshold
is at levels below 20 ng/L (20 parts/trillion). Viticultural and enological efforts to reduce
the IBMP to below human perception levels have been hindered by the inadequacies of
present methods available to identify and quantify them in complex juice and wine
matrices. This thesis concerns a revised assay method employing techniques of head-
space solid phase microextration coupled with stable isotope dilution gas-
chromatography-mass-spectrometry, following a revised vapor extraction method and
purification on a strong acid cation exchange resin column. The method detects IBMP at

typical concentrations in both grape juice and wine.

This thesis is dedicated to my mother and father, for whom I have looked to for guidance
and wisdom my entire life. I want to thank my loving wife for her unconditional love and

support.

iii

ACKNOWLEDGMENTS

I thank Dr. Stan Howell for his support and guidance throughout my academic
career at Michigan State University. I am appreciative of his keen sense of humor and
ability to make sound judgments on my behalf. I also thank Dr. Randolph Beaudry for
his un-wavered assistance in the laboratory. Without his aid, this project might not have
come to fruition.

I appreciate the moral support and friendship of Jen Dwyer, J on Treloar, Marcel
Lenz, Kacey Watts, Leah Clearwater, and Bill Nail. These colleagues and friends made
life as a graduate student enjoyable and most of all, memorable.

Lastly, I want to acknowledge the long hours of discussion and laboratory time I
was accompanied by my father. Because of him, I am a better scientist, and a better

person.

iv

TABLE OF CONTENTS

INTRODUCTION .................................................................................... 1
LITERATURE REVIEW ........................................................................... 3
MATERIALS AND METHODS ................................................................... 8
RESULTS AND DISCUSSION ................................................................... 13
CONCLUSIONS .................................................................................... 18
LITERATURE SITED .............................................................................. 19
SUBDIVISION ...................................................................................... 24
PAPER 1

A Simplified Method for the Detection and Quantification

of 2-Methoxy-3-Isobutylpyrazine in Grape Juice and Wine .................................. 25
PAPER 2

Analytical Method Development in Wine Chemistry,

A Narrative of the Process: Pitfalls and Breakthroughs ....................................... 33

Introduction

Michigan has 13,500 acres of grapes (1,500 of which are devoted to wine grapes),
and ranks fourth in the nation for total acres (Michigan Grape and Wine Industry
Council, 2002). More than 25 wineries in the state of Michigan produce approximately
200,000 cases annually; the vast majority of these are produced from Michigan grown
grapes (Michigan Grape and Wine Industry Council, 2002). For Michigan and other cool
climate wine producing regions, competition with the premium Viticultural and
enological areas of the world such as Australia, California, Chile, France, Germany and
Italy, requires that sound wines must be produced that meet the critical standards of the
wine consuming public at a competitive price per bottle.

One key element for cool climate areas competitiveness is the production of
Bordeaux-style wines from the grape cultivars Cabernet Sauvignon, Cabernet Franc,
Merlot, and Sauvignon Blanc. The most significant factor limiting acceptance of the red
wine cultivars is the presence of vegetative aromas and flavors in these wines when the
grapes are cultured in cool climate regions. A particular group of compounds, 2-
methoxy-alkylpyrazines (MP), are responsible for the vegetative and herbaceous aromas
(like green bell-peppers) found in some wines produced from the above cultivars in cool
climates. 2-Methoxy-3-isobutylpyrazine is the major contributing component of this
group of compounds.

The research described in this thesis was designed with the major goal to deveIOp
a method of detection and quantification for 2-methoxy-3-isobutylpyrazine that was

readily reproducible in a timely fashion and to be applicable to grape juice and wine

samples at any time during the growing season and the fermentation and post-

fermentation processing periods for wine.

Literature Review

Buttery, et al. (1969) determined that the aromatic, grassy, herbaceous, smelling
compound in green bell-peppers (Capcicum annuum) was 2-methoxy-3-isobutylpyrazine
(IBMP). IBMP is found in a wide variety of other fruits and vegetables including, peas,
tobacco, coffee, potatoes, and grapes (Czemy, et al., 2000; Maga, et al., 1973; Murray, et
al., 1975). IBMP is a strongly odorous compound with a lower human perception
threshold of about 2 parts per trillion (2 ng/L), in water (Buttery, et al., 1969; Seifert, et
al., 1970). Bayonove, et al. (1975) was the first to detect IBMP in Cabernet Sauvignon
(Vitis vinifera L. ), suggesting the compound gave the wine its vegetative, green bell-
pepper aromas.

IBMP can also be found in other vinifera grape cultivars: Cabernet Franc, Merlot,
and Sauvignon blanc (Augustyn, et al., 1982; Heymann, et al., 1986; Harris, et al., 1987;
Calo, et al., 1991; Allen, et al., 1991, 1993, 1994; Lacey, et al., 1991). Typical
herbaceous varietal characters of wines derived from these cultivars have been attributed
to the presence of very low levels of IBMP (Harris, et al., 1987). Initial efforts to detect
and quantify IBMP in grapes employing the GC-MS proved unsuccessful (Slingsby, et
al., 1980) and Allen, et al. (1994) suggested that improvement of the selectivity of
isolation and analysis was important.

It has been reported that IBMP concentrations in grapes decline during ripening
and that this phenomenon is linked to weather, the vine’s canopy status and yield, and the
vineyard management practices affecting the sun exposure of grape clusters (Lacey, et

al., 1991; Allen, et al. 1996; Roujou de Boube’e, et al. 2000). Roujou de Boubée, et al.

(2000) reported that, under comparable climactic conditions in Bordeaux vineyards, the
methoxypyrazine content of the grapes at veraison and variations during ripening are
strongly influenced by environmental and cultural conditions including soil type, pruning
and training system, density of plantation and crop load.

The extremely low detection threshold of IBMP requires specialized analytical
isolation techniques to enable detection of the compound. Such low levels are below the
detection abilities of high performance liquid chromatography (HPLC) (Heymann, et al.,
1986) and require the employment of gas chromatography-mass spectrometry (GC-MS)
(Allen, et al., 1994).

Detection and Quantification

Several analytical detection and quantification methods have been developed
specifically for methoxypyrazines and most have been inadequate to achieve rapid and
repeated analysis with great accuracy. The methods developed have mainly been in one
of two forms: a) liquid extraction; or b) solid phase microextraetion (SPME). Both
methods are coupled to GC-MS. Liquid extraction can be a very tedious and time
consuming procedure requiring multiple steps employing solvents to extract a target
compound from an initial matrix such as wine, into the solvent. SPME (Pawliszyn, 1997;
De Fatima Alpendurada, 2000) is a simple and effective adsorption and desorption
technique, which eliminates not only the need for solvents and complicated sample
preparations, but also the laborious techniques for concentrating and analyzing for
volatile or non-volatile compounds in liquid samples or headspace (De Fatima

Alpendurada, 2000).

SPME fibers consist of various types of polymer coatings fused to a silica fiber,
which can adsorb analytes from a sample by immersion or headspace extraction
(Supelco, 2003).

Headspace solid phase microextraetion (HS-SPME) analysis, coupled with a G0
MS, lends itself well to analysis of volatile compounds in any number of different types
of matrices, including wine. HS-SPME has been employed for extraction and
quantification of grape terpenoids (De la Calle Garcia, et a1. 1998), and the presence of
the soil fumigant methyl isoeyanate (Gandini, et al. 1997), and for the analysis of diacetyl
in wine (Hayasaka, et al. 1999). The method was also used by American Water Works
Association (AWWA) to analyze for methyl-isobomeol (MIB) and geosmin, reported as
being responsible for the “musty” odor occurring in drinking water at particular times of
the year.

HS-SPME was used with IBMP as an internal standard with the ability to quantify
MIB and geosmin down to a concentration of a few parts per trillion (AWWA method
6040). The AWWA method of detection for IBMP and other compounds in the parts-
per-trillion range is further reviewed in Volume 19.1 of The Reporter, published by
Sigma-Aldrich in conjunction with Supelco. Hartmann, et al. (2002) recently conducted
a study for the detection and quantification of IBMP in spiked model wines using the
techniques of HS-SPME, but reported that the method he developed was sensitive to only
the part-per-billion range, lacking the sensitivity to quantify common concentrations
found in wine, which are estimated between 10 and 15 ng/L (10-15 ppt) depending on the

grape variety (Allen, et al., 1994; Roujou de Boubée, er al., 2000).

Sample Concentration

Several methods of liquid separation with additional concentration of target
compound exist for the detection and quantification of IBMP. Harris, et al. (1987)
developed two assays for the detection of IBMP in Sauvignon blanc grapes and wine that
uses a stable isotope (deuterated-isobutylmethoxypyrazine, 2-(2H3)methoxy-3-
isobutylpyrazine) as an internal standard. In their process, the analyte itself (IBMP) was
isotopically labeled, then added as an internal standard prior to analysis and was carried
through all isolation, evaporation, and analytical steps, compensating for any losses
throughout isolation and evaporation and for variations in injection volumes and
detection efficiency (Allen, et al., 1994).

Allen, et al. (1994) developed a similar method that was applicable to red grape
juice and red wine as well as white grape juice and white wine. Their assay required a
sample of juice or wine (~24O mL) to be distilled at atmospheric pressure. The distillate
was then stirred with ion exchange resin, which adsorbed the IBMP. The target
compound was then stripped off the resin with a strong base and IBMP was extracted
from the aqueous phase with diehloromethane. The dichloromethane containing the
IBMP was evaporated down to 20 uL, which concentrated the IBMP, and then a 1 uL
injection was made into a GC-MS for analysis.

Roujou de Boubée, et al. (2000) modified the above assay further by performing
the distillation step in a volatile Cash still, atypical piece of equipment frequently used in
an enological laboratory for the determination of volatile acidity and alcoholic strength in
wines. Roujou de Boubée, et al. (2000) reported that the use of a volatile Cash still

extracted between 85 to 100% of IBMP as compared to ~6% by traditional distillation.

Also, rather than stirring the distillate in the presence of ion exchange resin, the distillate
was percolated through a micro-column of resin with a peristaltic pump. The ion
exchange resin was then placed in a glass micropipette, using a glass wool plug placed in
the small end of the pipette to prevent loss of ion exchange resin. The dichloromethane
was concentrated to less than 10 uL prior to an injection of sample into the GC-MS for
analysis.

The primary limitations of this approach include: limited sample size, the need for
solvents and organic phase extraction, the requirement of a peristaltic pump, and the
extended GC-MS run time. From start to finish this method required greater than two
hours time. What is needed to achieve the desired result is a method of extraction,
concentration, and quantification that is robust and quick. This thesis effort was designed

to accomplish that goal.

Materials and Methods

Isolation of 2-methoxy-3-isobutylpyrazine (Final Protocol) Roujou de Boubée, et
al. (2000) has developed a method for the analysis of the IBMP content in grapes during

the ripening period, which was adopted for our analytical procedure as follows:

Ten mL of 10% NaOH plus 20 ng of internal standard [2-(2H3)methoxy-3-
isobutylpyrazine] was added to the wine or the centrifuged must sample (300 mL).
IBMP was then extracted using the process originally devised by Allen et al. (1994), later
modified by Roujou de Boubée, et al. (2000), but simplified as follows:

1) Distillation - Distillation of 300 mL of wine was performed by steam extraction
using a common enological laboratory apparatus known as a volatile Cash still (Figure 1),
typically used to measure the volatile acidity and alcoholic strength of wine. It consists
of extracting volatile compounds of wine introduced in the bubble chamber by steam
over a period of approximately 15 min. The still used was based on typical designs, but
the size was increased to yield a 1 L internal sample chamber and a 3 L external boiling
chamber (as compared to 250 mL, and 1 L respectively, in a common still). The added
volume eliminated foam-over, which was a problem in the common still’s volume. The
steam then passed through a glass condensing-column with cold tap water (Temp. ~ 5°C)
providing condensation of the distillate. This method was reported to extract between 85
and 100% of the IBMP in the sample compared to about 6% by distillation (Roujou dc
Boubée, et al. 2000). The larger still size is a key improvement, allowing for variable
sample sizes; the volume of a low IBMP level sample can simply be increased.

Following concentration, IBMP levels in the sample will be within the detection range of

the GC-MS.

 

(Figure l — Common and custom made Cash volatile still sizes)

2) Ion Exchange - Instead of either agitating the distillate in the presence of strong
acid cation-exchange resin (Allen, et al. 1994), or percolating distillate through a small
IX resin bed contained in a Pasteur pipet (Roujou de Boubée, et al. 2000), the distillate
was passed through a column in the following manner. Approximately 1 g of strong acid
cation ion exchange (IX) resin (Dowex® 50WX4-200, Aldrich Cat. No.: 42,209-6) was
placed in a 50 mL burette with a glass wool plug in the narrow section and a stop cock to
control flow rate. Prior to each use the D( resin was cleaned and re-generated (put into
the acid form) by passing the following through the column, respectively: 1) 12 mL of
10% NaOH (aq); 2) 25 mL of ~4% HCl (aq); and 3) 100 mL of ultrapure HPLC grade
water (Personal Communication, S. Najmy, 2002). Afier the ion exchange resin was

rinsed with 100 mL of ultrapure HPLC grade water, the steam extract was passed through

the column at a flow rate of about 10 mL/min. After all of the distillate had passed
though the IX resin bed, the column was rinsed with 10 mL of ultrapure water.

3) _Elu_tio_n - IBMP was eluted from the column by passing 12mL of 10% NaOH
(aq) through the column at a flow rate of ~10 mL/min. The eluate was collected in a 40
mL amber vial. Three g of NaCl was added to the vial along with a micro stir bar. The
vial was then sealed with a “mininert valve”(Supeleo Cat. No.: 33304). The vial was
placed on a heating stir-plate and agitated while being sampled at 40° C.

4) Sample analysis - A SPME fiber (Supelco Cat. No.: 57348-U) was inserted
through the “mininert valve” and held in the sample headspace for 10 minutes. The fiber
was immediately inserted into a GC-MS unit for a desorption time of 5 minutes. A
cryofocusing trap filled with liquid nitrogen was placed in the GC oven allowing a small
portion of the column to be immersed in the liquid nitrogen during the desorption period.
After five minutes desorption time, the cryofocus trap was quickly removed; the GC oven
shut, and the GC-MS run started.

5) Gas Chromatographv-MassfiSpectrometrtI Method - The system used to analyze
for IBMP was a Hewlett Packard 6890 CC. coupled with a Leco Pegasus 11 MS.
attached to Dell Dimension computer running Pegasus vs. 1.33 GC-MS analysis
software. The capillary column was a Supeleowax 10 (30m x 0.2 mm id; 0.2 pm film
thickness; Cat. No.: 24169). The vector gas was helium (Grade 5.0; BOC Gases), with a
flow rate of 1.2 mL/min. Injection was made in splitless mode allowing for all of the
desorbed compound to mi grate through the machine. The temperature gradient from 40

°C to 80°C was 100 °C/min, and then the ramping rate was changed to 50 °C/min until

the final temperature of 240 °C was reached, then held for l min. Inlet temperature was

10

set at 220 °C and the transfer line was set at 200 °C. Each injection lasted 4.6 minutes.
Ion selection range was from mass/charge (m/z) 36 and 174.

The system was calibrated by adding increasing concentrations of IBMP (0-100
ng/L for wine) to a Pinot noir wine produced by Michigan State University’s research
winery following typical production methods. The internal standard concentration was
80 ng/L. Each of the samples prepared in this way were extracted according to the

method previously described (Figure 2).

 

 

F ’1
Standard Curve

4.00

 

R2 = 0.9747

y = 0.0153x + 1.7363 /

 

9°
U"!
o
l
l

N-IBMP/D-I S
N w
Ln CD
C) C)

 

 

 

1.50 _ i I l l I

O 20 4O 60 80 100 120
L _ ng/L

 

 

 

(Figure 2 — Initial standard curve developed in a wine matrix spiked with varying
levels of IBMP; x-axis: ng/L IBMP added, y-axis: ratio of IBMP peak area over internal
standard peak area)

The above standard curve was produced a second time to establish its
reproducibility. A 12% ethanol and 88% water matrix was used to produce a curve to

establish a difference in standard curves based on matrix type. It was also observed that

11

after completing a single adsorbtion/desorption and run on the GC-MS, samples were
completely taxed. A decay curve was produced by spiking a 12% ethanol/88% water
matrix with 1000 ng/L IBMP and repeatedly sampling it to determine the number of

sample replications required to tax the sample.

12

Results and Discussion
Standard Curve #2

The standard curve was reproduced in the same manner as above. The slope of
the line was identical to the first curve produced, however the slope intercept was
different. The change in slope intercept does not effect the concentration calculation for
unknown samples. Because IBMP was added to a wine that contained an initial quantity
of IBMP, the y-intercept will be somewhere above zero. When samples with unknown
levels of IBMP are run for quantification, it can be assumed that the y-intercept is zero
because no addition of IBMP will be made. The regression equation for wine is as
follows: IBMP concentration (ng/L) = (A/Ais)/0.0153 (A, area of IBMP peak; Ais, area of

internal standard peak) (Figure 3).

T"— __ ’ 5 ’l

 

 

 

Std. Curve Repeat l
3.00 2
R = 0.9581
y = 0.0153x + 1.2848 /
2.50 -t --

 

 

 

 

N-IBMP/I.S.
N
8

 

 

 

 

H
U1
0

.\

0

1 . 00 T . l . i l
0 20 40 60 80 100 120

L________ n n9/L

 

 

 

 

 

 

 

(Figure 3 - Repeated standard curve developed in a wine matrix spiked with varying

l3

levels of IBMP; x-axis: ng/L IBMP added, y-axis: ratio of IBMP peak area over internal
standard peak area)
Standard Curve in 12% Ethanol

A standard curve was also developed in a matrix of 12% ethanol and 88% water
(Figure 4). Increasing concentrations of IBMP (0-100 ng/L) were added to the
ethanol/water solution. The internal standard concentration was 80 ng/L. The slope of
the line (y=0.0095x + 0.1349) was different from the previous standard curves developed
in a wine matrix. It is important to develop standard curves in a matrix that is similar to
the matrix to be analyzed. This is consistent with the work done by Roujou de Boubée, et

al. (2000) in which a standard curve was developed for both must and wine.

12% EtOH/Water Std. Curve l

 

 

 

 

 

 

 

 

 

1.20 1

vi 1 00 _._____y -_—-_fi0.0095x + 0.1349 /_o _g l
- ' R2 = 0.9345 .
E 0-80 ____ / -wwe .-._ l
E 0.60 44-- ~~ — ,/ ~ ,Lm__- I
l

i 0.20 t—’ ~——~—- - —+~-—— - - a l
0.00 c l

0 20 40 6O 80 100 120 ‘

l

ng/L

l1_____

(Figure 4 — A standard curve developed in a 12% ethanol/88% water matrix spiked with
varying levels of IBMP; x-axis: ng/L IBMP added, y-axis: ratio of IBMP peak area over
internal standard peak area)

Reproducibility (CD

14

The reproducibility of the IBMP analysis on a series of eight measurements all spiked
with 50 ng/L of IBMP was high (Table 1).

Reproducibility of IBMP Analysis in Wine

Sample # ng/L
56.0
52.1
56.5
46.8
57.7
56.6
51.1
45.7
av(n=8) 52.8

SD 4.60
CV(°/o) 8.72

 

mVOIU'I-wat-t

 

(Table 1)
Double Blind Study
A more rigorous study was conducted in which a neutral party spiked six wine
samples with varying levels of IBMP that remained unknown to the analyst until after the
quantification data was collected and IBMP concentrations were calculated. The
calculated concentrations were compared to the actual levels added to the wine and were
found to be in high correlation (Table 2).

Double Blind Study
Actual Calculated

 

IBMP IBMP
(ng/L) Ins/L)
33 30
67 66
100 101
16 15
33 35

 

(Table 2 — Double blind study of wine spiked with varying levels of IBMP)

15

Single Sample DecayI Curve

It was observed that after a sample had completed the extraction and
concentration preparations, been adsorbed with a SPME fiber, desorbed into the GC-MS
and run, that the sample could not be adsorbed for a second run. The quantity of IBMP
was taxed from the vial and could not be used for further analysis. An experiment was
conducted in which a 12% ethanol/88% water solution was spiked with 1000 ng/L (1000
ppt) of internal standard. This concentration is between 10 and 100 times the
concentration range commonly found in wine (10-100 ng/L). The sample was then
adsorbed with a SPME fiber and then desorbed and run on the GC-MS multiple times to
determine the number of sampling replicates requires to tax a sample of very high

concentration, relative to wine samples (Figure 5).

I w— " _ _ .___ WI "T T

l Sampling Replication

 

 

 

 

 

 

 

 

 

GC-MS Response

 

 

 

 

 

L Adsorbtion #-_ __ _ _ l

 

(Figure 5 — Decay curve determining the number of sample replicates required to

tax a wine sample spiked with 1000 ng/L IBMP)

l6

In Figure 5, the lowest point on the curve (seventh replicate) is a false data point that has
been added to demonstrate a zero response from the GC-MS due to the IBMP levels
being lower that the detection limits of the instrument. After the sixth replicated
sampling the quantity of IBMP remaining in the vial was below the detection limits of the

GC-MS.

17

Conclusions

This method for detection and quantification of 2-meth0xy-3-isobutylpyrazine has
good reproducibility and offers the analyst a faster, simpler, less complicated assay than
was previously available. The distillation process has been enhanced to accommodate
varied sample sizes between zero and 500 mL. The improved method employs solid
.phase microextraetion, eliminating the need for solvents and reducing the labor in
concentrating and analyzing volatile compounds in the headspace down to an
adsorption/desorption step. The GC-MS run time is estimated to be 40 minutes shorter
than previously published methods. The method will work on juice or clarified must
samples so that individual wine production steps or vineyard practices may be evaluated
for their impact on IBMP concentration in the finished wine.

While this simplified method can be improved, the advantages of the combined
analytical techniques in the assay allow for faster sample preparation and reduced
analysis time, and should result in expanded research efforts aimed at IBMP reduction
through Viticultural methods such as clonal selection, crop control, and canopy
management and enological approaches such as yeast choice and malolactic strain,
fermentation conditions and blending. This is a method of detection and quantification
for 2-methoxy-3-isobutylpyrazine that is readily reproducible in a timely fashion and is
not only applicable to grape juice and wine at any time during the growing season but
also throughout the fermentation and post-fermentation processing of wine. We believe
it is a valuable tool for viticulture and enological studies focused on cultivars of red and

white Bordeaux wine cultivars in cool climates.

18

10.

11.

Literature Sited

. Allen, M.S.; Lacey, M. J. Methoxypyrazine grape flavour: influence of climate,

cultivar and viticulture. Vitic. Enol. Sci. 1993, 48, 211-213

Allen, M.S.; Lacey, M. J.; Boyd, S. Determination of methoxypyrazines in red
wines by stable isotope dilution gas chromatography mass spectrometry. J. Agric.
Food. Chem. 1994, 42, 1734-1738.

. Allen, M.S.; Lacey, M. J.; Boyd, S. Methoxypyrazines in red wines: Occurrence

of 2-Methoxy-3-(1-methylethyl)pyrazine. J. Agric. Food Chem. 1995, 43, 769-
772.

Allen, M.S.; Lacey, M. J .; Harris, R. L. N.; Brown, W. V. Contribution of
methoxypyrazines t0 Sauvignon blanc wine aroma. Am. J. Enol. Vitic. 1991, 42,
109-112.

American Water Works Association, Method 6040. Constituent Concentration by
Gas Extraction. Standard Methods for the Examination of Water and Wastewater.

Barker, D. A.; Capone, D. L.; Pollnitz, A. P.; McLean, H. J .; Francis, 1. L.;
Oakey, H.; Sefton, M. A. Absorption of 2,4,6-trichloroanisole by wine corks via

the vapour phase in an enclosed environment. Australian J. of Grape and Wine
Research. 2001, 7, 40-46.

Bellavia, V.; Natangelo, M.; F anelli, R.; Rotilio, D. Analysis of benzothiazole in
Italian wines using headspace solid-phase microextraetion and gas
chromatography-mass spectrometry. J. Agric. Food. Chem. 2000, 48, 1239-1242.

Boison, J. O. K.; Tomlinson, R. H. New sensitive method for the examination of
the volatile flavor fraction of Cabernet sauvignon wines. J. Chromatography.
1990, 522, 315-327.

Buttery, R. G.; Seifert, R. M.; Guadagni, D. G.; Ling, L. C. Characterization of
some volatile constituents of bell peppers. J. Agric. Food. Chem. 1969, 17, 1322-
1327.

Csiktusnadi Kiss, G. A.; Forgacs, E.; Cserhati, T.; Candeias, M.; Vila-Boas, L.;
Bronze, R.; Spranger, I. Solid-phase extraction and high-performance liquid

chromatographic separation of pigments of red wines. J. Chromatography A.
2000, 889, 51-57.

Czemy, M.; Grosch, W. Potent odorants of raw arabica coffee. Their changes
during roasting. J. Agric. Food Chem. 2000, 48, 868-872.

19

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

De Fatima Alpendurada, M. Solid-phase microextraetion: A promising technique
for sample preparation in environmental analysis. J. Chromatography A. 2000,
889, 3-14.

Dela Calle Garcia, D.; Reichenbacher, M.; Hurlbeck, C.; Bartzsch, C.; Feller, K-
H. Analysis of wine bouquet components using headspace solid-phase
microextraetion-capillary gas chromatography. .1. High Resol. Chromatogr. 1998,
21, 373-3 77.

Di Tommaso, D.; Calabrese, R.; Rotilio, D. Identification and quantitation of
hydroxytyrosol in Italian wines. J. High Resol, Chromatogr. 1998, 21, 549-553.

Dubourdieu, D.; Tominaga, T.; Masneuf, I.; Peyrot Des Gachons, C.; Murat, M.
The role of yeasts in grape flavor development during fermentation: the example
of Sauvignon blanc. Proc. ASE V 5 Oth Anniversary Annual Meeting, Seattle,
Washington. June 19-23, 2000.

Edson, C. E.; Howell, G. S.; Flore, J. A. Influence of crop load on photosynthesis
0f Seyval grapevines 1. Single leaf whole vine response pre- and post-harvest. Am.
J. Enol. Vitic. 1993, 44, 139-147.

Fischer, C.; Fischer, U. Analysis of cork taint in wine and cork material at
olfactory subthreshhold levels by solid phase microextraetion. J. Agric. Food.
Chem. 1997, 45, 1995-1997.

Gandini, N.; Riguzzi, R. Headspace solid-phase microextraetion analysis of
methyl isothiocyanate in wine. J. Agric. Food. Chem. 1997, 45, 3092-3094.

Giannakopoulos, P.I.; Markakis, P.; Howell, G.S. The influence of malolactic
strain on the fermentation on wine quality of three eastern red wine grape
cultivars. Am. J. Enol. Vitic. 1984, 35, 1-4.

Harris, R. L. N.; Lacey, M. J .; Brown, W. V.; Allen, M. S. Determination of 2-
methoxy-B-alkylpyrazines in wine by gas chromatography/mass spectrometry.
Vitis. 1987, 26, 201-207.

Hartmann, P.J.; McNair, H.M; Zoecklein, B.W. Measurement of 3-Alkyl-2-
Methoxypyrazine by Headspace Solid-Phase Microextraetion in Spiked Model
Wines. Am. J. Enol. Vitic. 2002, 54, 285-288.

Hashizume, K.; Kida, S.; Samuta, T. Effects of steam treatment of grape cluster
stems on the methoxypyrazine, phenolic, acid, and mineral content of red wines
fermented with stems. J. Agric. Food. Chem. 1998, 46, 4382-4386.

20

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

Hashizume, K.; Kida, S.; Samuta, T. Green odorants of grape cluster stem and
their ability to cause a wine stemmy flavor. J. Agric. Food Chem. 1997, 45, 1333-
1337.

Hayasaka, Y.; Bartowsky, E. J. Analysis of diacetyl in wines using solid-phase
microextraetion combined with gas chromatography-mass spectrometry. J. A gric.
Food. Chem. 1999, 47, 612-617.

Heymann, H.; Noble, A. C.; Boulton, R. B. Analysis of methoxypyrazines in
wines 1. Development of a quantitative procedure. J. Agric. Food. Chem. 1986,
34, 268-271.

Ibafiez, E.; Bernhard, A. R. Solid-phase microextraetion of pyrazines in model
reaction systems. J. Sci. Food. Agric. 1996, 72, 91-96.

Jones, R. G.; Pyrazines and related compounds. I. A new synthesis of
hydroxypyrazines. J. Am. Chem. Soc. 1949, 71, 78-81.

Karmas, G.; Spoerri, P. E. The Preparation of hydroxypyrazines and derived
chloropyrazines. J. Am. Chem. Soc. 1952, 74, 1580-1584.

Lacey, M. J .; Allen, M. S.; Harris, R. L. N.; Brown, W. V. Methoxypyrazines in
Sauvignon blanc grapes and wines. Am. J. Enol. Vitic. 1991, 42, 103-108.

Maga, J. A.; Sizer, C. E. Pyrazines in foods. A review. J. Agric. Food Chem.
1973, 21, 22-30.

Matisova, E.; Sedlakova, J .; Slezackova, M. Solid-phase microextraetion of
volatile polar compounds in water. J. High. Resol. Chromatogr. 1999, 22, 109-
115.

Mestres, M.; Busto, 0.; Guasch, J. Headspace solid-phase microextraetion
analysis of volatile sulphides and disulphides in wine aroma. Journal of
Chromatography A. 1998, 808, 211-218.

Miller, D. P.; Howell, G. S. The effects of various carbonic maceration treatments
on must and wine composition of Mareehal foch. Am. J. Enol. Vitic. 1989, 40,
170-174.

Miller, D. P.; Howell, G. S.; Michaelis, C. S.; Dickmann, D. I. The content of
phenolic acid and aldehyde flavor of white oak as affected by site and species.
Am. J. Enol. Vitic. 1992, 43, 333-338.

Miller, D. P.; Howell, G. S.; Striegler, R. K. Reproductive and vegetative

response of mature grapevines subjected to differential cropping stresses. Am. J.
Enol. Vitic. 1993, 44, 435-440.

21

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

Miller, M. E.; Stuart, J. D. Comparison of gas-sampled ans SPME-sampled static
headspace for the determination of volatile flavor components. Analytical

Chemistry. 1999, 71, 23-27.

Murry, K. E.; Whitfield, F. B. The occurrence of 3-alkyl-2-meth0xypyrazines in
raw vegetables. J. Sci. Fd. Aric. 1975, 26, 973-986.

Noble, A. C.; Arnold, R. A.; Buechsenstein, J .; Leach, E. J .; Schmidt, J. 0.; Stem,
P. M. Modification of standardized system of wine aroma terminology. Am. J.
Enol. Vitic. 1987, 38, 143-146.

Pawliszyn, J. Applications of solid phase microextraetion (Book), Royal Society
of Chemistry, ISBN 0-85404-525-2, 1999.

Roberts, D. D.; Pollien, P.; Milo, C. Solid-phase microextraetion method
development for headspace analysis of volatile flavor compounds. J. Agric. Food.
Chem. 2000, 48, 2430-2437.

Rocha, S.; Ramalheira, V.; Barros, A.; Delgadillo, I.; Coimbra, M. A. Headspace
solid phase microextraetion analysis of flavor compounds in wine. Effects of the

matrix volitle composition in the relative response factors in a wine model. J.
Agric. Food. Chem. 2001, 49, 5142-5151.

Roujou de Boubée, D.; Cumsille, A. M.; Pons, M.; Dubourdieu, D. Location of 2-
methoxy-3-isobutylpyrazine in Cabernet sauvignon grape bunches and its
extractability during vinification. Am. J. Enol. Vitic. 2002, 53, 1-5.

Roujou de Boubée, D.; Van Leeuwen, C.; Dubourdieu, D. Organoleptic impact of
2-methoxy-3-isobutylpyrazine on red Bordeaux and Loire wines. Effect of
environmental conditions on concentrations in grapes during ripening. J. Agric.
Food. Chem. 2000, 48, 4830-4834.

Seifert, R. M.; Buttery, R. G.; Guadagni, D. G.; Black, D. R.; Harris, J. G.
Synthesis of some 2-methoxy-3-alkylpyrazines with strong bell pepper-like odors.
J. Agric. Food Chem. 1970, 18, 246-249.

Supelco. SPME — A fast and inexpensive approach to trace organic analysis. The
Reporter. 2001, Vol. 19.1.

Supelco Bulletin 869A — Solid phase microextraetion: solventless sample

preparation for monitoring flavor compounds capillary gas chromatography.
1998.

Supelco Bulletin 928 — Solid phase microextraetion troubleshooting guide. 2001.

22

48.

49.

50.

51.

Supelco Bulletin 929 - A practical guide to quantitation with solid phase
microextraetion. 2001.

Urruty, L.; Montury, M.; Braci, M.; F oumier, J.; Dournel J. Comparison of two
recent solventless methods for the determination of procymidone residues in
wines: SPME/GC/MS and ELISA tests. J. Argic. Food. Chem. 1997, 45, 1519-
1522.

Wada, K.; Shibamoto, T. Isolation and identification of volatile compounds from
wine using solid phase microextraetion, gas chromatography, and gas
chromatography/mass spectrometry. J. Agric. Food. Chem. 1997, 45, 4362-4366.

Weber, J .; Beeg, M.; Bartzsch, C.; Feller, K. Improvement of the chemometric

variety characterization of wines by improving the detection limit of aroma
compounds. J. High Resol. Chromatoge. 1999, 22, 322-326.

23

The following papers are written in the format
of the journal in which they are

intended to be submitted for publication.

24

Technical Brief

A Simplified Method for the Detection and
Quantification of 2-Methoxy-3-Isobutylpyrazine in

Grape Must and Wine

A simplified method was developed to rapidly and accurately quantify 2-
methoxy-3-isobutylpyrazine combining previous techniques of stable isotope
dilution gas chromatography-mass spectrometry with headspace solid phase

microextraetion.

Key words: 2-methoxy-3-isobutylpyrazine, headspace solid phase

microextraetion, volatile cash still.

A particular group of compounds, 2-meth0xy-alkylpyrazines (MP), are responsible
for the vegetative and herbaceous green bell-pepper aromas and flavors found in wines
produced from cultivars Cabernet Sauvignon, Cabernet Franc, Merlot, and Sauvignon
blanc when grown in cool climates. The major contributing component of this group of

compounds is 2-meth0xy-3-isobutylpyrazine (IBMP) is (Allen et al. 1994). IBMP is a

25

strongly odorous compound with a low human perception threshold of about two parts-
per-trillion (ng/L) in water (Allen et al. 1994; Buttery et al. 1969; Seifert et al. 1970).
The extremely low detection threshold of IBMP requires specialized analytical isolation
techniques in order to be able to detect the compound in wine and berries. In order to
implement vineyard management and cellar practices that would ultimately reduce the
perception of IBMP in wine, which is commonly organolepticly evaluated as negative,
quantification of the target compound is imperative to demonstrate a reduction in
concentration. Several analytical detection and quantification methods have been
developed specifically for methoxypyrazines and most have been shown to be inadequate
to allow for rapid, accurate, or repeated analysis. This report concerns efforts to combine
methods of stable isotope dilution gas chromatography-mass spectrometry (GC-MS)
(Allen et al. 1994; Roujou de Boubée et al. 2002, 2000) with the simplicity and ease of
solid phase microextraetion (Pawliszyn, 1997). Using this approach a method has been
developed to analyze for IBMP in musts or wines in the cellar and fruit being grown in

the vineyard.

Materials and Methods

Distillation. Distillation of 300 mL of wine was performed by steam extraction using
a custom-made enological laboratory apparatus known as a ‘Cash Volatile Still’ which is
typically used to measure the volatile acidity and alcoholic strength of wine. The
modified still was based on typical designs, but the volume was increased to yield a 1 L

internal sample chamber and a 3 L external boiling chamber. The added volume

26

prevented foam-over in the larger than average sample size and provided additional
volume for variable samples sizes (Figure 1).

(Figure l-Common and custom made Cash volatile still sizes)

 

Ion Exchange. The distillate was passed through an ion exchange column in the

following manner: a) approximately 1 g of strong acid cation ion exchange (IX) resin
(Dowex® 50WX4-200, Aldrich, Milwaukee, WI) was placed in a 50 mL burette with a
glass wool plug in the narrow section, and a stopcock to control flow rate; b) the ion
exchange resin was rinsed with ~100 mL of HPLC grade water, and then the steam
extract was passed through the column at a flow rate of about 10 mL/min; c) after all of
the distillate had passed though the DC resin bed, the column was rinsed with 10 mL of

HPLC water.

27

Elution. IBMP was eluted from the column by passing 12 mL of 10% NaOH
through the column. The eluate was collected in a 40 mL amber vial. Three grams of
NaCl were added to the vial along with a micro stir bar. The vial was sealed with a
‘mininert valve’ (Cat. No.: 33304, Supelco, Bellefonte, PA) and placed on a heat/stir-
plate allowing the sample to equilibrate to 40°C and to be agitated while being sampled.

Sample analysis.

A 2 cm-50/30um divinylbenzene/CarboxenTM/polydimethylsiloxane solid phase
microextraetion (SPME) fiber (Cat. No.: 57348-U, Supelco, Bellefonte, PA) was inserted
through the mininert valve and held in the sample headspace for an adsorption time of 10
min. The fiber was immediately inserted into a GC-MS unit for a desorption time of 5
min. A cryofocusing trap filled with liquid nitrogen was placed in the GC oven allowing
a small portion of the column to be immersed in the liquid Nitrogen during the desorption
period. After 5 min. desorption time, the cryofocus trap was quickly removed, the GC
oven shut, and the GC-MS run initiated.

GC-MS method. The system used to analyze for IBMP was a Hewlett Packard (Palo
Alto, CA) 6890 GC coupled with a Leco (St. Joseph, MI) Pegasus II MS attached to a
Dell (Round Rock, TX) Dimension computer running Pegasus (Leco, St. Joseph, MI)
version 1.33 GC-MS analysis software. The capillary column was a Supelcowax 10
(30m x 0.2 mm id; 0.2 pm film thickness; Cat. No.: 24169, Supelco, Bellefonte, PA).
The vector gas was Helium, with a flow rate of 1.2 mL/min. Injection was made in
splitless mode. The temperature gradient from 40 °C to 80°C was 100 °C/min. The

ramping rate was then changed to 50 °C/min until the final temperature of 240 °C was

reached, then held for one min. Inlet temperature was set at 220 °C and the transfer line

28

was set at 200 °C. Each injection lasted 4.6 minutes. Ion selection range was from

mass/charge (m/z) between 36 and 174.

Results and Discussion

Quantitative Assessment. To quantify the naturally occurring concentration of
IBMP in a sample, the analyte itself (IBMP) can be isotopically labeled, and added at a
known concentration as an internal standard. This labeled internal standard (2-
(2H3)methoxy-3-isobutylpyrazine) will be carried through all isolation, distillation, and
analytical steps. This compensates for any losses incurred during isolation and
evaporation and for variations in SPME fiber adsorption/desorption and detection
efficiency as shown effective in preexisting analytical methods (Allen et al. 1994; Harris
et al. 1987; Roujou de Boubée et al. 2002, 2000).

Standard Curves and Reproducibility. The standard curves for this assay were
developed by the addition of increasing concentrations of IBMP to wine. The area of the
IBMP peaks (m/z 124) compared to that of the internal standard peaks (m/z 127) were in
linear correlation with the IBMP concentration added to the wine. The regression
equation for wine is as follows: IBMP concentration (ng/L) = (A/A,S)/0.0153 (A, area of
IBMP peak; Ais, area of internal standard peak) (r2 = .975). The reproducibility of the
IBMP analysis on a series of eight measurements all spiked with 50 ng/L of IBMP was

high (Table 1).

29

Reproducibility of IBMP Analysis in Wine

Sample # ng/ L
56.0
52. 1
56.5
46.8
57.7
56.6
51.1
45.7
av(n=8) 52.8
SD 4.60
CV(°/o) 8.72

(Table 1- Reproducibility of IBMP analysis of 8 samples of wine spiked with 50 ng/L

 

CDVOXUI-wat-t

 

ofIBMP)

Double Blind Study. A more rigorous study was conducted in which a neutral party
spiked six wine samples with varying levels of IBMP that remained unknown to the
analyst until after the quantification data was collected and IBMP concentrations were
calculated. The calculated concentrations were compared to the actual levels added to the

wine and were found to be in high correlation (Table 2).

Double Blind Study
Actual Calculated

 

IBMP IBMP
(ng/L) (an/L)
33 30
67 66
100 101
16 15
33 35

 

(Table 2- Double blind study of wine spiked with varying levels of IBMP)

30

Conclusions

This simplified method for detection and quantification of 2-methoxy-3-
isobutylpyrazine offers the analyst a faster, simpler, less complicated assay with good
reproducibility. The distillation process has been enhanced to accommodate varied and
larger sample sizes. The improved method employs solid phase microextraetion,
eliminating the need for solvents and simplifying the labor in concentrating and analyzing
volatile compounds in the headspace. The GC-MS run time is also much shorter than
previously published methods. The method also will work on juice or clarified must
samples so that individual wine production steps or vineyard practices may be evaluated

for their impact on IBMP concentration in the finished wine.

31

Literature Cited

. Allen, M.S., M. J. Lacey. 1993. Methoxypyrazine grape flavour: influence of
climate, cultivar and viticulture. Vitic. Enol. Sci. 48:211-213.

. Allen, M.S., M. J. Lacey, S. Boyd. 1994. Determination of methoxypyrazines in
red wines by stable isotope dilution gas chromatography mass spectrometry. J.
Agric. Food. Chem 42:1734-1738.

. Buttery, R. G., R. M. Seifert, D. G. Guadagni, L. C. Ling. 1969. Characterization
of some volatile constituents of bell peppers. J. Agric. Food. Chem 17:1322-
1327.

. Harris, R. L. N., M. J. Lacey, W. V. Brown, M. S. Allen. 1987. Determination of

2-methoxy-3-alky1pyrazines in wine by gas chromatography/mass spectrometry.
Vitis. 26:201-207.

. Pawliszyn, J. 1997. Applications of Solid Phase Microextraetion: Theory and
Practice. John Wiley & Sons, Inc., New York.

. Roujou de Boubée, D., A. M. Cumsille, M. Pons, D. Dubourdieu. 2002. Location
of 2-mcthoxy-3-isobutylpyrazine in Cabernet sauvignon grape bunches and its
extractability during vinification. Am. J. Enol. Vitic. 53:1-5.

. Roujou dc Boubée, D., C. Van Leeuwen, D. Dubourdieu. 2000. Organoleptic
impact of 2-meth0xy-3-isobutylpyrazine on red Bordeaux and Loire wines. Effect
of environmental conditions on concentrations in grapes during ripening. J. Agric.
Food. Chem. 48:4830-4834.

. Seifert, R. M., R. G. Buttery, D. G. Guadagni, D. R. Black, J. G. Harris. 1970.

Synthesis of Some 2-Methoxy-3 -Alkylpyrazines with Strong Bell Pepper-Like
Odors. J. Agric. Food. Chem. 18:246-249.

32

Analytical Method Development in Wine Chemistry, A
Narrative of the Process: Pitfalls and Breakthroughs

A narrative of the process to develop a simplified analytical method for the
detection and quantification of 2-methoxy-3-isobutylpyrazine in grape juice and

wine.

Key words: Methoxypyrazine, headspace solid phase microextraetion, volatile

cash still, gas chromatography-mass spectrometry.

The research efforts regarding the impact of methoxypyrazines in wine date back
slightly more than a decade. There have been several key improvements in the analytical
methods for both detection and quantification. With the advancements in equipment and
analytical techniques, scientists are able to apply different and often advantageous
practices to their research. This is one such project. The goal of this paper is to give a
narrative of the analytical wine chemistry techniques and approaches applied to
improving and simplifying the methods of detecting and quantifying 2-methoxy-3-

isobutylpyrazine (IBMP) in grape juice and wine.

Previous Research Efforts
The extremely low detection threshold of IBMP requires specialized analytical
isolation techniques in order to detect the compound in wine and berries. Such low levels

are below the detection abilities of high performance liquid chromatography (HPLC)

33

(Heymann et al. 1986) and require the employment of gas chromatography-mass
spectrometry (GC-MS) (Allen et al. 1994). The abilities of analytical chemistry labs, and
especially less equipped enological laboratories, are taxed when striving to detect the
lower thresholds of IBMP in wine (10 ng/L). To put it into perspective: ten parts per
trillion is analogous to the diameter of a human hair compared to the diameter of the
earth, or, stated in terms more closely related to the wine industry, ten drops of vermouth
in a pool of gin the size of a football field, 43 feet deep. It is the extremely low naturally
occurring level of IBMP that makes detection and quantification difficult.

Detection and Quantification.

Several analytical detection and quantification methods have been developed
specifically for methoxypyrazines and most have ultimately been shown to be impractical
for rapid and repeated analysis with adequate accuracy. The methods developed have
mainly followed one of two approaches: liquid extraction or solid phase microextraetion
(SPME). Both methods are typically coupled with GC-MS. SPME is a simple and
effective adsorption and desorption technique, which eliminates not only the need for
solvents and complicated sample preparations, but also the laborious techniques for
concentrating and analyzing for volatile or non-volatile compounds in liquid samples or
headspace (Pawliszyn 1999). Headspace (HS)-SPME analysis, coupled with GC-MS,
lends itself well to analysis of volatile compounds in any number of different types of
matrices, including wine. HS-SPME has been employed for a range of extraction efforts
such as for terpenoids in wine (De la Calle Garcia et al. 1998), and for the presence of the
soil fumigant methyl isoeyanate in wine (Gandini et al. 1997). Hayasaka, et al. (1999)

used the method coupled with GC-MS for the analysis of diacetyl in wine. The method

34

was also used by American Water Works Association (AWWA) to analyze for methyl-
isobomeol (MIB) and geosmin, reported as being responsible for the “musty” odor
occurring in drinking water at particular times of the year. HS-SPME was used with
IBMP as an internal standard with the ability to quantify MIB and geosmin down to a
concentration of a few parts per trillion (AWWA method 6040). The AWWA method of
detection for IBMP and other compounds in the parts-per-trillion range is further
reviewed in Volume 19.1 of The Reporter (2001), published by Sigma-Aldrich in
conjunction with Supelco. Hartmann et al. (2002) recently conducted a study for the
detection and quantification of IBMP in model wines using the techniques of HS-SPME,
but reported the method developed was sensitive to only mg/L levels, lacking the
sensitivity to quantify common concentrations in wine, which are estimated between 10
and 15 ng/L (10-15 ppt) depending on the grape variety (Allen et al. 1994; Roujou dc
Boube'e et al. 2000).

Concentration.

Several methods of liquid separation with additional concentration of the target
compound exist for the detection and quantification of IBMP. Harris et al. (1987)
developed two assays for the detection of IBMP in Sauvignon blanc grapes and wine that
use a stable isotope (deuterated-isobutylmethoxypyrazine, 2-(2H3)meth0xy-3-
isobutylpyrazine) as an internal standard. In their process, the analyte itself (IBMP) was
isotopically labeled, then added as an internal standard prior to analysis and was carried
through all isolation, evaporation, and analytical steps, compensating for any losses
throughout isolation and evaporation and for variations in injection volumes and

detection efficiency (Allen et al. 1994). Allen et al. (1994) developed a similar method

35

applicable to red grape juice and red wine as well as white grape juice and white wine.
Their assay required a sample of juice or wine (~240 mL) be distilled at atmospheric
pressure. The distillate was then stirred with ion exchange (IX) resin, which adsorbed the
IBMP. The target compound was then stripped off the resin with a strong base and IBMP
was extracted from the aqueous phase with diehloromethane. The dichloromethane
containing the IBMP was evaporated down to 20 uL, which concentrated the IBMP, and
then 1 uL was injection into a GC-MS for analysis.

Roujou de Boubée et al. (2000) modified the above assay further by performing
the distillation step in a Cash volatile still: a typical piece of equipment frequently used in
an enological laboratory for the determination of volatile acidity and alcoholic strength in
wines. Roujou de Boubée reported that the use of a Cash volatile still extracted between
85 to 100% of IBMP as compared to ~6% by traditional distillation. Also, rather than
stirring the distillate in the presence of ion exchange resin, the distillate was percolated
through a micro-column of resin with a peristaltic pump. The ion exchange resin was
then placed in a glass micropipette, using a glass wool plug placed in the small end of the
pipette to prevent loss of ion exchange resin. The dichloromethane was concentrated to
less than 10 uL prior to an injection of sample into the GC-MS for analysis.

This was the state of the literature when our efforts were initiated. The goal here
was to employ steam distillation, ion exchange resin, solid phase microextraetion, and
gas chromatography-mass spectroscopy in varying ways to speed-up, simplify and

improve accuracy of the analysis for IBMP.

36

Materials and Methods
Initial Modification Efforts: Improving concentration and extraction.

Initial efforts were focused on direct HS-SPME analysis of wine samples. The
naturally occurring levels of IBMP in wine are low in concentration and HS-SPME
coupled with GC-MS was found to lack the sensitivity necessary to detect natural levels.
Attempts to produce standard curves by spiking wine with IBMP and analyzing by HS-
SPME proved unsuccessful due to variation in GC-MS response and the ug/L (part per
billion) spiked levels required for detection. For example when the HS of a series of two
or more wine samples spiked with the same level of IBMP were analyzed, there was no
reproducibility. In some runs the peak area was very small, and very large in others. The
simplicity and speed with which a sample can be analyzed with SPME was a logical
modification that could eventually be added to an assay to detect and quantify IBMP.

The extraction and concentration procedures originally developed by Allen et al.
(1994) have been shown to work but lacked simplicity and case because their method
required liquid phase extraction steps and evaporation techniques, both requiring a
relatively high degree of skill. Modifications in their procedures were made, allowing for
more rapid and more-easily conducted sample preparation. The first such modifications
adopted were the distillation parameters, followed by the ion exchange techniques as
follows:

Distillation: Increase the quantity of wine distilled.

Distillation of 300 mL of wine was performed by steam extraction using a

common enological laboratory apparatus known as a volatile Cash still, typically used to

measure the volatile acidity and alcoholic strength of wine. The still used was based on

37

typical designs, but the size was increased to yield a l L internal sample chamber and a 3
L external boiling chamber (over 3x that of traditional Cash still size). The added volume
eliminated foam-over, which was a problem in the common still’s volume. Increasing
the size of the still also allowed for much larger sample sizes. A wine that contained a
low quantity of IBMP, sufficient to tax the detection abilities of the GC-MS could be
accommodated by the analyst’s flexibility to use more starting material (wine or juice),
which contains additional IBMP.

Ion exchange.

Instead of agitating the distillate in the presence of strong acid cation-exchange
resin (Allen et al. 1994), or percolating distillate through a small IX resin bed contained
in a Pasteur pipet (Roujou de Boubée et al. 2000), the distillate was passed through a
column in the following manner: approximately 1 g of strong acid cation IX resin
(Dowex® 50WX4-200, Aldrich, Milwaukee, WI) was placed in a 50 mL burette with a
glass wool plug in the narrow section and a stop cock to control flow rate. Use of a
burette eliminates the need for peristaltic pumps as used by Roujou de Boubée et al.
(2000) and Allen et al. (1994). Prior to each use the IX resin was cleaned and
regenerated (put into the acid form) by passing the following through the column,
respectively: 1) 10 mL of 10% NaOH; 2) 25 mL of ~3.5% HCl; and 3) 100 mL of HPLC
grade water (S. Najmy 2002, personal communication). After the IX resin was rinsed
with 100 mL of HPLC grade water, the steam extract was passed through the column at a
flow rate of approximately 10 mL/min. After all of the distillate had passed though the

IX resin bed, the column was rinsed with 10 mL of HPLC water. Regeneration of the

38

resin allows continual use of the material and eliminates the time needed to load and
unload the column between runs.
Second Modification Efforts: Refined methods to elute column.

The distillation and ion exchange modifications made during initial efforts
remained in practice throughout the efforts to modify the assay for the detection and
quantification of IBMP. The appropriate volume of strong base to elute IBMP from the
IX resin was not known. A recovery study was conducted to determine a practical
volume and multiple of elutions required to sufficiently elute IBMP at a level within the
detection abilities of the GC-MS.

Elution.

Four mL of 10% NaOH was arbitrarily set as a standard elution volume. It was
observed that a small percentage of the IBMP elutes from the column with a single
elution. To increase the quantity of IBMP eluted, the equivalent of three elutions was

deemed practically advantageous (Figure l).

39

l Elutions vs. GC-MS Response
150000

 

100000

MS Count

50000 ,

 

 

 

10ppb in Non-eluate lst Elution 2nd 3rd Elution
water Elution

l Elution 1,2,&3 =3 mL (12 mL total) ,

 

(Figure 1- Elution volume and replication to efficiently elute IBMP from column)

The following modification was adopted: IBMP was eluted from the column by
passing 12mL of 10% NaOH (aq) through the column at a flow rate of ~1 0 mL/min. The
eluate was collected in a 40 mL amber vial. Three grams of NaCl was added to the vial
along with a micro stir bar. The vial was then sealed with a “mininert valve”(Cat. No.:
33304,Supelco, Bellefonte, PA). The vial was placed on a heat/stir plate and agitated
while being sampled at 40° C.

Sampling.

Attempts were then made to concentrate the eluted IBMP from the aqueous phase
with diehloromethane. The process of evaporating the dichloromethane containing the
IBMP down to 20 uL, which concentrated the IBMP, and then injecting a l 11L injection
into the GC-MS for analysis was found to be tedious, time consuming and a complex,

error producing procedure. At this point HS-SPME analysis seemed a logical alternative

40

to direct injection, eliminating the need for dichloromethane as well as the slow organic
phase extraction and concentration process.

A direct HS-SPME analysis of the eluate proved successful for IBMP detection in
spiked ethanol/water solutions, must, and wine. Unfortunately, the problem of variation
in GC-MS response from one run to the next also accompanied HS-SPME analysis. At
this stage, the assay allowed for the detection of naturally occurring levels of IBMP, but
not for its quantification.

Modification — Internal standard.

This hurdle was overcome by the addition of an internal standard (deuterated-
isobutylmethoxypyrazine, 2-(2H3)methoxy-3-isobutylpyrazine), as had proved successful
for Harris et al. (1987). The internal standard was added prior to analysis and, was carried
through all isolation, evaporation, and analytical steps, compensating for any losses
throughout isolation and evaporation and for variations in injection volumes and
detection efficiency (Allen et al. 1994).

The sampling procedure was developed and implemented as follows: A 2 cm-
50/30um divinylbenzene/CarboxenTM/polydimethylsiloxane SPME fiber (Cat. No.:
57348-U, Supelco, Bellefonte, PA) was inserted through the “mininert valve” and held in
the sample headspace for 10 minutes. The fiber was immediately inserted into a GC-MS
unit for a desorption time of 5 minutes. A cryofocusing trap filled with liquid nitrogen
was placed in the GC oven allowing a small portion of the column to be immersed in the
liquid nitrogen, during the desorption period. After five minutes desorption time, the

cryofocus trap was quickly removed; the GC oven was shut, and the GC-MS run started.

41

GC-MS Reduced Run Time.

Previously developed methods of detection and quantification employed for GC-
MS run times were nearly one hour in duration. This seemed unnecessary and inefficient
for only two target compounds (IBMP and internal standard). After only a few series of
trials, the GC-MS method and run time was successfully modified and reduced to less
than five minutes. The protocol is as follows: The system used to analyze for IBMP was
a Hewlett Packard (Palo Alto, CA) 6890 GC coupled with a Leco (St. Joseph, MI)
Pegasus 11 MS attached to a Dell (Round Rock, TX) Dimension computer running
Pegasus (Leco, St. Joseph, MI) version 1.33 GC-MS analysis software. The capillary
column was a Supelcowax 10 (30m x 0.2 mm i.d.; 0.2 um film thickness; Cat. No.:
24169, Supelco, Bellefonte, PA). The vector gas was Helium, with a flow rate of 1.2
mL/min. Injection was made in splitless mode. The temperature gradient from 40 °C to
80°C was 100 °C/min. The ramping rate was then changed to 50 °C/min until the final
temperature of 240 °C was reached, then held for one min. Inlet temperature was set at
220 °C and the transfer line was set at 200 °C. Each injection lasted 4.6 minutes. Ion
selection range was from mass/charge (m/z) between 36 and 174.

Final Protocol.

With modifications made to the distillation parameters, the ion exchange
techniques, the replacement of liquid injections with HS-SPME, and the shortened GC-
MS run time, a method for the detection and quantification of IBMP was in place, taking
an average of an hour per sample (Figure 3). Simplifications from previous methods are

shown in Table 1.

42

Standard curve development.

Initial efforts to develop standard curves for the quantification of IBMP were
unsuccessful. Originally faux-wines were spiked with increasing amounts of IBMP and
run through the sample preparation procedure with the expectation that the curve could
be extrapolated to relate directly to wine samples. Several problems caused this approach
to fail. Primarily, fluctuations in the GC-MS response caused multiple identical samples
to have large variations in peak area, preventing direct comparison from one
chromatogram to the next. The addition of an internal standard compensated for response
fluctuations, maintaining the deuterated (D)-IBMP in appropriate ratios with the added
naturally occurring IBMP. Additionally, wine is a very complex matrix; when wine
samples were spiked with D-IBMP as well as the naturally occurring IBMP, the GC-MS
responded with consistent peak ratios. It is important to develop standard curves in the
matrix to be analyzed.

Standard curves.

The system was calibrated by adding increasing concentrations of IBMP (0-100
ng/L) to wine. The internal standard concentration was 80 ng/L. Each of the samples
prepared in this way was extracted according to the method above. The ratio of the area
of the naturally occurring IBMP to the D-IBMP was graphed versus the increasing

concentration of the naturally occurring IBMP (Figure 2).

43

 

 

 

Standard Curve

 

R2 = 0.9747
3.50 +- -

y = 0.0153x + 1.7363 /.

 

 

N-IBMP/D-I S
N DJ
U1 0
O O

 

2.00 ___-
/

1.50 i , . .

l 0 20 40 60 80 100 120 .
l______ _ N__/L_______________-_ _flg

(Figure 2 -— Initial standard curve developed in a wine matrix spiked with varying levels

 

 

 

 

 

 

of IBMP; x-axis: ng/L IBMP added, y-axis: ratio of IBMP peak area over internal

standard peak area)

Results and Discussion

The standard curve is the cornerstone of this assay, allowing one to quantify the
unknown naturally occurring IBMP in wine. With the added quantity of internal standard
known, a ratio of the naturally occurring IBMP over the D-IBMP can be determined and
the unknown quantity of IBMP computed based on the equation of the standard curve.
Because IBMP was added to a wine that contained an initial quantity of IBMP, the y-
intercept will be somewhere above zero. When samples with unknown levels of IBMP
are run for quantification, it can be assumed that the y-intercept is zero because no

addition of IBMP will be made. The regression equation for wine is as follows: IBMP

44

concentration (ng/L) = (A/Ais)/0.015 3 (A, area of IBMP peak; Ais, area of internal
standard peak). The curve was reproduced and the slope was identical (results not

shown).

45

Conclusions

This modified method for detection and quantification of 2-methoxy-3-
isobutylpyrazine offers the analyst a faster, simpler, less complicated assay. The
distillation process has been enhanced to accommodate varied and larger sample sizes.
The improved method employs solid phase microextraetion, eliminating the need for
solvents and simplifying the labor in concentrating and analyzing volatile compounds in
the headspace. The GC-MS run time is much shorter than in previously published
methods. This simplified method can be improved. However, the advantages and
combined analytical techniques in the above assay allow for faster sample preparation
and analysis time, helping to encourage greater efforts related to conditions leading to
IBMP reduction into the vineyard and cellar. This is a method of detection and
quantification for 2-methoxy-3-isobutylpyrazine that is readily reproducible and can be

conducted in a timely fashion.

Summary of Simplifications for the Detection and Quantification of IBMP in Wine

 

Category Previous methods’ technique Simplification

Distillation Distillation/ small Cash still Custom Cash Still

IX resin Peristaltic pump, pipette 50 mL burette/ regeneration
Elution 1 mL 10% NaOH 12 mL 10% NaOH

Sample volume 100-240 mL 2 300 mL

Sampling Direct Injection HS-SPME

GC run time > 45 min < 5 min

Total run time Estimated at > 2 hrs/ sample 3 1 hr /sample

Cost / sample Estimated at $100/ sample 3 $25/ sample

 

(Table 1- simplifications made to IBMP detection and quantification)

46

(Figure 3- IBMP extraction flowchart)

Methoxypyrazine Extraction Flowchart

IX Resin Regeneration

1. Add 10 mL of
10% NaOH to flask.

 

—>

 

12. Following elution, the lX resin is
washed w/ lOmL of HPLC water and
then regenerated with 25 ml of 3.5%
HCl (aq).

13. Following regen w/ HCl (aq) the
column is then rinsed with a further
100 ml of HPLC water and ready for
use again.

14. IX resin must be “regenerated”
prior to every use, including when new
resin is used!

 

 

 

 

Cash
Still

  
 

 

6. Wash resin with lOOmL of

HPLC water - discard.

 

 

utsar x1 . ] azzaan '11“ 05‘

 

4. Distill wine.

5. Collect
180 mL ofdistillatc.

Beaker

2. Add 300 mL of
wine/juice to flask.

 
  

 

 
 

3. Add 20 ng (80ppt)
internal standard:
[2-(2H3)methoxy-3-
isobutylpyrazine].

 

 

 

 
 
      
  
 

   
 
   
 

 

7. Pass distillate through burette
- discard.

 

 

 

A

8. Rinse with lOmL HPLC
water - discard.

 

 

./

 

10. Elute IBMP with 12 mL of
10% NaOH - collect in 40 mL
amber vial.

 

 

 

 

mini-stir bar.

 

9. Weigh 3 g ofNaCl
into vial and drop in a

 

 

47

l 1. Analyze by GC/MS.

 

10.

11.

Literature Sited

. Allen, M.S., M. J. Lacey, S. Boyd. 1994 Determination of methoxypyrazines in

red wines by stable isotope dilution gas chromatography mass spectrometry. J.
Agric. Food. Chem. 42:1734-1738.

American Water Works Association, Method 6040. Constituent Concentration by
Gas Extraction. Standard Methods for the Examination of Water and Wastewater.

De la Calle Garcia, D., M. Reichenbacher, C. Hurlbeck, C. Bartzsch, K-H. Feller.
1998. Analysis of wine bouquet components using headspace solid-phase

microextraetion-capillary gas chromatography. J. High Resol. Chromatogr. 21 ,
373-377.

Gandini, N., R. Riguzzi. 1997. Headspace solid-phase microextraetion analysis of
methyl isothiocyanate in wine. J. Agric. Food. Chem 45:3092-3094.

Harris, R. L. N., M.J. Lacey, W.V. Brown, M.S. Allen. 1987. Determination of 2-
methoxy-3-alkylpyrazines in wine by gas chromatography/mass spectrometry.
Vitis. 26:201-207.

Hartmann, P.J., H.M. McNair, B.W. Zoecklein. 2002. Measurement of 3-Alkyl-2-
Methoxypyrazine by Headspace Solid-Phase Microextraetion in Spiked Model
Wines. Am. J. Enol. Vitic. 54:285-288.

Hayasaka, Y., E]. Bartowsky. 1999. Analysis of diacetyl in wines using solid-
phase microextraetion combined with gas chromatography-mass spectrometry. J.
Agric. Food. Chem. 47:612-617.

Heymann, H., A.C. Noble, R.B. Boulton, 1986. Analysis of methoxypyrazines in
wines 1. Development of a quantitative procedure. J. Agric. Food. Chem. 34:268-
271.

Pawliszyn, J. 1999. Applications of Solid Phase Microextraetion: Theory and
Practice. John Wiley & Sons, Inc., New York.

Roujou de Boubée, D., C. Van Leeuwen, D. Dubourdieu. 2000. Organoleptic
impact of 2-methoxy-3-isobutylpyrazine on red Bordeaux and Loire wines. Effect
of environmental conditions on concentrations in grapes during ripening. J. Agric.
Food. Chem. 48:4830-4834.

Supelco. 2001. SPME — A fast and inexpensive approach to trace organic
analysis. The Reporter. Vol. 19.1.

48

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

l11111111111IIIIIIIIIIIII

l
1