PDF Archive

Easily share your PDF documents with your contacts, on the Web and Social Networks.

Share a file Manage my documents Convert Recover PDF Search Help Contact



J Environ Sci Health B 42, 2007, 887 897 .pdf



Original filename: J Environ Sci Health B 42, 2007, 887-897.pdf

This PDF 1.4 document has been generated by pdftk 1.41 - www.pdftk.com / itext-paulo-155 (itextpdf.sf.net-lowagie.com), and has been sent on pdf-archive.com on 03/11/2015 at 02:17, from IP address 71.17.x.x. The current document download page has been viewed 337 times.
File size: 169 KB (12 pages).
Privacy: public file




Download original PDF file









Document preview


This article was downloaded by:[Rayne, Sierra]
On: 3 November 2007
Access Details: [subscription number 783726712]
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Environmental Science
and Health, Part B
Pesticides, Food Contaminants, and Agricultural
Wastes
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713597269

4-Ethylphenol and 4-ethylguaiacol in wines: Estimating
non-microbial sourced contributions and toxicological
considerations
Sierra Rayne ab; Nigel J. Eggers b
a
Pacific Agri-Food Research Centre, Agriculture and Agri-Food Canada, British
Columbia
b
Irving K. Barber School of Arts & Sciences, The University of British Columbia at
Okanagan,

Online Publication Date: 01 November 2007
To cite this Article: Rayne, Sierra and Eggers, Nigel J. (2007) '4-Ethylphenol and 4-ethylguaiacol in wines: Estimating
non-microbial sourced contributions and toxicological considerations', Journal of Environmental Science and Health, Part
B, 42:8, 887 - 897
To link to this article: DOI: 10.1080/03601230701623365
URL: http://dx.doi.org/10.1080/03601230701623365

PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article maybe used for research, teaching and private study purposes. Any substantial or systematic reproduction,
re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly
forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents will be
complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be
independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings,
demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or
arising out of the use of this material.

Downloaded By: [Rayne, Sierra] At: 20:15 3 November 2007

Journal of Environmental Science and Health Part B (2007) 42, 887–897
C Taylor & Francis Group, LLC
Copyright
ISSN: 0360-1234 (Print); 1532-4109 (Online)
DOI: 10.1080/03601230701623365

4-Ethylphenol and 4-ethylguaiacol in wines: Estimating
non-microbial sourced contributions and toxicological
considerations
SIERRA RAYNE1,2 and NIGEL J. EGGERS2
1
2

Pacific Agri-Food Research Centre, Agriculture and Agri-Food Canada, British Columbia
Irving K. Barber School of Arts & Sciences, The University of British Columbia at Okanagan

Received April 24, 2007.

Analyses of commercially available wines suggested non-Brettanomyces sources of 4-ethylphenol and 4-ethylguaiacol. Grapes, enological additions, exposure to plastics, and oak-barrel aging were potential inputs considered. Investigations of whole grape bunch
samples from two major red wine Vitis vinifera cultivars (L. cv. Cabernet Franc and Pinot Noir), a commercial mannoprotein additive,
and three commercial enological tannin additions indicated they are not likely significant sources of these compounds. Studies on
15 commercial oak barrelled red wines from six Vitis vinifera cultivars (L. cv. Cabernet Franc, Cabernet Sauvignon, Dunkelfelder,
Merlot, Pinot Meunier, and Pinot Noir), and a review of volatile phenol extraction from toasted oak wood, suggested that oak-aging
may produce concentrations of up to 50 µg L−1 4-ethylphenol and 4-ethylguaiacol. Thus, following potential Brettanomyces-sourced
aroma impacts in wine using 4-ethylphenol and/or 4-ethylguaiacol concentrations as proxies should only be considered reliable at
analyte levels >100 µg L−1 . A review of worldwide 4-ethylphenol and 4-ethylguaiacol concentrations in wine, consumption patterns,
and available toxicological data also suggested that levels of 4-ethylphenol being observed in wines worldwide do not warrant concerns
about acute or long-term effects. While little is known about the toxicology of 4-ethylguaiacol, it is unlikely that elevated concentrations
will pose any health-related concerns.
Keywords: 4-Ethylphenol; 4-ethylguaiacol; wine; grape; oak; non-Brettanomyces sources; toxicology; consumption patterns; health
impacts.

Introduction
Brettanomyces is an important yeast genus known to play
a significant role in the aroma of red wines.[1,2] Two compounds considered to be a major part of the Brettanomyces
aroma in wines are 4-ethylphenol and 4-ethylguaiacol, although other volatile phenols,[3] acetic acid,[4] isovaleric
(3-methylbutyric) acid,[5] and tetrahydropyridines[6] have
been linked to Brettanomyces-sourced organoleptic defects.
Brettanomyces-sourced 4-ethylphenol and 4-ethylguaiacol
are formed by the decarboxylation and subsequent reduction of the corresponding grape-derived hydroxycinnamic acids (p-coumaric acid and ferulic acid, respectively)
(Fig. 1).[7,8]
The Brettanomyces metabolites 4-ethylphenol and 4ethylguaiacol can produce phenolic, animal, and stable
Address correspondence to Sierra Rayne National Bioproducts and Bioprocesses Program, Pacific Agri-Food Research
Centre, Agriculture and Agri-Food Canada, 4200 Highway 97,
Summerland, British Columbia, Canada, V0H 1Z0; E-mail:
raynesierra@yahoo.ca

wine odors at concentrations in the range of several hundred µg L−1 for 4-ethylphenol, to as low as about 50
to 100 µg L−1 for 4-ethylguaiacol.[9–11] Worldwide concentrations of 4-ethylphenol vary widely, with levels averaging about 400 to 500 µg L−1 in France (ranging up
to 6 mg L−1 ),[10] about 400 to 800 µg L−1 in Australia
(ranging up to 2.7 mg L−1 ),[12–14] and near 50 µg L−1 in
the Okanagan Valley appellation of Canada (ranging up
to 540 µg L−1 ).[15,16] 4-Ethylguaiacol concentrations show
similar spatial distributions among various winemaking
regions, but the average levels are generally 2- to 10-fold
lower than those of 4-ethylphenol. At our present level of
understanding, it appears that Brettanomyces is the only
microorganism that can produce levels of 4-ethylphenol
and 4-ethylguaiacol in wines near their corresponding odor
thresholds. While other species such as lactic acid bacteria have been shown to produce low levels of these analytes from the p-coumaric and ferulic acid precursors in
culture media,[17] Lactobacillus spp. are currently regarded
as non-significant contributors to 4-ethylphenol and 4ethylguaiacol concentrations in wines. However, it is important to note that our understanding of microbial sources of

Downloaded By: [Rayne, Sierra] At: 20:15 3 November 2007

888

Rayne and Eggers
O
HO

OH

cinnamate decarboxylase

vinylphenol reductase
HO

HO

Brettanomyces mediated
p-coumaric acid

-CO2

4-vinylphenol

4-ethylphenol

O
vinylphenol reductase

cinnamate decarboxylase
HO

OH

HO

HO

Brettanomyces mediated
-CO2

H3CO

H3CO

ferulic acid

4-vinylguaiacol

H3CO
4-ethylguaiacol

Fig. 1. Structures and formation pathways of 4-ethylphenol and 4-ethylguaiacol in wines.

4-ethylphenol and 4-ethylguaiacol is evolving, and is the
subject of substantial interest. Future findings may allow a
more quantitative consideration of non-Brettanomyces microbial sources of these analytes in wine, and their comparison to inputs from the abiotic sources discussed below.
Brettanomyces (the asexual, nonsporulating form) and
Dekkera (its sexual, sporulating form) are ubiquitous in
the vineyard and winery, and are likely to be present in the
water, the soil, the grapes and must, and throughout the
winery.[18] Once this yeast is established in a winery, it is difficult to eliminate.[13,18] Spoilage of wine by Brettanomyces
can be devastating and wineries have had to shut down to remove this contaminant. For these reasons, it is of value to be
able to rapidly and reliably distinguish between low-levels of
Brettanomyces derived 4-ethylphenol and 4-ethylguaiacol
and other sources of these analytes.
Thus, given the generally low levels of 4-ethylphenol and
4-ethylguaiacol we are observing in local Okanagan barreled reds (typically <50–60 µg/L for both analytes),[15–16]
we sought to better understand what the likely nonBrettanomyces sourced concentrations could be. Such
knowledge would then imply a limiting concentration below
which it cannot be reliably inferred whether the presence of
these compounds is coming by way of a Brettanomycesmediated path, or simply other non-Brettanomyces viticultural and enological sources. Given the current lack of
quantitative information on non-Brettanomyces microbial
sources of 4-ethylphenol and 4-ethylguaiacol in wines, the
work described here is not able to assess the potential contributions from other microorganisms towards total nonBrettanomyces sources of the analytes. This paper describes
our findings regarding the relative contributions of nonBrettanomyces sourced 4-ethylphenol and 4-ethylguaiacol
inputs to wines, and provides an estimated concentration
range below which it is unlikely to allow distinction between Brettanomyces and non-Brettanomyces sources of
these compounds.
As well as its presence as an aroma compound in wines, 4ethylphenol is widely used as source material in the production of reactive polymers, antioxidants, pharmaceuticals,
agriculture chemicals, and dyes, and is in the Organization

for Economic Cooperation and Development (OECD) list
of high production volume (HPV) compounds, and as a
plasticizing agent in plastics. However, despite its commercial use in high volumes, relatively little open-source toxicological work has been conducted on 4-ethylphenol and
4-ethylguaiacol. It is from this perspective that the current
work also considers the potential toxicological implications
of elevated volatile phenol concentrations in wines.

Materials and methods
Extraction of wine sample
Fifteen individual samples of barrelled red wines were
collected from three commercial wineries in the Okanagan Valley appellation of British Columbia, Canada during November and December 2006. Samples were collected from 225 L oak barrels using a 50 mL glass volumetric pipette that had been sterilized with neat ethanol
prior to use. Samples were placed in 50 mL amber glass
jars for transport and storage prior to analysis. Collected
wines were stored at 4◦ C prior to analysis, with storage
times ranging from <24 hrs to 7 days. Ten commercial
Riesling wines were obtained from a local liquor store
(Kelowna, British Columbia, Canada). The selection included seven Canadian Vintners Quality Alliance (VQA)
wines and three German Qualit¨atswein Riesling wines. The
Canadian Rieslings were from the Okanagan Valley, while
the German Rieslings were from the Pfalz, Mosel-SaarRuwer, and Rheingau regions.
A sample of wine (5.00 mL) was pipetted into a 10-mL
test tube. The sample was spiked with 2.03 µg of 4ethylphenol-d3 and 2.36 µg of 4-ethylguaiacol-d3 as internal standards. The solution was shaken for 10 s to ensure
mixing, and diethyl ether:pentane (2 mL of a 1:2 v/v solution) was added to the vial and shaken again for 20 s.
The sample was centrifuged until two distinct layers formed
(typically <1 min). A portion (about 1 mL) of the organic
phase was transferred directly from the test tube to a
3 mL vial, capped, and the extract injected directly into

889

Downloaded By: [Rayne, Sierra] At: 20:15 3 November 2007

4-ethylphenol and 4-ethylguaiacol in wines
the gas-chromatograph/mass-spectrometer (GC/MS) under the instrumental conditions described below.
Extraction of enological additions sample
Non-glycosidically bound 4-ethylphenol and 4ethylguaiacol were extracted from the commercial
enological additions (one mannoprotein extract, and three
enological tannin extracts; see below for details) by placing
about a 5.0 g sample of product (accurately weighed) in
a 50 mL Erlenmeyer flask, spiking with 90 µL (13.23 µg
of 4-ethylphenol-d3 and 9.18 µg of 4-ethylguaiacol-d3 ) of
mixed internal standard solution, adding 15 mL of diethyl
ether:pentane (1:2 v/v), and stirring the suspension with
a Teflon-covered magnetic bar for 15 min. A 1 to 2 mL
aliquot of the suspension was filtered through a pasteur
pipette packed with about 1 cm of Kimwipe directly into a
GC-MS microvial.
Glycosidically bound 4-ethylphenol and 4-ethylguaiacol
were extracted from the commercial enological additions
by placing about a 5.0 g sample of product in a 100 mL
Erlenmeyer flask, spiking with 90 µL (13.23 µg of 4ethylphenol-d3 and 9.18 µg of 4-ethylguaiacol-d3 ) of mixed
internal standard solution, adding 40 mL of a pH 5 phosphate/citrate buffer and 1 g of a commercial β-glucosidase
extract (NOVAROM G, Novo-Nordisk A/S, Bagsværd,
Denmark; 900 FDU 20◦ C g−1 ), and stirring the suspension with a Teflon-covered magnetic bar for 24 hours at
room temperature (about 25◦ C). The resulting solution was
then extracted by adding 15 mL of diethyl ether:pentane
(1:2 v/v), and stirring the suspension with a Teflon-covered
magnetic bar for 15 min. A 1 to 2 mL aliquot of upper (organic layer) of the suspension was filtered through a pasteur
pipette packed with about 1 cm of Kimwipe directly into a
GC-MS microvial.
Extraction of grape sample
Non-glycosidically bound 4-ethylphenol and 4ethylguaiacol were extracted from grape bunches by
adding between 250 to 300 g of sample (including whole
berries and stems) and 100 mL of deionized water to a
commercial food blender, spiking with 100 µL (14.7 µg
of 4-ethylphenol-d3 and 10.2 µg of 4-ethylguaiacol-d3 ) of
mixed internal standard solution, and blending on low
speed for 15 min until a fine consistency was achieved.
Fifty mL of diethyl ether:pentane (1:2 v/v) was added
to the blender and the solution blended on low speed
for 1 min. A 5 mL aliquot of the emulsified solution was
transferred to a small test tube, centrifuged for 5 min, and
the top organic layer (typically about 0.5 mL) placed in a
GC-MS microvial for analysis.
Glycosidically bound 4-ethylphenol and 4-ethylguaiacol
were extracted from grape bunches by adding between 250
to 300 g of sample (including whole berries and stems)
and 500 mL of deionized water to a commercial food

blender, spiking with 100 µL (14.7 µg of 4-ethylphenol-d3
and 10.2 µg of 4-ethylguaiacol-d3 ) of mixed internal standard solution, adding 3 g of a commercial β-glucosidase extract (NOVAROM G, Novo-Nordisk A/S, Bagsværd, Denmark; 900 FDU 20◦ C g−1 ), and blending on low speed for
15 min until a fine consistency was achieved. The resulting solution was then allowed to sit for 24 hours at room
temperature (about 25◦ C), after which 50 mL of diethyl
ether:pentane (1:2 v/v) was added to the blender and the
solution blended on low speed for 1 min. A 5 mL aliquot of
the emulsified solution was transferred to a small test tube,
centrifuged for 5 min, and the top organic layer (typically
about 0.5 mL) placed in a GC-MS microvial for analysis.
Instrumental analysis
4-Ethylphenol and 4-ethylguaiacol were analyzed on a
gas-chromatograph (Thermo Scientific Trace GC; Thermo
Fisher Scientific, Inc., Waltham, MA, USA) coupled to a
mass spectrometer (Thermo Scientific DSQ MS; Thermo
Fisher Scientific, Inc., Waltham, MA, USA). The GC column was an Agilent/J&W (Agilent Technologies, Santa
Clara, CA, USA) DB-1701 [(14%-cyanopropyl-phenyl)methylpolysiloxane] with dimensions of 30 m length ×
0.25 mm inside diameter × 0.25 µm film thickness and
ultra-high purity helium at 1.2 mL/min as the carrier gas.
The GC injector was operated in the split/splitless mode,
with the splitless time of 1.00 min followed by a split flow of
50 mL min−1 . The GC injector temperature was constant
at 220◦ C over the course of a sample run, with the oven
temperature held at 40◦ C for 1 min, ramped to 260◦ C at
8◦ C min−1 , and held at 260◦ C for 1 min, for a total run time
of 30 min (with an equilibration time of 0.5 min prior to
injection). The MS ion source temperature was 200◦ C and
the GC-MS transfer line temperature was 250◦ C. MS scans
were obtained in the selected ion monitoring (SIM) mode
operating at unit resolution with an emission current of
100 µA and a dwell time of 100 ms at each of the following
masses: m/z 107, 122, 125, 137, and 155.
Further details regarding synthesis and characterizing
of the stable isotope derivative internal standards, sample
workup, instrumental conditions, and the criteria used for
identification and quantitation are described in detail in our
previous work.[16]
Data analysis
For quantitative analysis, concentrations below the method
detection limits (MDLs; 0.5 µg/L for 4-ethylphenol and
0.1 µg/L for 4-ethylguaiacol) were set at one-half the MDL
(i.e., 0.25 µg/L for 4-ethylphenol and 0.05 µg/L for 4ethylguaiacol). Univariate relationships between variables
were investigated by parametric linear regression. Population means were tested by parametric single-factor analysis
of variance (ANOVA) for two groups of samples, or by the
parametric multiple comparisons Tukey-Kramer test with

Downloaded By: [Rayne, Sierra] At: 20:15 3 November 2007

890

Rayne and Eggers

pairwise comparisons for one-way layout design for three or
more groups of samples. All statistical tests used the software program KyPlot (v.2.0 b.15; KyensLab Inc., Tokyo,
Japan).
Toxicological meta-analysis
An exhaustive literature search was conducted to determine the range of 4-ethylphenol and 4-ethylguaiacol levels reported in wines worldwide, as well as any chronic
or acute toxicological studies and corresponding doseresponse relationships on these two compounds. The results
of this literature survey, coupled with available worldwide
wine consumption datasets, were used in the assessment.
In addition, a search was undertaken to determine if any
countries/regions are actively regulating, or planning to
formally regulate, levels of these compounds in alcoholic
beverages. No regions appear to undertaking or planning
such regulatory action.

ling grown in Ohio, USA.[19] This work initially implied
to us that 4-ethylphenol concentrations up to 70 µg L−1
could arise from the grapes and/or winemaking processes
in the absence of Brettanomyces yeasts. Our results with
the ten commercial Riesling wines reported here support
these previous findings, and conservatively indicate that 4ethylphenol concentrations less than about 50–100 µg L−1
should be considered ambiguous in terms of source (i.e.,
grapes, winemaking materials/processes common to both
red and white varietals, Brettanomyces, etc.).
Furthermore, there does not appear to be a measurable source of 4-ethylguaiacol in Riesling wines, suggesting that there are likely negligible sources of this analyte
in the corresponding grapes, or in the winemaking materials or processes that would be common to other red
and white varietals. Thus, we estimated that non-oak and
non-Brettanomyces sources of 4-ethylguaiacol contribute
<1 µg L−1 to the finished wine, although further studies
were undertaken to support this hypothesis, as are described
below.

Results and discussion
Contribution of mannoproteins, enological tannins,
and grapes

Analyte levels in commercially available
non-oaked riesling white wines
To help us in our non-Brettanomyces sourced background
investigations of 4-ethylphenol and 4-ethylguaiacol in oak
barrelled red wines, we initially analyzed ten commercial
non-oaked bottled Riesling white wines from Canada and
Germany (vintage years 2002 through 2004) (Table 1). The
Riesling wines contained between 5 to 25 µg L−1 of 4ethylphenol (mean = median = 17 µg L−1 ). All but one
Riesling wine had <0.1 µg L−1 of 4-ethylguaiacol. The remaining 2002 German Riesling had a 4-ethylguaiacol level
at 1.4 µg L−1 .
We have not seen previous reports of 4-ethylguaiacol in
Riesling wines, but a prior study did find 4-ethylphenol
concentrations between 30 to 70 µg L−1 in a 1992 RiesTable 1. Concentrations (µg L−1 ) of 4-ethylphenol and 4ethylguaiacol in ten commercial non-oaked bottled Riesling white
wines.
Wine
ID
1
2
3
4
5
6
7
8
9
10

Vintage

Country of
origin

4-ethylphenol

4-ethylguaiacol

2002
2002
2003
2003
2003
2003
2003
2004
2004
2004

Germany
Canada
Canada
Germany
Canada
Canada
Germany
Canada
Canada
Canada

25.5
20.9
17.6
19.8
19.9
5.4
13.1
17.3
13.7
14.8

1.4
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1

Mannoproteins are believed to promote the growth of malolactic bacteria, contribute to protein and tartrate stability of white wines, interact with aroma and phenolic compounds of red wines, decrease the astringency and bitterness of tannins, and increase the body of wine. They are
used in bottle fermentation of sparkling wine, barrel aging of still wines, and in the production of flor sherry-type
products.[20] Because of their use in winemaking for red
and white wines, we extracted and analyzed a commercial mannoprotein sample (Bio-Springer, Maisons-Alfort,
France) in duplicate, with and without exposure to a commercial β-glucosidase enzyme extract, and found no measurable 4-ethylphenol (<0.5 ng analyte g−1 of material) or
4-ethylguaiacol (<0.1 ng analyte g−1 of material) (Table 2).
At the recommended dosing rate of 0.4 g mannoprotein per liter of wine, our results indicate that this product would contribute <0.0002 µg L−1 of 4-ethylphenol and
<0.00004 µg L−1 of 4-ethylguaiacol to the finished wine.
While we have not performed an exhaustive survey on available commercial mannoprotein products, the insignificant
analyte yields observed for the current sample suggest that
these enological additions are not significant background
sources of 4-ethylphenol and 4-ethylguaiacol to wines. Even
if a different commercial mannoprotein had 4-ethylphenol
and 4-ethylguaiacol yield factors four orders of magnitude
greater than the product we analyzed, this hypothetical
product would only contribute <2 µg L−1 of 4-ethylphenol
and <0.4 µg L−1 of 4-ethylguaiacol to the finished wine.
In addition we analyzed three commercially available
enological tannin products (one from American Tartaric
Products, Inc., Larchmont, NY, USA; and two from

891

Downloaded By: [Rayne, Sierra] At: 20:15 3 November 2007

4-ethylphenol and 4-ethylguaiacol in wines

Table 2. Concentrations (ng g−1 ) of 4-ethylphenol and 4-ethylguaiacol in three commercial enological tannins, one commercial
mannoprotein, and whole grape bunches from two major red grape varieties, and the corresponding contributions to finished wines
Yield factor
(ng analyte g−1 material)

Dosing/production rate
(g material L−1 wine)

4-ethylphenol 4-ethylguaiacol
Commercial product
Mannoprotein
Without β-glucosidase exposure
With β-glucosidase exposure
Enological tannin (medium toast oak)
Without β-glucosidase exposure
With β-glucosidase exposure
Enological tannin (lightly toasted
American oak)
Without β-glucosidase exposure
With β-glucosidase exposure
Enological tannin (equal portions
grape/wood tannins)
Without β-glucosidase exposure
With β-glucosidase exposure
Whole grape bunches
Vitis vinifera L. cv. Cabernet Sauvignon
Without β-glucosidase exposure
With β-glucosidase exposure
Vitis vinifera L. cv. Pinot Noir
Without β-glucosidase exposure
With β-glucosidase exposure

Contribution to finished wine
(µg analyte L−1 wine)
4-ethylphenol 4-ethylguaiacol

<0.5
<0.5

<0.1
<0.1

0.40
0.40

<0.0002
<0.0002

<0.00004
<0.00004

<0.5
<0.5

16 ± 5a
18 ± 3

0.05–0.20
0.05–0.20

<0.0001
<0.0001

0.0006 to 0.004b
0.0008 to 0.004

<0.5
<0.5

12 ± 4
11 ± 4

0.05–0.20
0.05–0.20

<0.0001
<0.0001

0.0004 to 0.003
0.0004 to 0.003

<0.5
<0.5

<0.1
<0.1

0.05–0.20
0.05–0.20

<0.0001
<0.0001

<0.00004
<0.00004

0.062 ± 0.010 0.0087 ± 0.009 1450
0.052 ± 0.009 0.0095 ± 0.005 1450

0.075–0.10
0.059–0.088

0.012 to 0.014
0.013 to 0.015

0.048 ± 0.019 0.0089 ± 0.005 1450
0.057 ± 0.015 0.0084 ± 0.009 1450

0.042–0.097 0.012 to 0.014
0.061–0.0104 0.011 to 0.014

a

Error bars indicate the range of duplicate analyses for values above the method detection limits.
Wine contribution ranges were calculated using the following combinations: lowest observed yield factor × lowest dosing/production rate, and
highest observed yield factor × highest dosing/production rate.
b

Scott Laboratories, Petaluma, CA, USA) each in duplicate, and in the presence and absence of β-glucosidase.
None of the three commercial enological tannins had 4ethylphenol yield factors above the method detection limits
(<0.5 ng g−1 of material) in either the presence or absence
of β-glucosidase. Using the suggested dosing rates for these
products (0.05 to 0.2 g per liter of wine) as given by their
respective suppliers, the yield factors suggest that these materials will impart <0.0001 µg L−1 of 4-ethylphenol to the
finished wine.
In contrast, measurable yields of 4-ethylguaiacol were
extracted from the two entirely oak-derived enological tannins in the presence and absence of β-glucosidase (Table 2),
but no significant difference in yields was observed upon
enzyme exposure (p > 0.82). The medium toasted oak tannin yielded about 17 ng g−1 of 4-ethylguaiacol, compared to
about 12 ng g−1 for the lightly toasted American oak. The
average yields for these materials are not significantly different (p = 0.60). A sample of grape/wood enological tannin
(manufactured from equal source amounts of grape and
oak) contained <0.1 ng g−1 extractable 4-ethylguaiacol.
However, even at the higher extraction rates from the two
oak-derived tannin products, and at their maximum suggested dosing rates, these materials could not contribute

more than 0.004 µg L−1 of 4-ethylguaiacol to the finished
wine. As with the mannoprotein analyses, and although we
have not conducted an exhaustive survey on available commercial enological tannin products, the insignificant analyte yields observed for the current samples indicate that
these enological additions are not likely to be significant
background sources of 4-ethylphenol and 4-ethylguaiacol
to wines. Furthermore, the absence of significantly higher
yields of the 4-ethylguaiacol from the two commercial entirely oak-derived enological tannins upon β-glucosidase
exposure suggests that both aqueous and organic solvent
extractable forms of these analytes are dominantly in the
aglycone form within the product, not the glycosylated
form.
Having ruled out substantial 4-ethylphenol and 4ethylguaiacol contributions from mannoproteins and enological tannins towards levels in finished wines, we further extracted and analyzed in duplicate the following two
varieties of Vitis vinifera whole grape bunches (including
stems, seeds, skins, and juice): Vitis vinifera L. cv. Cabernet Sauvignon, and Vitis vinifera L. cv. Pinot Noir. Because these analytes could occur as their aglycones or as
glycosides in the grapes that could be liberated during
fermentation and aging, we further analyzed the samples

Downloaded By: [Rayne, Sierra] At: 20:15 3 November 2007

892
in duplicate with, and without, exposure to a commercially available β-glucosidase enzyme preparation. None of
the grape samples, either before or after exposure to βglucosidase, provided a significant analyte yield, although
measurable quantities of both 4-ethylphenol (ranging from
0.048 ± 0.019 to 0.062 ± 0.010 ng g−1 ) and 4-ethylguaiacol
(ranging from 0.0084 ± 0.009 to 0.0095 ± 0.005 ng g−1 )
were observed (Table 2). There were no significant differences in analyte yields in either grape variety before and
after exposure to β-glucosidase (p > 0.50).
On average, grape bunches are about 75% juice by
weight,[21] and thus, assuming all juice is quantitatively extracted and converted to a finished wine gives a production
factor of about 1350 g of grape bunch per liter of wine. Discussions with the winery contributing grapes samples as
part of the current study indicate their production factor
was 1450 g of grape bunch per liter of wine for the varieties being investigated. Using this production factor, our
results suggest that these two grape varieties can contribute
– at most – between 0.042–0.10 µg L−1 of 4-ethylphenol and
0.011–0.015 µg L−1 of 4-ethylguaiacol via raw grapes to the
finished wine.
Other than our current work, there appear to have
been few investigations considering 4-ethylphenol and 4ethylguaiacol contributions to wines from the grapes (and
no published numbers to the best of our knowledge). Levels of 4-ethylphenol at 1 µg L−1 and 4-ethylguaiacol below
detection limits in a Spanish red wine before oak aging
were reported,[22] but this includes any influence from the
vinification process. It is of note that after six months of
barrel aging in American and French oak, these Spanish
wines contained 18 to 42 µg L−1 of 4-ethylphenol and 6 to
11 µg L−1 of 4-ethylguaiacol (in free-run and press wines
with and without enological tannin additions). As is discussed below, these analyte levels are within our estimated
range of non-Brettanomyces derived sources, and are likely
due to extraction from the oak barrel during extended
aging.
Extraction from aging in oak barrels
Our findings that neither enological additions nor raw
grapes are likely contributors to more than 0.1 µg L−1 of 4ethylphenol and 0.01 µg L−1 of 4-ethylguaiacol in finished
wines led to an examination of the literature on extraction of
these compounds from the longer aging period in oak barrels for some red and white varieties. The details of aroma
compound extraction from oak wood have been studied in
some detail, and two mechanisms have been proposed:[23]
(1) a fast direct extraction of compounds already present in
the wood, and (2) a slow indirectss extraction from degradation of the wood.
Recent work has extended our understanding of the processes involved, and describe that during the first stage in
the extraction of compounds from oak wood, the wine penetrates slowly into the wood, with extraction kinetics in-

Rayne and Eggers
fluenced by wine penetration rate.[24] Over this ‘wetting’
timeframe—estimated to last around 4 months [25] —the
surface area of wetted wood continues to increase and extraction kinetics vary depending on the particular physicochemical characteristics of the compound. After the wetting
is completed, the system can be considered at steady-state
with temporally constant extraction rate coefficients. In
sum, the final process of extraction results from the kinetically opposing processes of extraction and sorption.
A previous work, in which toasted oak samples were
extracted with a hydroalcoholic model solution,[26] also
greatly assists in establishing non-Brettanomyces sourced
background levels of 4-ethylphenol and 4-ethylguaiacol
in oaked wines. This study gave the following extractive
yields (µg of analyte per gram of French oak) under light
(surface temperature of 115–125◦ C), medium (200–215◦ C),
and heavy (220–230◦ C) toasts: light toast, 4-ethylphenol
yield of 0.30 ± 0.24 µg g−1 (error bars are standard deviations) and 4-ethylguaiacol yield of 0.04 ± 0.02 µg g−1 ;
medium toast, 4-ethylphenol yield of 0.29 ± 0.04 µg g−1
and 4-ethylguaiacol yield of 0.14 ± 0.03 µg g−1 ; and heavy
toast, 4-ethylphenol yield of 0.26 ± 0.19 µg g−1 and 4ethylguaiacol yield of 0.04 ± 0.02 µg g−1 .
Full barrels of about 200 L capacity have about
90 cm2 L−1 of wood surface, and thus, each millimeter of
wine penetration into the oak (and ensuing analyte extraction) would contribute extraction from about 9 cm3 of oak
per liter of wine, or assuming a wood density of about
0.6 g cm−3 , about 5.4 g L−1 of oak.[27] Based on experience
in visual examinations of oak barrel cross-sections after use,
a total penetration depth of up to about 6 mm may be extracted by red wine given enough time.[27] This depth would
yield a total oak mass of 32.4 g L−1 extracted per barrel.
Using the extraction factors given above, and assuming
a 225 L barrel, gives an oak-derived 4-ethylphenol concentration ranging between 2 and 18 µg L−1 depending on
toasting level (and including variability within a toasting
level), and a corresponding range of 0.3 to 6 µg L−1 for 4ethylguaiacol using analogous calculations. It is generally
acknowledged that toasting is difficult to keep uniform,[27]
such that while winemakers may specify light, medium, or
heavy toasts, these vary by portion of a barrel, cooper,
and order. Hence, it is likely of little value to attempt
developing rigorous background extraction yields for 4ethylphenol and 4-ethylguaiacol linked to stated toasting
levels. Rather, it is probably more accurate to state that
concentrations of 4-ethylphenol at <20 µg L−1 and concentrations of 4-ethylguaiacol at <6 µg L−1 could be coming
from the toasted oak itself.
It is also possible that the model hydroalcoholic solution
used for determining 4-ethylphenol and 4-ethylguaiacol extraction factors from toasted oak[26] was not a good model
extractive for these compounds relative to real wines, and
may have underestimated actual yields. The model solution
did not contain tannins and other compounds that may
help solubilize 4-ethylphenol and 4-ethylguaiacol beyond

893

Downloaded By: [Rayne, Sierra] At: 20:15 3 November 2007

4-ethylphenol and 4-ethylguaiacol in wines
what a 10–15 v/v% ethanol solution can accomplish. Studies that develop extraction yields from oak using real
wine with known negligible starting (i.e., pre-extraction) 4ethylphenol and 4-ethylguaiacol concentrations are needed
to better resolve these issues, and to help establish reliable extraction factors of these compounds from nonBrettanomyces contaminated oak.
It has also been noted that the rate of extraction from oak
is rapid (on the order of weeks to extract up to >90% of
the analyte),[27] although ensuing barrel fillings yield progressively less extractive quantities per unit time (i.e., to
achieve the same levels of analyte, longer periods of time
are required for used barrels compared to their new and
freshly toasted counterparts). Conversely, barrels still yield
these compounds even after multiple uses—and even without fresh shaving and toasting between uses—possibly because of continuous flaking and cracking that opens up new
wood for eventual wine contact.
The source of 4-ethylphenol and 4-ethylguaiacol in the
oak is also of interest, and two hypotheses can be put forward for further testing:[27] (1) the microbially mediated
conversion of precursor glycosides to the corresponding 4ethylphenol and 4-ethylguaiacol aglycones, similar to the reported microbial transformation of scopolin and coumarin
glycosides in the oak to their aglycones (note that slower
abiotic hydrolysis may also contribute to this source); and
(2) direct formation of the free phenols and guaiacols during pyrolysis of carbohydrates and fragmentation of preexisting larger phenols (polyphenols and the oak lignin itself). Our work above on the oak-derived enological tannins
suggests that in some cases, the majority of these analytes
may be present in the wood in aglycone form, although
further work is necessary to extend these preliminary findings from a commercial oak tannin product to whole oak
barrels.
The extraction quantities of 4-ethylphenol and 4ethylguaiacol from the oak barrel may also depend on
the relative humidity in the cellar. As has been previously noted,[27] extractions within the oak require free liquid to diffuse back into the body of the barreled wine,
and that evaporation from the barrel must occur at—or
in front of—the free liquid boundary. Hence, a deeper
layer of wood would be fully wetted and extracted at
higher relative humidity values (which also leads to another corollary—maturation in oak should be more rapid
at higher cellar relative humidity values). This higher relative humidity effect on increasing phenolic extraction from
wood has been previously reported for brandies,[28] and
would reasonably be expected to extend to wines. We have
recently reported on 4-ethylphenol and 4-ethylguaiacol
levels in barrelled red wines,[15] but a re-analysis of our
dataset looking for correlations between cellar relative
humidity and analyte concentrations in the likely nonBrettanomyces sourced range (<50 µg L−1 ) found no significant (p > 0.05) relationships. This absence of any humidity
effects in our low-level 4-ethylphenol and 4-ethylguaiacol

Table 3. Concentrations (µg L−1 ) of 4-ethylphenol and 4ethylguaiacol in 15 commercial 2006 vintage oak barrelled
Okanagan red wines.

Winery
ID
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3

Variety
Dunkelfelder
Dunkelfelder
Pinot Noir
Pinot Noir
Merlot
Merlot
Merlot
Cabernet Franc
Cabernet Franc
Pinot Meunier
Pinot Noir
Cabernet
Sauvignon
Cabernet
Sauvignon
Merlot
Merlot

Length of
time in
barrel before
sampling
4-ethylphenol 4-ethylguaiacol
21 days
21 days
18 days
18 days
12 days
4 days
4 days
2 days
2 days
11 days
17 days
7 days

<0.5
<0.5
1.3
1.2
<0.5
1.5
1.1
<0.5
1.4
3.0
1.1
<0.5

<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1

7 days

1.4

3.0

9 days
9 days

<0.5
<0.5

4.8
4.3

dataset may be the result of the various factors at work
(e.g., different wineries, barrel types and ages, etc.) that
confound underlying univariate relationships. Additional
studies are warranted using controlled conditions and replicate wines/barrels to better understand the relationship between volatile phenol extraction into the wine and cellar
temperature and humidity conditions.
To better understand the short-term (<30 days) effects
of oak-barrel aging on 4-ethylphenol and 4-ethylguaiacol
concentrations, we sampled and analyzed 15 commercial
barrelled red wines from six different Vitis vinifera cultivars
(L. cv. Cabernet Franc, Cabernet Sauvignon, Dunkelfelder,
Merlot, Pinot Meunier, and Pinot Noir) in three wineries
having oak-barrel aging periods prior to sampling ranging
from two to 21 days (Table 3). The observed 4-ethylphenol
concentrations were all <3 µg L−1 and 4-ethylguaiacol levels were all <5 µg L−1 (with 80% of the values below the
0.1 µg L−1 method detection limit). These samples are a
strong composite of the sources of 4-ethylphenol and 4ethylguaiacol that may come from the grapes, enological
additions, the general winemaking process, and short-term
oak-barrel aging.
Other studies support the notion of measurable 4ethylphenol and 4-ethylguaiacol levels prior to extended
oak aging. 4-Ethylguaiacol levels of about 40 µg L−1 at the
start of barrel aging in oak were reported,[29] which subsequently increased to about 80 µg L−1 (the majority of the
increase coming between days 180 and 270 of barrel aging),
and little or no significant future change in levels during up
to a year of subsequent bottle aging. As well, high initial
levels of 4-ethylphenol (about 200 µg L−1 ) were found in the
study wine prior to aging. The 4-ethylphenol concentrations
during barrel and bottle aging followed a similar pattern to

Downloaded By: [Rayne, Sierra] At: 20:15 3 November 2007

894
that of 4-ethylguaiacol, with levels approximately doubling
to about 400 µg L−1 between days 180 and 270 of barrel aging, and likely a non-significant change in concentrations
over the subsequent one year of bottle aging.
4-ethylphenol from plastics in the winery
The compound 4-ethylphenol is widely used as source
material in the production of reactive polymers, antioxidants, pharmaceuticals, agriculture chemicals, dyes, as a
plasticizing agent in plastics, and is in the Organization
for Economic Cooperation and Development list of highproduction volume compounds.[30] Hence, 4-ethylphenol
could enter the wine at some point between the start of
grapegrowing and end of vinification as a component of
an agricultural chemical (i.e., possibly from the vineyard,
and thereby present from the start) or as the wine contacts various plastics in the winery itself. Based on our
work above that shows grapes contribute a negligible 4ethylphenol loading towards a finished wine, it is unlikely
that pre-vinification processes or viticultural chemical additions are relevant sources of this analyte.
It is difficult to conceive that in-winery sources would
account for a significant portion of 4-ethylphenol levels in
the hundreds to thousands of µg L−1 , but they could potentially contribute to levels in the 1 to 50 µg L−1 range
(e.g., 10 µg L−1 in a 225 L barrel only requires a total of
about 2 mg of material). Namely, a source of this compound
would need to be found in a winery (e.g., plastic material),
that source would need to contact the wine, and that exposure would need to result in a material transfer into the
wine. We are unable to find any data on leaching rates of
compounds like 4-ethylphenol from plastics into hydroalcoholic solutions that would mimic the matrix behavior of
wines. In addition, the wide range of plastic types in use in
a commercial winery makes such a source apportionment
study a difficult task.
However, recent work has shown styrene contamination
in wines from contact with synthetic materials such as closures and plastic or fiberglass containers during storage or
transport.[31] To extend these novel and important precedents to 4-ethylphenol seems reasonable, provided the appropriate source and exposure routes are present. These
researchers further noted that other volatile substances
which can be associated with polystyrene materials, such
as toluene, ethylbenzene, propylbenzene, and other alkylbenzenes, may cause container-related wine contamination
in storage or transit from leaching by unsuitable or faulty
epoxy-lined or fiberglass tanks or plastic containers.32 For
related plasticizing agents like alkylphenols, the risk is worthy of detailed investigations.
Toxicological considerations
Among the few toxicological studies, Thompson et al.[33]
found that 4-ethylphenol was metabolized to a reactive

Rayne and Eggers

HO
4-ethylphenol

enzyme mediated
oxidation
-H2

O

p-quinone methide

Fig. 2. Schematic of oxidative para-quinone methide formation
from 4-ethylphenol.

quinone methide intermediate (Fig. 2) by liver enzymes (although at relatively high lethal concentration, LC50 , values
near 115 mg L−1 – well above the highest ever reported levels in wine that range up to 5 mg L−1 (see, e.g., ref.[10,12] ),
and that this mechanism was important in the measured
cytotoxic effects.
The quinone methides—such as can be obtained oxidatively from 4-ethylphenol—can be formed during P450 dependent metabolism of alkylphenols, including the popular butylated hydroxytoluene (BHP) antioxidant that
is added to a variety of consumer materials (including
foodstuffs).[34] Quinone methide production is toxicologically significant, via any manner of human exposure, as
they are highly reactive electrophiles which can alkylate
cellular macromolecules such as proteins and DNA.[35–36]
Quinone methides can also be formed photochemically in
ultraviolet light,[37] and via dissolved oxygen in liquids and
the bloodstream, indicating there are likely several possible pathways by which 4-ethylphenol sourced from wines
could result in formation of these reactive intermediates in
vivo. Much analogous work is being undertaken in the environmental chemistry and toxicology fields regarding the
relationships between precursor compounds that may yield
reactive species, and toxicological responses in the body,
although the studies are very complex when attempting to
make reliable linkages, and often a low degree of confidence
arises between the true cause-effect connections.
While such long-term dose-response studies regarding
contaminants like 4-ethylphenol and biological effects in
humans from wine consumption are not a significant focus
in current wine chemistry research efforts, regulatory agencies worldwide are expending significant efforts towards
better understanding trace contaminant sources and effects
in drinking waters. It appears to be inevitable that the regulatory focus will also soon include wines and other highvolume alcoholic beverages (one could argue that some
portion of the population drinks nearly as much wine as
potable water in a day, thus making contaminant sources
in wines potentially more important than in drinking water
for this subset of society). As well, the present discussions—
and efforts (such as those demonstrated here) to place wine
chemistry studies in an adjunct toxicological perspective—
will help prepare the industry for answering future questions regarding long-term health impacts of the variety of
substances present in wide-ranging concentrations within
the consumer product.
Along a similar research path as Thompson et al.[33]
but using animal-dosing studies, Takahashi et al.[34] have


Related documents


am j enol vitic 59 2008 92 97
aust nz grapegrower winemaker 521 2007 64 69
october 2006 brett project update poster 4nov06
preliminary brett results poster 4sep06
aust nz grapegrower winemaker 516 2007 52 59
j chromatogr a 1167 2007 195 201


Related keywords