Food Chemistry 215 (2017) 341–346

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Determination of free fatty acids in beer

Elisabetta Bravi a,⇑, Ombretta Marconi a, Valeria Sileoni b, Giuseppe Perretti b
a
Italian Brewing Research Center (Centro di Eccellenza per la Ricerca sulla Birra – CERB), University of Perugia, Via S. Costanzo n.c.n., 06126 Perugia, Italy
b
Department of Agricultural, Food and Environmental Sciences, University of Perugia, Via S. Costanzo, n.c.n., 06126 Perugia, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Free fatty acids (FFA) content of beer affects the ability to form a stable head of foam and plays an
Received 29 October 2015 important role in beer staling. Moreover, the presence of saturated FAs is related sometimes to gushing
Received in revised form 13 July 2016 problems in beer. The aim of this research was to validate an analytical method for the determination of
Accepted 28 July 2016
FFAs in beer. The extraction of FFAs in beer was achieved via Liquid–Liquid Cartridge Extraction (LLCE),
Available online 29 July 2016
the FFAs extract was purified by Solid Phase Extraction (SPE), methylated by boron trifluoride in
methanol, and injected into GC-FID system. The performance criteria demonstrate that this method is
Chemical compounds studied in this article:
suitable for the analysis of medium and long chain FFAs in beer. The proposed method was tested on four
Caprylic acid (PubChem CID: 379)
Capric acid (PubChem CID: 2969)
experimental beers.
Lauric acid (PubChem CID: 3893) Ó 2016 Elsevier Ltd. All rights reserved.
Myristic acid (PubChem CID: 11005)
Myristoleic acid (PubChem CID: 5281119)
Palmitic acid (PubChem CID: 985)
Palmitoleic acid (PubChem CID: 445638)
Stearic acid (PubChem CID: 5281)
Oleic acid (PubChem CID: 445639)
Linoleic acid (PubChem CID: 5280450)
Linolenic acid (PubChem CID: 5280934)

Keywords:
Analysis
Beer
Free fatty acids
GC-FID
LLCE
SPE

1. Introduction
Although free fatty acids (FFA) are quantitatively only minor con-
stituents of beer their presence has long been considered to have
adverse effects on beer quality. A stable foam is an important aspect
of beer quality as perceived by the consumer (Segawa, Yamashita,
Mitani, & Masachika, 2002). Beer contains foam-positive com-
pounds (proteins and iso-a-acids) and foam-negative compounds
(alcohols and FFAs). FFAs weaken the foam film by absorbing on
the gas bubble surface or by interaction with foam-generating pro-
teins. Moreover, the detrimental effect caused by the different FFAs
is related to their surface absorption properties. Considering that the
molecular hydrophobicity of the saturated FAs increased as chain-
⇑ Corresponding author.
E-mail address: e_bravi@yahoo.it (E. Bravi).
length increased and that the unsaturated FAs are more hydrophilic
compared with the saturated with the same carbon-chain length
(double bonds are more hydrophilic than single bonds). Therefore,
an unsaturated FAs absorbs more weakly at the surface of gas-
bubbles than a saturated FAs of the same chain length (Dickie,
Cann, Norman, Bamforth, & Muller, 2001; Segawa et al., 2002).
Beer consists of many compounds (volatile and non-volatile) that
affect beer flavor, many of these aroma compounds are synthesized
by yeasts during fermentation, others derive directly from the raw
materials. FFAs represent one group of these compounds that affect
beer taste, and also the foam stability of beer (Bamforth, 1985;
Clapperton, 1978; Horak et al., 2009; Meilgaard, Dalgliesh, &
Clapperton, 1979). Medium-chain fatty acids such as hexanoic,
octanoic, and decanoic acids, responsible for rancid or goaty flavor
characteristics, are formed by yeasts during fermentation, and their
formation is influenced by yeast strain, original gravity, wort com-
position, and degree of aeration (Horak et al., 2009). Off-flavors

http://dx.doi.org/10.1016/j.foodchem.2016.07.153
0308-8146/Ó 2016 Elsevier Ltd. All rights reserved.
342 E. Bravi et al. / Food Chemistry 215 (2017) 341–346
due to these acids normally arise from excess formation during fer-
mentation and not for other causes such as infection or raw materi-
als. Since the aroma contributions for medium-chain FFAs are
additive, the rancid and goaty off-flavors can occur in beer even
though none of the individual FAs are present above threshold con-
centrations (Horak et al., 2009). Long-chain unsaturated FAs, such as
linoleic and linolenic acids, directly deriving from raw materials or
from the yeast metabolism, are of great importance because their
oxidative degradation may lead to the formation of a characteristic
aging flavor (Bravi, Perretti, Buzzini, Della Sera, & Fantozzi, 2009;
Horak et al., 2009; Kaneda, Kano, Kamimura, Osawa, & Kawakishi,
1990; Vanderhaegen, Nevin, Verachtert, & Derdelinckx, 2006) and,
although their level is normally very low in beer, an unexpected
increase, coupled with wrong storage conditions, can result in unde-
sirable stale flavor (Vanderhaegen et al., 2006). Finally, saturated
FAs, derived from both raw materials and yeast, seem to be related
to the spontaneous over foaming (gushing) in beer (Carrington,
Collett, Dunkin, & Halek, 1972; Christian, Titze, & Ilberg, 2011;
Muller, Schmid, Becker, & Gastl, 2010).
Due to their involvement in different aspects of beer quality,
simple and reliable methods for analyzing FFAs content and profile
in beer are desirable. Both HPLC and GC method has been used for
the determination of FAs in beer (Bravi, Perretti, Montanari, Favati,
& Fantozzi, 2007; Irwin & Thompson, 1987; Kaneda et al., 1990;
Kobayashi, Kaneda, Kano, & Koshino, 1993; Ng, 2002; Van der
Meersche, Devreux, & Masschelein, 1979). Chromatographic meth-
ods for analyzing fatty acids required a sample preparation step.
The extraction is a very critical step which often needs complex
procedures resulting in low recoveries. In the past, the approaches
to the problem of extraction and concentration of the FFAs in beer
have been based on steam distillation (Irwin & Thompson, 1987;
Van der Meersche et al., 1979) and liquid-liquid solvent extraction
techniques (Chen, Jamieson, & Van Gheluwe, 1980; DeVries, 1990;
Kaneda et al., 1990; Kobayashi et al., 1993; MacPherson & Buckee,
1974; Ng, 2002; Taylor & Kirsop, 1977). The steam distillation tech-
niques are time-consuming and labor intensive. Liquid-liquid
extraction procedures suffer from problems of emulsion formation,
responsible for analyte loss; moreover, they are complex, time-
consuming and labor-intensive and require large volumes of
organic solvents. On the other hand, the extraction of samples with
small volumes of solvent results in poor reproducibility and
extractability of some FAs because of their low solubility in the sol-
vent (Bravi, Benedetti, Marconi, & Perretti, 2014; Hawthorne, Jones,
Barrett, Kavanagh, & Clarke, 1986; Horak et al., 2009).
Some authors developed more modern procedures, based on
the use of solid-phase extraction (SPE). These procedures are quite
easy and fast, if compared with the previously mentioned method-
ologies, reproducible and allow to minimizing solvent consuming.
Battistutta, Buiatti, Zenarola, and Zironi (1994) reported on the
developed of a method for the determination of medium chain
FAs in beer by SPE, thus excluding the long chain FAs which are
also potentially dangerous for beer flavor. Schutz and Back
(2005) developed a method for the analysis of long chain FAs by
SPE, thus excluding the medium chain FAs (potentially dangerous
for beer flavor and foam). Other authors (Horak, Culik, Jurkova,
Cejka, & Kellner, 2008) reported on the Stir Bar Sorptive Extraction
(SBSE) of medium chain FAs in beer, thus excluding the long chain
ones. Moreover, low recoveries of short chain FAs and an interfer-
ence by ethanol concentration of beer, that could reduce the versa-
tility of the method, were observed. Horak et al. (2009) discussed
the advantages and the limitations of three different extraction
methods (SPE, used as reference, SPME, and SBSE) for the determi-
nation of FFAs in beer. For all the three methods, the medium chain
were determined as free fatty acids while the long chain as methyl
esters, so two run into GC-FID system were needed. The SPME is
limited because allows only the determination of medium chain
FAs, the SBSE is a time-consuming procedure, the SPE technique
is quite fast and lead good recoveries and reproducibility. Never-
theless, no validation procedure was carried out and the suitability
of the method for the quality control for the brewing process and
the assessment of the final product quality not guaranteed.
In order to determine low amounts (ppm) of FFAs in a aqueous
matrix, such as beer samples, the development of a simple,
accurate and efficient extraction method which avoids emulsion
formation and which allows the simultaneous extraction of a wide
range of fatty acids (with different chain length) is required. The
same authors reported on the development of a method for the
determination of medium and long chain FFAs in beer wort using
a sorbent assisted liquid–liquid extraction (by the use of the inert
support of ChemElut cartridges) that, avoiding emulsion formation,
facilitate efficient extraction of small amount of FFAs. In order to
evaluate the quality and to predict the stability of beer, by modify-
ing the method developed for the beer wort, the authors developed
and validate a reliable method for the quantification of FFAs in
beer. The proposed method was tested on four experimental beers.
2. Experimental methods
2.1. Materials
The unmalted and the flocked T. monococcum (emmer), variety
Monlis, both dehulled, samples were supplied by Prometeo srl
(Urbino, Italy).
Twenty commercial beer samples of the same brand and batch
were purchased from local market and were used for the develop-
ment and validation of the method. Beer experimental samples
were supplied by CERB (Italian Brewing Research Centre, University
of Perugia) and were used to test the developed method. Amber
beer sample (AB) was obtained using a mixture of different malts,
80% of Pilsner, 15% of CararedÒ, and 5% of CaramunichÒ (Weyer-
mann, Germany); blond beer sample (BB) was obtained using 80%
of Pilsner, 10% of CarapilsÒ, 5% of CarahellÒ (Weyermann, Ger-
many), and 5% of crude wheat (Prometeo srl, Italy); an emmer beer
sample (EB) was obtained using 75% of Pilsner, 15% of CarapilsÒ
(Weyermann, Germany), and 10% of crude emmer (Prometeo srl,
Italy); finally, an emmer malt beer (EMB) was obtained using a mix-
ture of two malts obtained by experimentally malted emmer, 95%
of Pils emmer and 5% of Caramel emmer. Emmer was malted in
an automatic micromalting system from Custom Laboratory
Products (Scotland) following the procedure described by Mayer,
Marconi, Perretti, Sensidoni, and Fantozzi (2011).
Two liters of each experimental beer sample were use for the
application of the developed method, all the experimental beer
samples were unfiltered.
The experimental beer samples were prepared as follows:
mashing was performed in a 110 L pilot plant and was carried
out on by infusion for all the considered beer samples.
For AB the temperature profile of mashing was as follows: (i)
first steady-state at 52 °C for 5 min, (ii) first rise of temperature
from 52 to 65 °C in 12 min, (iii) second steady-state at 65 °C for
46 min, second rise of temperature from 65 to 72 °C in 8 min, (iv)
third steady-state at 72 °C for 20 min, (v) final rise from 72 to
76 °C in 5 min, and final steady state at 76 °C for 5 min.
For BB the temperature profile of mashing was: (i) first steady-
state at 52 °C for 10 min, (ii) first rise of temperature from 52 to
65 °C in 13 min, (iii) second steady-state at 65 °C for 44 min, sec-
ond rise of temperature from 65 to 72 °C in 7 min, (iv) third
steady-state at 72 °C for 20 min, (v) final rise from 72 to 78 °C in
10 min, and final steady state at 78 °C for 3 min.
For EMB the temperature profile of mashing was as follows: (i)
first steady-state at 52 °C for 30 min, (ii) first rise of temperature
 E. Bravi et al. / Food Chemistry 215 (2017) 341–346
from 52 to 67 °C in 10 min, (iii) second steady-state at 67 °C for
15 min, second rise of temperature from 67 to 72 °C in 10 min,
(iv) third steady-state at 72 °C for 20 min, (v) final rise from 72
to 76 °C in 5 min, and final steady state at 76 °C for 10 min.
For EB the temperature profile of mashing was: (i) first steady-
state at 52 °C for 10 min, (ii) first rise of temperature from 52 to
65 °C in 14 min, (iii) second steady-state at 65 °C for 45 min, sec-
ond rise of temperature from 65 to 72 °C in 6 min, (iv) third
steady-state at 72 °C for 20 min, (v) final rise from 72 to 76 °C in
8 min, and final steady state at 76 °C for 10 min.
For all the considered beer samples, after the step of mashing,
the mash was filtered in a lauter tun. Sparging was executed twice
with 3 L and once with 2 L of brewing water at 78 °C. Next, the
mash was boiled for 60 min until reaching approximately 12°P.
At the beginning of this phase, hop pellets (Columbus, Centellian
and Summit variety, for RB; Saphir and Cascade variety, for BB;
Saphir variety, for EMB; and Saaz variety for EB), were added. After
boiling and elimination of the hot trub, the wort was cooled to
15 °C. Aeration of the wort was achieved.
For AB the pitching was carried out at 20 °C, by using a commer-
cial yeast (Safale Fermentis, S-05, Marcq-en-Baroeul Cedex,
France), 70 g/hL (dry weight) after rehydration with sterile water.
The fermentation temperature was maintained at 20 °C for 3 days.
For BB the pitching was carried out at 20 °C, by using a commer-
cial yeast (Safale, Fermentis, US-05, Marcq-en-Baroeul Cedex,
France), 70 g/hL (dry weight) after rehydration with sterile water.
The fermentation temperature was maintained at 20 °C for 7 days.
For EB and EMB samples the pitching was carried out at 20 °C,
by using a commercial yeast (Safale, Fermentis, WB-06, Marcq-
en-Baroeul Cedex, France), 70 g/hL (dry weight) after rehydration
with sterile water. The fermentation temperature was maintained
at 20 °C for 3 days.
For all the beer samples, at the end of fermentation, after cool-
ing and when the final attenuation was achieved, the maturation
lasted for a week at 4 °C. At the same time, the discharge of the
yeast from the bottom of the tank was performed. After this phase
of maturation, the beer was refermented in the bottle with the
addition of sucrose (4.5 g/L) and placed in a thermostat at 22 °C
for 2 weeks. At the end of the re-fermentation, the beer was stored
at 4–5 °C until analysis.
2.2. Other materials
HPLC grade n-hexane, diethyl ether, methanol, chloroform,
2-propanol, RPE grade glacial acetic acid, boron trifluoride (BF3)
methanol complex solution (13–15%), GC standards, SupelcoÒ 37
component FAME MIX, caprylic acid (P99%), margaric acid
(P98%), linoleic acid (P99%), oleic acid (P99%), and methyl non-
adecanoate (P98%) were purchased from Sigma-Aldrich S.r.l.
(Milan, Italy). Internal Standard (IS) solution was prepared by
diluting accurately weighed amounts (approx. 10 mg) of methyl
nonadecanoate to 200 ml with hexane. Diatomaceous earth car-
tridges (ChemElut), SPE C18 cartridges (SampliQ C18), and SPE
amino cartridges (SampliQ aminopropyl) were obtained from
Agilent Technologies (Milan, Italy). A VAC Elut 20 Tall Glass
Manifold (Agilent Technologies, Milan, Italy) was used.
2.3. Equipment and chromatographic conditions
An Agilent Model 6850 gas chromatograph equipped with a
Flame Ionization Detector (FID), a capillary inlet system, a DB-23
(60 m
0.25 mm
0.25 lm) column, and a Model Maestro MPS
2XL multipurpose sampler with a 10 ll syringe (Gerstel Inc., Balti-
more, MD) was employed. The programmed oven temperature
was: 130 °C for 1 min, 130–170 °C at 6.5 °C/min, 170–215 °C at
2.75 °C/min, 215 °C for 12 min, 215–230 °C at 40 °C/min, 230 °C
343
for 3 min. The carrier gas (H2) flow rate was 1.7 ml/min. Constant
pressure mode was set up. The split ratio was set at 50:1. The tem-
peratures of injector and of detector were 270 °C and 280 °C,
respectively. Peak areas were measured by using an Agilent MSD
Chemstation for HRGC-FID.
2.4. Extraction procedure
Beer samples were degassed and concentrated in a rotary evap-
orator (LABOROTA SEM-320, Resona Technics, Gossau, Switzer-
land) before the FAs determination.
2.4.1. Liquid–liquid cartridge extraction procedure
The extraction was carried using porous diatomaceous earth
(ChemElut cartridges). A sample of 40 ml of concentrated beer
was loaded on a dry ChemElut cartridge and, after 15 min, the col-
umn was washed with two 20 ml portions and with two 10 ml por-
tion of n-hexane/diethyl ether (1/1; v/v) to extract the lipid
fraction. The effluents, containing the lipid fractions, were col-
lected in a 100 ml bottom flask and evaporated to dryness under
vacuum at 37 °C. The purification and the esterification of the FFAs
portion were conducted following the procedure reported by Bravi
et al. (2014) for the determination of FFAs content in beer wort.
The extracted lipids, dissolved in chloroform, were loaded in a
aminopropyl cartridge previously conditioned with 4 ml of n-
hexane. Then, 4 ml of chloroform-isopropanol (2/1; v/v) were
added to remove all the neutral lipids, and the FFA fraction was
eluted with 4 ml of 2% acetic acid in diethyl ether. The methylation
was performed using 1 ml of BF3 methanolic solution (13–15%) at
100 °C for one minute. The FAME fraction was dissolved in 1 ml of
IS solution and injected into GC-FID system.
2.5. Validation
Three related performance characteristics, trueness (how close
the mean of ten replication is to a reference value), precision
(how close results are to one another), and uncertainty (is an inter-
val associated with a measurement result which expresses the
range of values that can reasonably be attributed to the quantity
being measured) were used to describe the quality of the results
obtained with the developed method (AOAC International, 2013;
Eurachem Group, 2014). Each step of the entire methodology (con-
centration of beer sample, liquid–liquid cartridge extraction, SPE
purification, methylation, and GC-FID analysis) was replicated ten
times in order to validate the method.
2.6. Statistical analysis
The FFAs determination on different experimental beer samples
was performed in triplicate and the data were analyzed using
MATLAB Statistics Toolbox (version 7.6.0, The Mathworks Inc., Nat-
ick, MA, USA) to perform the appropriate statistical tests (One-Way
and Two-Ways ANOVA). Values were considered significantly dif-
ferent at P < 0.001.
3. Results and discussion
The method development, its validation for the determination
of FFAs in beer and its application as study case are reported.
3.1. Method development and validation
Since the extraction is the critical time-consuming step in the
available methods (Chen et al., 1980; DeVries, 1990; Hawthorne
et al., 1986; Horak et al., 2009; Irwin & Thompson, 1987; Kaneda
344 E. Bravi et al. / Food Chemistry 215 (2017) 341–346

et al., 1990; Kobayashi et al., 1993; MacPherson & Buckee, 1974;
Ng, 2002; Taylor & Kirsop, 1977; Van der Meersche et al., 1979),
in the proposed method the FFAs were extracted from the beer
samples using a single step extraction procedure with ChemElut
cartridges. The several advantages of these columns are previously
described by the same authors for the beer wort analyses (Bravi
et al., 2014). In order to improve the extraction of medium chain
fatty acids, present in beer in higher amount then in wort and
responsible for rancid or goaty flavor characteristics (Horak et al.,
2009), a mixture of n-hexane/diethyl ether was used with ChemE-
lut cartridges. The saturated medium chain FAs (as octanoic,
decanoic, and dodecanoic acids) have a small polar portion and a
larger non-polar rod-like hydrocarbon chain that makes these
molecules behave more like a non-polar compound. Thus, the
non-polar eluent n-hexane promote the elution of saturated short
and medium chain FAs. Different ratio of n-hexane/diethyl ether
were tested to verify the efficiency of the solvent mixture and of
the extraction method; the commercial beer sample was fortified
with 2 mg/L of caprylic, margaric, oleic and linoleic acid (saturated
medium and long chain, unsaturated and polyunsaturated long
chain FAs) and the recoveries were determined by analyzing each
of spiked samples six times. The mean percentage recoveries of
each spiking fatty acid extracted with the different extraction mix-
ture tested are reported in Table 1. It is evident that the best sol-
vent mixture for the extraction of FFAs of beer is n-hexane/
diethyl ether, 50/50 (v/v), that allows the higher recovery percent-
ages of all tested FAs (78% caprylic, 77% heptadecanoic, 104% oleic,
and 102% linoleic acid, respectively). The use of ChemElut car-
tridges allowed therefore the rapid and solvent-saving extraction
of lipids from the aqueous beer matrix, after the extraction the
FFAs were separated from the lipid extract by SPE, methylated
and injected to GC-FID system.
The trueness of the method was determined by replicate analy-
sis (n = 10) of test sample unspiked and spiked with the four differ-
ent FAs (caprylic, margaric, oleic and linoleic acid) at three
different concentrations (0.4, 1, and 2 mg/L of beer). The trueness
was calculated to compare the difference between mean spiked
value and mean value with the added concentration of analyte
and was expressed as the relative spike recovery, R0 % (Eurachem
Guide, 2014). The obtained results are reported in Table 2. The
trueness values ranged between 77% and 104%. The recovery
acceptability is function of the analyte concentration and of the
purpose of the analysis. In the present study (considering the
ppm content of FFAs in beer) the recovery limit it is expected to
be between 80 and 115%, according to the AOAC (2002). Therefore,
the results we obtained are in the requested range, only two data
points for caprylic acid were lower than 80%, but the mean of the
recoveries for this fatty acid was 80%, which fall in the range of
acceptability.
The within-laboratory precision was tested by analyzing beer
samples fortified with 0.4, 1, and 2 mg/L of the four different med-
ium and long chain fatty acids ten times. The relative standard
deviation of repeatability (%RSDr) values were calculated from
these data. The results are shown in Table 2. The precision ranged
from 0.4% to 2.0% and is within acceptability criteria for %RSD (±2%)
(Ahuja & Scypinski, 2001; Food and Drug Administration, ORA
Laboratory Procedure, 2014; Shabir, 2003; Sistla, Tata, Kashyap,
Chandrasekar, & Diwan, 2005).
The uncertainty of the developed method was determined cal-
culating the Horwitz ratio value (HorRat), the relative results are
reported in Table 2. The HorRat is a variation to the Horwitz equa-
tion to be used on single-laboratory validation study (AOAC
International, 2013). The predicted repeatability relative standard
deviation (PRSDr) for the levels analyzed was calculated according
to the Horwitz equation: PRSDr = 2C0.1505, where C is the mass
fraction. The ratio of the RSDr calculated from the data to the
PRSDr calculated from the Horwitz equation results in the HorRat
value: HorRat = RSDr/PRSDr. Accepted values are between 0.3 and
1.3, under repeatability conditions. The acceptability of the HorRat
values for within-laboratory relative standard deviations on the
low side (values <0.3) may indicate unreported averaging or excel-
lent training and experience (AOAC International, 2013).
The system suitability, expressed as the %RSD of retention time
and peak area for each FAME, the range of linearity, the LOD, deter-
mined as three times the signal-to-noise ratio, and the LOQ, as ten
times the signal-to-noise ratio, of each FAME were calculated and
reported in the analysis of FFAs in wort by GC-FID by the same
authors (Bravi et al., 2014).

Table 1
Recovery of fatty acids, based on one concentration level (2 mg/L) fatty acids spiked in beer, as a function of different solvent mixture.

Fatty acids Extraction mixture

Diethyl ether n-hexane/diethyl ether (30/70, v/v) n-hexane/diethyl ether (50/50, v/v) n-hexane/diethyl ether (70/30, v/v)
Recovery % (mean ± SD) Recovery % (mean ± SD) Recovery % (mean ± SD) Recovery % (mean ± SD)
Caprylic acid (C8:0) 25.2 ± 0.3 35.0 ± 0.0 77.5 ± 0.0 50.0 ± 0.1
Margaric acid (C17:0) 55.6 ± 0.1 57.3 ± 0.3 77.3 ± 0.2 61.7 ± 0.1
Oleic acid (C18:1) 63.0 ± 0.0 78.8 ± 0.2 104.0 ± 0.0 69.9 ± 0.1
Linoleic acid (C18:2) 76.0 ± 0.1 79.6 ± 0.4 101.9 ± 0.1 75.2 ± 0.1

Mean of six replications; SD, standard deviation.

Table 2
Trueness and precision and Horwitz ratio (HorRat) of the optimized method, based on three concentration levels (0.4, 1 and 2 mg/L) of fatty acids spiked in beer.

Fatty acids Concentration level (mg/L)

0.4 1 2
Trueness R0 % Precision % HorRat Trueness R0 % Precision % HorRat Trueness R0 % Precision % HorRat
(mean ± SD) RSDr (mean ± SD) RSDr (mean ± SD) RSDr
Caprylic acid (C8:0) 77 ± 1 1.9 0.1 86 ± 2 2.0 0.1 78 ± 1 1.2 0.1
Margaric acid (C17:0) 80 ± 2 2.0 0.1 81 ± 1 1.1 0.1 81 ± 1 0.9 0.1
Oleic acid (C18:1) 83 ± 1 1.0 0.1 94 ± 4 1.2 0.1 104 ± 3 1.4 0.1
Linoleic acid (C18:2) 84 ± 2 0.4 0.1 96 ± 2 0.7 0.1 102 ± 0 1.2 0.1

Mean of ten replications. SD, standard deviation; R0 , recovery; RSD, relative standard deviation; HorRat, Horwitz ratio.
 E. Bravi et al. / Food Chemistry 215 (2017) 341–346 345

Table 3
Free fatty acids patterns of the different experimental beer samples.

Fatty acids AB BB EMB EB

Mean ± SD (mg/L) Mean ± SD (mg/L) Mean ± SD (mg/L) Mean ± SD (mg/L)
Caprylic acid (C8:0) nd 0.0363a ± 0.0011 0.0368a ± 0.0014 0.0483b ± 0.0019
Capric acid (C10:0) 0.0442a ± 0.0008 0.0872b ± 0.0006 0.3086c ± 0.0013 0.0937d ± 0.0022
Lauric acid (C12:0) 0.1192a ± 0.0008 0.4951c ± 0.0011 0.3570b ± 0.0014 0.5593d ± 0.0015
Myristic acid (C14:0) 0.1163a ± 0.0007 0.1195a ± 0.0008 0.1380b ± 0.0012 0.1715c ± 0.0030
Myristoleic acid (C14:1) 0.0431a ± 0.0021 0.0714b ± 0.0009 0.0712b ± 0.0020 0.0733b ± 0.0017
Palmitic acid (C16:0) 0.9814b ± 0.0006 0.9296a ± 0.0022 1.8188d ± 0.0257 1.3165c ± 0.0015
Palmitoleic acid (C16:1) 0.0513a ± 0.0010 0.0731b ± 0.0021 0.0951c ± 0.0028 0.0792b ± 0.0017
Stearic acid (C18:0) 0.8820a ± 0.0019 1.1875b ± 0.0057 1.6850c ± 0.0034 1.6664c ± 0.0220
Oleic acid (C18:1) 0.9353a ± 0.0012 1.3246b ± 0.0200 2.3723c ± 0.0051 2.4378d ± 0.0035
Linoleic acid (C18:2) 2.4649b ± 0.0016 3.7143d ± 0.0079 1.8807a ± 0.0036 3.3900c ± 0.0021
Linolenic acid (C18:3 x3) 0.8142a ± 0.0011 1.3430b ± 0.0157 nd 1.8006c ± 0.0271

Mean of three replications. SD, standard deviation; values in the same row followed by different superscript letters are statistically different (P < 0.001). AB, amber beer; BB,
beer with 5% of crude wheat; EMB, emmer malt beer; EB, beer with 10% of crude emmer; nd, not detectable.

Considering the performance criteria, the developed method
can offer an useful tool for quality control in brewing industries.
Considering the influence of the raw materials on the FA pattern
of finished beer (Bravi, Sensidoni, Floridi, & Perretti, 2009), in the
study case the proposed method was tested on experimental beers
brewed with different raw materials (barley malt, malted emmer,
crude wheat, and crude emmer).
3.2. Application of developed method
The method developed in this study was employed for the
determination of FFAs in four different experimental beer samples
(AB, BB, EMB, and EB). The results are reported in Table 3. The FFA
content and profile of beer is affected by the metabolically active
yeast cells dispersed in unfiltered beers, as well as raw materials
(malts, adjuncts, and hops) and mashing. The FAs content of beer
is also influenced by the process of maturation, during the matura-
tion yeast growth determines an increased of biosynthesized FAs
(Bravi, Perretti, et al., 2009). Moreover, the yeast strain and its
metabolism and catabolism influences the FAs in beer samples.
The EB and the BB showed the higher content of FFAs,
11.638 mg/L and 9.382 mg/L respectively; probably the presence
of crude cereals as adjunct influenced the lipid content of final
products. Moreover, different malt and hops with different FA con-
tent, that influence the FA pattern of finished beer, were used.
The lower content of FFAs was observed for amber beer sample
(6.450 mg/L), the higher temperatures necessary to obtain more
coloured malt during malting process could probably reduce the
lipid content of malted cereals.
In the analyzed beer samples, the most representative fatty
acids were the unsaturated oleic and linoleic acids, but also the
linolenic acid (not detectable only in EMB sample), and the satu-
rated palmitic and stearic acids.
For amber beer the more representative FA was linoleic
(2.465 mg/L), followed by palmitic (0.981 mg/L), oleic (0.935 mg/
L), stearic (0.882 mg/L), and linolenic (0.814 mg/L) acids,
respectively. For the crude wheat added BB, linoleic acid was the
major FA (3.714 mg/L), followed by linolenic (1.343 mg/L), oleic
(1.325 mg/L), stearic (1.188 mg/L), palmitic (0.930 mg/L), and
lauric (0.495 mg/L) acids. This beer, added with wheat, showed
higher level of linoleic acid (principal component of wheat lipids)
in comparison with all-malt beer (AB), this is probably due to the
presence of crude wheat and to the fatty acid profile of its lipids
(Macmurray & Morrison, 1970; Welch, 1975). Oleic acid was the
dominant FA (2.372 mg/L) in EMB, followed by linoleic
(1.880 mg/L), palmitic (1.819 mg/L), and stearic (1.685 mg/L) acids.
The FA composition of T. dicoccum, showing an higher content of
oleic acid compared with FA pattern of barley malt (Suchowilska
et al., 2009), probably influenced the oleic content of final beer,
as well as, the yeast strain and its metabolism that also deeply
affect the FAs content in beer (Bravi, Perretti, et al., 2009). For
the crude emmer added beer (EB), linoleic acid (3.390 mg/L) was
the major FA, followed by oleic (2.438 mg/L), linolenic (1.800 mg/
L), stearic (1.666 mg/L), palmitic (1.317 mg/L), and lauric
(0.559 mg/L) acids. The presence of crude emmer may influence
the FAs level in final beer, in fact the linoleic acid is the major con-
stituent of the lipids of T. dicoccum, followed by oleic acid. The oleic
acid, in particular, was found at higher percentage (26%) in emmer
than in barley malt (12%) (Anness, 1984; Bravi, Marconi, Perretti, &
Fantozzi, 2012; Suchowilska et al., 2009).
4. Conclusions
In conclusion, a simple and reliable GC-FID method with liquid–
liquid cartridge extraction and SPE purification was developed for
the quantitative determination of FFAs in beer. The LLCE with
ChemElut cartridges and n-hexane/diethyl ether mixture allows
the efficient extraction (recovery between 78% and 104%) of
lipophilic analytes (medium and long chain fatty acids), in traces
in aqueous medium, from a difficult matrix as beer, that it is gen-
erally recognized in the literature for its dishomogeneity (liquid,
gas, solids). The performance criteria of the proposed method
demonstrated that it is useful for the analysis of a wide range of
FFAs in different beer samples at suitable levels. The applicative
analyses of FAs in experimental beers confirmed the validity of
the developed method when applied on beer samples experimen-
tally brewed with different raw materials (malt, hop, and crude
cereals) and different yeast, highlighting their influence on the
FAs pattern of final beer.
References
Ahuja, S., & Scypinski, S. (2001). Handbook of modern pharmaceutical analysis
(pp. 415–443). New York: Academic Press.
Anness, B. J. (1984). Lipids of barley, malt and adjuncts. Journal of the Institute of
Brewing, 80, 315–318.
AOAC. (2002). Guidelines for single laboratory validation of chemical methods for
dietary supplements and botanicals.
AOAC International (2013). Definitions and calculations of Horrat values from
interlaboratory data. Accessed 1 October 2015 www.aoac.org.
Bamforth, C. W. (1985). The foaming properties of beer. Journal of the Institute of
Brewing, 91, 370–383.
Battistutta, F., Buiatti, S., Zenarola, C., & Zironi, R. (1994). Rapid analysis of free
medium chain fatty acids and related ethyl esters in beer using SPE and HRGC.
Journal of High Resolution Chromatography, 17, 662–664.
Bravi, E., Benedetti, P., Marconi, O., & Perretti, G. (2014). Determination of free fatty
acids in beer wort. Food Chemistry, 151, 374–378.
Bravi, E., Marconi, O., Perretti, G., & Fantozzi, P. (2012). Influence of barley variety
and malting process on lipid content of malt. Food Chemistry, 135, 1112–1117.
346 E. Bravi et al. / Food Chemistry 215 (2017) 341–346
Bravi, E., Perretti, G., Buzzini, P., Della Sera, R., & Fantozzi, P. (2009). Technological
steps and yeast biomass as factors affecting the lipid content of beer during the
brewing process. Journal of Agricultural and Food Chemistry, 57(14), 6279–6284.
Bravi, E., Perretti, G., Montanari, L., Favati, F., & Fantozzi, P. (2007). Supercritical fluid
extraction for quality control in beer industry. Journal Supercritical Fluids, 42,
342–346.
Bravi, E., Sensidoni, M., Floridi, S., & Perretti, G. (2009). Fatty acids composition in
beer lipids and determination of corn adjuncts. MBAA Technical Quarterly, 46(4).
http://dx.doi.org/10.1094/TQ-46-3-0916-01.
Carrington, R., Collett, R. C., Dunkin, I. R., & Halek, G. (1972). Gushing promoters and
suppressants in beer and hops. Journal of the Institute of Brewing, 78(3),
243–254.
Chen, E. C.-H., Jamieson, A. M., & Van Gheluwe, G. (1980). The release of fatty acids
as a consequence of yeast autolysis. Journal of the American Society of Brewing
Chemists, 38, 148–153.
Christian, M., Titze, J., & Ilberg, V. (2011). Chemical structure of model substances
related to their gushing-inducing and -suppressing activity. Journal of the
American Society of Brewing Chemists, 69(3), 170–179.
Clapperton, J. F. (1978). Fatty acids contributing to caprylic flavour in beer. The use
of profile and threshold data in flavour research. Journal of the Institute of
Brewing, 84, 107–112.
DeVries, K. (1990). Determination of free fatty acids in wort and beer. Journal of the
American Society of Brewing Chemists, 48(1), 13–17.
Dickie, K. H., Cann, C., Norman, E. C., Bamforth, C. W., & Muller, R. E. (2001). Foam-
negative materials. Journal of the American Society of Brewing Chemists, 59(1),
17–23.
Eurachem Working Group. (2014). Eurachem guide. The fitness for purpose of
analytical methods. A laboratory guide to method validation and related topics.
Middlesex, U.K.
Food and Drug Administration, ORA Laboratory Procedure (2014). Methods, method
verification and validation, laboratory manual, (Vol. II, Document number:
ORA-LAB.5.4.5). Accessed 12/10/2015 http://www.fda.gov/downloads/
ScienceResearch/FieldScience/LaboratoryManual/UCM092147.pdf.
Hawthorne, D. B., Jones, R. D., Barrett, P. A., Kavanagh, T. E., & Clarke, B. J. (1986).
Methods for the analysis of C4 to C10 fatty acids in beer, wort and carbohydrate
syrup. Journal of the Institute of Brewing, 92, 181–184.
Horak, T., Culik, J., Cejka, P., Jurkova, M., Kellner, V., Dvorak, J., et al. (2009). Analysis
of free fatty acids in beer: Comparison of solid-phase extraction, solid-phase
microextraction, and stir bar sorptive extraction. Journal of Agricultural and Food
Chemistry, 27, 11081–11085.
Horak, T., Culik, J., Jurkova, M., Cejka, P., & Kellner, V. (2008). Determination of free
medium-chain fatty acids in beer by stir bar sorptive extraction. Journal of
Chromatography A, 1196–1197, 96–99.
Irwin, A. J., & Thompson, D. J. (1987). A rapid method for the extraction and analysis
of beer flavour components. Journal of the Institute of brewing, 93, 113–115.
Kaneda, H., Kano, Y., Kamimura, M., Osawa, T., & Kawakishi, S. (1990). Analysis of
long-chain fatty acids in beer by HPLC-fluorescence detection method. Journal of
Agricultural and Food Chemistry, 38, 1363–1367.
Kobayashi, N., Kaneda, H., Kano, Y., & Koshino, S. (1993). Determination of fatty acid
hydroperoxides produced during the production of wort. Journal of the Institute
of Brewing, 99, 143–146.
Macmurray, T. A., & Morrison, W. R. (1970). Composition of wheat-flour lipids.
Journal of the Science of Food and Agriculture, 21, 489–544.
MacPherson, J. K., & Buckee, G. K. (1974). Estimation of free fatty acids (C8–C18) in
wort and beer. Journal of the Institute of Brewing, 80, 540–544.
Mayer, H., Marconi, O., Perretti, G., Sensidoni, M., & Fantozzi, P. (2011). Investigation
of the suitability of hulled wheats for malting and brewing. Journal of the
American Society of Brewing Chemists, 69(2), 116–120.
Meilgaard, M. C., Dalgliesh, C. E., & Clapperton, J. F. (1979). Beer flavour terminology.
Journal of the Institute of Brewing, 85, 38–42.
Muller, M. P., Schmid, F., Becker, T., & Gastl, M. (2010). Impact of different hop
compounds on the over foaming volume of beer caused by primary gushing.
Journal of the Institute of Brewing, 116(4), 459–463.
Ng, L. K. (2002). Analysis by gas chromatography/mass spectrometry of fatty acids
and esters in alcoholic beverages and tobaccos. Analytica Chimica Acta, 465,
309–318.
Schutz, M., & Back, W. (2005). The influence of mash acidification on long-chain
fatty acid content in wort determinate by a new developed method.
Monatsschrift fur Brauwissenschaft, 58, 6–10.
Segawa, S., Yamashita, S., Mitani, Y., & Masachika, T. (2002). Analysis of detrimental
effect on head retention by low molecular weight surface-active substances
using surface excess. Journal of the American Society of Brewing Chemists, 60,
31–36.
Shabir, G. A. (2003). Validation of high-performance liquid chromatography
methods for pharmaceutical analysis: Understanding the differences and
similarities between validation requirements of the US Food and Drug
Administration, the US Pharmacopeia and the International Conference on
Harmonization. Journal of Chromatography A, 987, 57–66.
Sistla, R., Tata, V. S. S. K., Kashyap, Y. V., Chandrasekar, D., & Diwan, P. V. (2005).
Journal of Pharmaceutical and Biomedical Analysis, 39, 517–522.
Suchowilska, E., Wiwart, M., Borejszo, Z., Packa, D., Kandler, W., & Krska, R. (2009).
Discriminant analysis of selected yield components and fatty acid composition
of chosen Triticum monococcum and Triticum dicoccum and Triticum spelta
accessions. Journal of Cereal Science, 49, 310–315.
Taylor, G. T., & Kirsop, B. H. (1977). Rapid method for the analysis of the fatty acids
of fermenting wort and beer (C6–C10) and yeast (C8–C18). Journal of the
Institute of Brewing, 83, 97–99.
Van der Meersche, J., Devreux, A., & Masschelein, C. A. (1979). Formation des acides
volatils dans la maturation de la biere. In Proceedings of the European Brewery
Convention congress, Berlin (pp. 787–800). Zoeterwoude, The Netherlands: EBC.
Vanderhaegen, B., Nevin, H., Verachtert, H., & Derdelinckx, G. (2006). The chemistry
of beer aging – A critical review. Food Chemistry, 95, 357–381.
Welch, R. (1975). Fatty acid composition of grain from winter and spring sown oats,
barley and wheat. Journal of the Science of Food and Agriculture, 26, 429–435.