fermentation

Article
Yeast Diversity and Physicochemical Characteristics
Associated with Coffee Bean Fermentation from the
Brazilian Cerrado Mineiro Region
Dão Pedro de Carvalho Neto 1 , Gilberto Vinícius de Melo Pereira 1 , Valcineide O. A. Tanobe 1 ,
Vanete Thomaz Soccol 1 , Bruno José G. da Silva 2 , Cristine Rodrigues 1 and
Carlos Ricardo Soccol 1, *
1 Bioprocess Engineering and Biotechnology Department, Federal University of Paraná (UFPR),
81531-970 Curitiba-PR, Brazil; daopcn@gmail.com (D.P.d.C.N.); gilbertovinicius@gmail.com (G.V.d.M.P.);
valcitanobe@gmail.com (V.O.A.T.); vanetesoccol@gmail.com (V.T.S.); cristinelabor@gmail.com (C.R.)
2 Chemistry Department, Federal University of Paraná (UFPR), 81531-970 Curitiba-PR, Brazil;
bruno@quimica.ufpr.br
* Correspondence: soccol@ufpr.br; Tel.: +55-41-33-613-191; Fax: +55-41-33-613-695

Academic Editor: Eleftherios H. Drosinos

Received: 20 January 2017; Accepted: 12 March 2017; Published: 16 March 2017

Abstract: The aim of this study was to evaluate the yeast diversity and physicochemical characteristics
of spontaneous coffee bean fermentation conducted in the coffee-producing region in Cerrado Mineiro,
Brazil. During 48 h of fermentation, yeast population increased from 6.60 to 7.89 log CFU·g−1 , with
concomitant pulp sugar consumption and organic acids production (mainly lactic (3.35 g·L−1 ) and
acetic (1.27 g·L−1 ) acids). According to ITS-rRNA gene sequencing, yeast population was mainly
represented by Saccharomyces sp., followed by Torulaspora delbrueckii, Pichia kluyveri, Hanseniaspora
uvarum, H. vineae and Meyerozyma caribbica. SPME-GC-MS analysis revealed a total of 25 volatile
organic compounds with predominance of hydrocarbons (9 compounds) and higher alcohols
(6 compounds). The resulting fermented, roasted coffee beans were analyzed by diverse chemical
analysis methods, including Fourier Transform Infrared (FTIR) spectroscopy and mineral and
thermogravimetric analysis. The thermal decomposition of the coffee beans occurred in four stages
between 90 and 390 ◦ C, with significant mass loss (68%) after the second stage at 190 ◦ C. FTIR
spectroscopy confirmed the presence of the main organic functions associated with coffee aroma,
such as aromatic acids, ketones, aldehydes and aliphatic esters. The results presented in this study
enrich our knowledge concerning yeast diversity and physicochemical characteristics associated with
coffee bean fermentation, and can be used to promote a controlled on-farm processing.

Keywords: wet processing; coffee; yeasts; coffee fermentation; coffee beverage

1. Introduction
Coffea is a genus of flowering plants whose seeds are used to make coffee beverage. Although there
are many steps in the manufacturing of coffee beverage, microbial fermentation plays an important
role with great impacts on product quality and value [1–5]. Coffee fermentation consists of an on-farm
process, occurring in the so-called wet processing, during which microorganisms grow in the pulp
material that surrounds the seeds of the coffee fruit [6]. In this processing method, the pulp (the exocarp
and a part of the mesocarp) is removed mechanically. Subsequently, the beans are submitted to 24–48 h
of underwater tank fermentation to allow the microbial degradation of the remaining mesocarp
layer (called mucilage) adhering to the parchment. In addition, the microbiota responsible for the
fermentation (e.g., yeasts and lactic acid bacteria) may also contribute to the beverage’s sensory

Fermentation 2017, 3, 11; doi:10.3390/fermentation3010011 www.mdpi.com/journal/fermentation

Fermentation 2017, 3, 11 2 of 11

beverage’s sensory characteristics and other qualities due to the excretion of metabolites produced
Fermentation
during this2017, 3, 11 [6]. The main chemical changes that occur during coffee fermentation are pectin
process 2 of 11

degradation and microbial production of organic acids, ethanol, esters and other metabolites from
the carbohydrates [1,7,8].
characteristics and other qualities due to the excretion of metabolites produced during this process [6].
Yeasts are among the most frequently isolated microorganisms from fermenting coffee beans.
The main chemical changes that occur during coffee fermentation are pectin degradation and microbial
They are considered to be important to the fermentation performance and coffee flavor
production of organic acids, ethanol, esters and other metabolites from the carbohydrates [1,7,8].
development. Consequently, yeast is the microbial group most widely studied in coffee
Yeasts are among the most frequently isolated microorganisms from fermenting coffee beans.
fermentations, which metabolic function has been elucidated in recent studies [1,2]. The most
They are considered to be important to the fermentation performance and coffee flavor development.
frequently occurring yeast species during coffee processing are Pichia kluyveri, Pichia anomala,
Consequently, yeast is the microbial group most widely studied in coffee fermentations, which
Hanseniaspora uvarum, Saccharomyces cerevisiae, Debaryomyces hansenii and Torulaspora delbrueckii
metabolic function has been elucidated in recent studies [1,2]. The most frequently occurring
[9–11].
yeast species during coffee processing are Pichia kluyveri, Pichia anomala, Hanseniaspora uvarum,
Recent studies published by our research group have reported the yeast and bacteria diversity
Saccharomyces cerevisiae, Debaryomyces hansenii and Torulaspora delbrueckii [9–11].
associated with coffee bean fermentation in Brazil [1,4–6]. Pichia fermentans (YC5.2), Saccharomyces sp.
Recent studies published by our research group have reported the yeast and bacteria diversity
(YC9.15) and Lactobacillus plantarum (LPBR01) were studied as having a potential for use as starter
associated with coffee bean fermentation in Brazil [1,4–6]. Pichia fermentans (YC5.2), Saccharomyces sp.
cultures for coffee wet fermentation [4,5]. However, studies still are needed to improve the
(YC9.15) and Lactobacillus plantarum (LPBR01) were studied as having a potential for use as starter
knowledge of the microbiota present in coffee processing due to the variation in climate and altitude
cultures for coffee wet fermentation [4,5]. However, studies still are needed to improve the knowledge
[3]. Cerrado Mineiro is one of the largest coffee-producing regions in Brazil, located at Alto
of the microbiota present in coffee processing due to the variation in climate and altitude [3]. Cerrado
Paranaíba, Mineiro Triangle and Northwest Minas Gerais state. This region presents a uniform
Mineiro is one of the largest coffee-producing regions in Brazil, located at Alto Paranaíba, Mineiro
edaphoclimatic pattern with an average temperature of 23 °C and flat relief situated at 800–1300 m,
Triangle and Northwest Minas Gerais state. This region presents a uniform edaphoclimatic pattern
which enables the production of high-quality coffees [12]. To the best of our knowledge, there is no
with an average temperature of 23 ◦ C and flat relief situated at 800–1300 m, which enables the
study on the yeast diversity and physicochemical characteristics associated with coffee bean
production of high-quality coffees [12]. To the best of our knowledge, there is no study on the yeast
fermentation performed in this coffee-producing region. In this regard, the aim of this study was to
diversity and physicochemical characteristics associated with coffee bean fermentation performed
study the yeast diversity and physicochemical characteristics associated with coffee bean
in this coffee-producing region. In this regard, the aim of this study was to study the yeast diversity
fermentation during on-farm wet processing in the Cerrado Mineiro region.
and physicochemical characteristics associated with coffee bean fermentation during on-farm wet
processing in the Cerrado Mineiro region.
2. Materials and Methods
2. Materials and Methods
2.1. Spontaneous Coffee Bean Fermentation and Sampling
2.1. Spontaneous Coffee Bean Fermentation and Sampling
Coffee cherries (Coffea arabica) were manually harvested at the mature stage from the Fazenda
Shalon localized
Coffee cherriesin the Cerrado
(Coffea arabica)Mineiro region (Patrocí
were manually nio, at
harvested Minas Gerais stage
the mature State,from
Brazil).
the Figure
Fazenda 1
shows
Shalon the experimental
localized setup and
in the Cerrado analytical
Mineiro region procedures
(Patrocínio, of eachGerais
Minas step conducted in this
State, Brazil). study.
Figure The
1 shows
fruits were mechanically
the experimental depulped
setup and andprocedures
analytical approximately 75 kg
of each of depulped
step conductedbeans were
in this conveyed
study. in a
The fruits
clear mechanically
were water stream depulped
to tanks andandleft to ferment for
approximately 48 of
75 kg h depulped
in accordance
beanswith
werelocal wet processing
conveyed in a clear
method
water [4,5]. to
stream Every
tanks12and
h, liquid
left to fraction
ferment samples were
for 48 h in withdrawn
accordance withfrom
localthe
wetmiddle depth
processing of the tank
method [4,5].
fermentation
Every 12 h, liquidand fraction
transferred to the
samples werelaboratory
withdrawninfromice the
boxes for microbiological
middle depth of the tankand chemical
fermentation
analyses.
and transferred to the laboratory in ice boxes for microbiological and chemical analyses.

Figure 1. Experimental
Experimental setup
setup of
of the case study of coffee-processing experiments
experiments carried
carried out at the
Cerrado Mineiro region (Brazil).

2.2.
2.2. Yeast Isolation
Yeast Isolation
Ten millilitersofofeach
Ten milliliters eachsample
sample was
was added
added to mL
to 90 90 mL sterile
sterile saline-peptone
saline-peptone water,
water, followed
followed by
by serial
serial dilutions. Yeasts were enumerated by surface inoculation on Yeast extract
dilutions. Yeasts were enumerated by surface inoculation on Yeast extract peptone glucose (YEPG) peptone glucose
(YEPG) agar containing
agar containing 100 mg·L100 mg·L−1 chloramphenicol
−1 chloramphenicol to inhibit to inhibitgrowth.
bacterial bacterialPlating
growth. wasPlating was
performed
performed
with 100 µLwith
of each100dilution
μL of each dilution were
and cultures and incubated
cultures were
at 30incubated at 30
◦ C for 5 days [1].°CAccording
for 5 days to [1].
the
macroscopic observations, colonies of different types on YEPG medium were counted separately,
Fermentation 2017, 3, 11 3 of 11

and representatives isolated from different fermentation times were purified by repetitive streaking.
The purified isolates were stored at −80 ◦ C in YEPG broth containing 20% (v/v).

2.3. Identification of Yeast Isolates

Yeast cultures were grown under appropriate conditions, collected from agar plates with a sterile
pipette tip and resuspended in 50 µL of ultra-pure water. The suspension was heated for 15 min
at 95 ◦ C, and 1 µL of this suspension was used as a DNA template in PCR experiments. The 5.8S
ITS rRNA gene region of yeast isolates was amplified using the primers ITS1 and ITS4. The 55 µL
volume reaction consisted of 5.5 µL of 10× PCR buffer (Invitrogen, Carlsbad, CA, USA), 2 µL of MgCl2
(50 mM), 1.21 µL of dNTP Mix (10 mM), 4 µL of the combined forward and reverse primers (ITS1 and
ITS4), 0.4 µL of 5 U·µL−1 Platinum® Taq DNA polymerase (Invitrogen, Waltham, MA, USA). The 5.8S
ITS rRNA gene region was sequenced using an ABI3730 XL automatic DNA sequencer. The sequences
obtained were compared with sequences available in the GenBank database through a basic local
alignment search tool (BLAST).

2.4. High Performance Liquid Chromatography (HPLC) Analysis of Fermenting Coffee Beans Samples
The concentration of the reducing sugars (glucose and fructose) and organic acids (acetic, citric,
succinic, lactic, propionic and butyric acids) of fermenting coffee-pulp bean mass was monitored
during the course of fermentation. Samples of each time were centrifuged at 6000 g and filtered
through 0.22-µm pore size filter (Sartorius Stedim, Goettingen, Germany). The samples were analyzed
through a HPLC apparatus (Aglient Technologies 1260 Infinity Series; Aglient Technologies, Santa
Clara, CA, USA) equipped with a Hi-Plex H column (300 mm × 7.7 mm; Aglient Technologies, Santa
Clara, CA, USA) connected to a refractive index (RI) detector (Aglient Technologies, Santa Clara, CA,
USA). The column was eluated with a mobile phase containing 5 mM H2 SO4 at 60 ◦ C and a flow rate
of 0.6 mL·min−1 .

2.5. Physicochemical Characterization of Fermented and Roasted Coffee Beans

2.5.1. Volatile Organic Compounds Determination by Gas Chromatography Coupled to Mass

Spectrophotometry (GC-MS)
The resulting parchment coffee was sun dried until a water content of 12% (wet basis) was
achieved. The extraction of volatile compounds from the fermented, dried coffee samples (Figure 1) was
performed using a headspace vial coupled to an Solid phase microextraction (SPME) fiber (5% Carboxen
[CARB]/95% Polydimethylsiloxane [PDMS] df75 µm partially crosslinked, Supelco, St. Louis, MI,
USA). For each determination, 1 g of sample was stored in a 20 mL HS vial. The flask was heated
at 70 ◦ C for 10 min without shaking, followed by 15 min of fiber exposure in COMBI-PAL system
for balancing the volume within the vial. The compounds adsorbed by the fiber were desorbed into
the gas chromatograph injection system gas phase (CGMS-gun TQ Series 8040 and 2010 Plus GC-MS
Shimadzu, Tokyo, Japan) at 250 ◦ C. The compounds were separated on a column 95% PDMS/5%
PHENYL (30 m × 0.25 mm, 0.25 mm film thickness). The GC was equipped with an HP 5972 mass
selective detector (Hewlett Packard Enterprise, Palo Alto, CA, USA). Helium was used as carrier gas
at a rate of 1.0 mL·min−1 . Mass spectra were obtained by electron impact at 70 eV. The compounds
were identified by comparison to the mass spectra from the library database (Nist’98 and Wiley7n).

2.5.2. Mineral Analysis

Mineral analysis was performed of fermented, dried and roasted coffee beans (Figure 1).
The fermented, dried coffee samples were roasted in a semi-industrial roaster (Probatino, Leogap
model, Brazil) at 140 ◦ C for 30 min. For sample preparation, fermented and roasted cocoa samples
were transferred to a 250 mL volumetric flask and acidified with 5 mL of concentrated P.A. HNO3
and H2 O2 30%. Subsequently, the system was allowed to heating for 40 min at 80 ◦ C. The extract was
Fermentation 2017, 3, 11 4 of 11

filtered by a 0.45 µm pore size filter and the volume was completed to 100 mL. Reference solutions were
prepared using deionized water with resistivity of 18.2 MΩ·cm−1 through a Milli-Q water purification
system linked to a water distillatory Fisatom (Model 534, Brazil). Glassware used in this procedure
was subjected to a decontamination treatment with HNO3 10% (w/v) for 24 h prior to use.
For mineral content determination, an ICP-OES (Varian, Model ES 720, Palo Alto, CA, USA)
was used simultaneously with axial arrangement and solid-state detector. The torch was aligned
horizontally and vertically with a Mn2+ standard solution concentration of 5.0 mg·L−1 . The optical
system of the ICP OES was calibrated with multi-element stock solution of the scanned patterns.
Spectral lines were selected considering the absence of interferences and appropriate sensitivity for
determining elements in high and low concentrations. The operation conditions were as follows:
power of 1.10 kW, plasma gas flow of 15 L·min−1 , auxiliary gas flow of 1.5 L·min−1 , nebulizer pressure
of 180 kPa, triplicate time read of 3 s, stabilization time of 15 s, sample delay of 30 s, pump speed of
15 rpm and sample washing time of 3 s.

2.5.3. Fourier Transform Infrared (FTIR) Spectroscopy

Functional groups in samples of grounded coffee beans (fermented and roasted) were determined
by FTIR on a VERTEX 70 (Bruker, Billerica, MA, USA) containing a DRIFT accessory with 64 scans and
a 4 cm−1 resolution at the 4000 to 400 cm−1 wave length region. The samples were crushed, pulverized
and oven dried. Before determination, about 20 mg of the samples were mixed and homogenized with
100 mg of Potassium bromide (KBr), and the reads were recorded.

2.5.4. Thermal Stability

Thermal stability of fermented, dried coffee beans was evaluated by Thermogravimetry (TG)
analysis. Analyzes were performed under an O2 (g) atmosphere at 20 ◦ C·min−1 rates to a
maximum temperature of 800 ◦ C in a Setsys Evolution TG/DTA/DSC (SETARAM, Hillsborough, NJ,
USA) system.

3. Results and Discussion

3.1. Microbiological and Chemical Characterization of Coffee Bean Fermentation

Descriptions of the load of yeast in coffee fermentation have been provided by some
studies [9,10,13,14]. This has been reported to range between 2 to 7 log CFU·mL−1 , depending on the
study. Factors affecting the initial yeast load include the quality and integrity of the coffee beans and
the hygiene of fermentation tank, utensils and water used at the commencement of the fermentation
process [6]. In this study, the average number of yeast, which was 6.60 log CFU·mL−1 early in the
fermentation process, increased to 7.89 log CFU·mL−1 by the end of 48 h of fermentation (Figure 2).
This growth is favored by the ability of yeast cells to metabolize coffee pulp sugars as well to adapt and
to cope with the hostile environment and stress conditions prevailing in coffee fermentation matrix [1].
The yeast growth was accompanied by a regular consumption of pulp sugars (glucose and
fructose) and their conversion into organic acids. Lactic acid was the major metabolite produced,
reaching a concentration of 3.28 g·L−1 at 48 h, followed by acetic and succinic acids (1.27 and 0.30 g·L−1 ,
respectively). The low production of acetic acid (<1.5 g·L−1 ) and the absence of butyric and propionic
acids minimize the formation of off-flavors in the final beverage [15,16]. On the other hand, lactic acid
production can assist in the coffee-pulp acidification process without interfering in the final product
quality [6]. Such organic acids production during coffee fermentation is mainly associated with lactic
acid bacteria metabolism [5]. However, yeasts of the genera Saccharomyces, Pichia and Hanseniaspora
may also have produced a fraction of the concentration of organic acids found in the coffee pulp beans
mass in this study [17–19].
Fermentation 2017, 3, 11 5 of 11
Fermentation 2017, 3, 11 5 of 11

Fermentation 2017, 3, 11 5 of 11

Figure 2. Total yeast count, organic acids production and pulp-sugar consumption during coffee
Figure 2. Total yeast count, organic acids production and pulp-sugar consumption during coffee
bean fermentation.
Figure
bean 2. Total yeast count, organic acids production and pulp-sugar consumption during coffee
fermentation.
bean fermentation.
3.2. Yeast Identification
3.2. Yeast Identification
3.2. Yeast Identification
A total of 35 yeasts were isolated at the beginning and end of the fermentation process and
identified by 35
A total of yeasts were
ITS-rRNA gene isolated at the
sequencing beginning
(Figure 3). Theand endfrequently
most of the fermentation process
detected species wereand
A total of 35 yeasts were isolated at the beginning and end of the fermentation process and
identified by ITS-rRNA gene sequencing (Figure 3). The most frequently detected species
Saccharomyces sp. (17 isolates), Torulaspora delbrueckii (6 isolates) and Pichia kluyveri (7 isolates). were
identified by ITS-rRNA gene sequencing (Figure 3). The most frequently detected species were
Saccharomyces sp. (17 isolates),
However, Saccharomyces sp.Torulaspora
was founddelbrueckii (6 isolates)
to be dominant the Pichia
at and end ofkluyveri (7 isolates). process
the fermentation However,
Saccharomyces sp. (17 isolates), Torulaspora delbrueckii (6 isolates) and Pichia kluyveri (7 isolates).
Saccharomyces sp. was
proving its easy found to be
adaptation the dominant at the endenvironment.
coffee fermentation of the fermentation process
In addition, provingtoits
the ability
However, Saccharomyces sp. was found to be dominant at the end of the fermentation process
easy adaptation
metabolize to the
pulp coffee
coffee fermentation
pectin showed by environment. In addition,
some Saccharomyces the might
species abilitybe
to metabolize
proving its easy adaptation to the coffee fermentation environment. In addition, the ability to
considered an pulp
advantage
coffee pectin over
showedotherbynon-pectinolytic
some yeasts [1].
Saccharomyces species might be considered an advantage over other
metabolize pulp coffee pectin showed by some Saccharomyces species might be considered an
non-pectinolytic yeasts [1].
advantage over other non-pectinolytic yeasts [1].

Figure 3. Distribution and frequency of yeast isolated during coffee bean fermentation.

Figure
Figure 3. 3. Distributionand
Distribution andfrequency
frequency of
of yeast
yeast isolated
isolated during
duringcoffee
coffeebean
beanfermentation.
fermentation.
Fermentation 2017, 3, 11 6 of 11

Two isolates of Hanseniaspora uvarum and one isolate of each Meyerozyma caribbica, Torulaspora sp.
and Hanseniaspora vineae were identified at the beginning of the fermentation process. These yeast
species have been previously found in coffee processing environments [1,9,11], except for H. vineae
which was isolated for the first time. H. vineae is mainly associated with grapes and has been
demonstrated to increase fruity aromas of wine by producing a high amount of acetate esters, such
as 2-phenylethyl acetate and ethyl [20–22]. For coffee fermentation, these flavor-active esters could
attribute distinct fruity sensory notes to the coffee bean through their diffusion during the fermentation
process, enriching the flavor of the final beverage [6]. Thus, this yeast species should be included in
research programs for the selection and development of functional starter cultures.

3.3. Volatile Organic Compounds Determination of Fermented Cocoa Beans by Gas Chromatography Coupled to
Mass Spectrophotometry (GC-MS)
Yeast fermentation of pulp sugars produces a vast array of volatile metabolites that are well
known for their aromatic and flavorant properties [4,23]. These volatiles can diffuse into the coffee
beans which may influence in its chemical composition [4]. In this study, a total of 25 volatile organic
compounds were identified in the fermented, dried coffee bean samples by SPME-GC-MS analysis,
with predominance of hydrocarbons (9 compounds) and higher alcohols (6 compounds) (Table 1).
Although most of the compounds identified originate from the bean itself, some are known to be
related bacteria (i.e., nonanal, citric acid and heptanal) and yeast (i.e., hexane, heptane and thiophenes)
metabolism [24]. Despite the diffusion mechanism have not yet been elucidated, it is often referenced
in the literature that these volatile organic compounds diffuse into the beans [1,5,25]. Further research
to understand how these volatiles are conserved during roasting operation and reach the final coffee
product is required.

Table 1. Volatile compounds identified in fermented coffee samples by Gas Chromatography Coupled
to Mass Spectrophotometry (GC-MS).

Organic Functional Group Volatile Organic Compound

Nonanal
Aldehydes (2)
Heptanal
5-tert-Butyl-1,3-cyclohexanediol
2-Propyl-1-pentanol
Alcohol (6)
4-Ethyl-1-octyn-3-ol
2-ethyl-1-Decanol
13-(Benzyloxy) tridecanoic acid
Dodecanoic acid, 3-hydroxy-
Carboxylic acid (4)
Methacrylic acid
Acetic acid
Ester (1) Heptyl valerate
Heptane, 2,2,3,5-tetramethyl-Pentane, 2,2,3,4-tetramethyl-Heptane,
2,2,6,6-tetramethyl-Hexane, 2,2,5-trimethyl-Eicosane,
Hydrocarbons (9)
3-methyl-Dodecane, 2,6,11-trimethyl-Heptane,
5-ethyl-2,2,3-trimethyl-Hexadecane 9-Octadecene, 1,1-dimethoxy-
Sulfur Compounds (1) 2-Allyl-3-methyl-4-(phenylsulfanyl)-2,5-dihydrothiophene, 1,1-dioxide
Ketone (1) p-Benzoquinone
Pyrazine (1) 2-Isobutyl-3-methoxypyrazine
Fermentation 2017, 3, 11 7 of 11

3.4. Physicochemical Characterization of Coffee Beans

The mineral composition of fermented, dried and roasted samples is shown in Table 2. Potassium
displayed the higher amount among the minerals analyzed in fermented and roasted coffee samples
(12,453.10 and 13,117.50 mg·kg−1 , respectively), followed by phosphorus (1932.76 and 2110.71 mg·kg−1 ,
respectively), magnesium (1554.31 and 1772.65 mg·kg−1 , respectively) and calcium (1360.19 and
1192.32 mg·kg−1 , respectively). These results are in agreement with those found by Martín et al. [26],
except for the high aluminum content present in our samples. Although the Coffea arabica is not
reported as an aluminum accumulator, the levels of such metal in the analyzed samples may be
indicative of a soil with high availability of this metal for plant [27].

Table 2. Content of minerals on fermented and roasted coffee samples.

Trait
Minerals (mg·kg−1 )
Fermented Coffee Roasted Coffee
a
Al 234.71 ± 29.22 362.56 ± 106.12 b
Ba 3.40 ± 0.49 c 2.61 ± 0.09 c
B 7.28 ± 0.30 c 6.26 ± 0.21 c
Cd ND c ND c
Ca 1360.19 ± 20.53 d 1192.32 ± 0.88 e
Co ND c ND c
Cu 16.51 ± 0.38 c 17.57 ± 0.25 c
Fe 33.75 ± 1.32 c 33.12 ± .44 c
P 1932.76 ± 43.56 f 2110.71 ± 28.40 g
Li ND c ND c
Mg 1554.31 ± 17.47 h 1772.65 ± 24.87 i
Mn 16.32 ± 0.01c 17.43 ± 0.07 c
Mo ND c ND c
Ni ND c ND c
K 12453.10 ± 8.35 j 13117.50 ± 16.41 k
Se ND c ND c
Na 350.565 ± 17.83 b 380.99 ± 16.40 b
V 0.86 ± 0.05 c 0.47 ± 0.04 c
Zn 10.15 ± 0.12 c 9.78 ± 0.30 c
* Means of triplicate in each row bearing the same letters are not significantly different (p > 0.05) from one another
using Duncan’s Test (mean ± standard variation). ND: not detectable. Al = Aluminum; Ba = Barium; B = Boron; Cd
= Cadmium; Ca = Calcium; Co = Cobalt; Cu = Copper; Fe = Iron; P = Phosphorus; Li = Lithium; Mg = Magnesium;
Mn = Manganese; Mo = Molybdenum; Ni = Nickel; K = Potassium; Se = Selenium; Na = Sodium; V = Vanadium;
Zn = Zinc.

Over the last two decades, vibrational spectroscopy methods have proven to be a reliable and
fast technique for the identification and quantification of several primary and secondary metabolites
generated during fermentation processes or to estimate the quality of the food itself [28–32]. In this
study, FTIR spectroscopy analysis showed a quite similar spectrum for both fermented, dried and
roasted coffee beans (Figure 4). It was possible to verify the presence of the main organic functions
associated with the coffee aroma, such as aromatic acids (1700–1680 cm−1 ), ketones (1725–1705 cm−1 ),
aldehydes (1739–1724 cm−1 ) and aliphatic esters (1755–1740 cm−1 ). Those results corroborate the wide
variety of volatile organic compounds identified by the SPME-GC-MS technique (Table 1).
 Fermentation 2017, 3, 11 8 of 11
Fermentation2017,
Fermentation 2017,3,3,1111 8 8ofof1111

Figure4.4.Functional
Figure Functionalgroups
groupspresent
presentininfermented,
fermented,drieddriedandandroasted
roasted samplesofofcoffee
coffeebeans
beans
Figure 4. Functional groups present in fermented, dried and roasted samples samples
of coffee beans determined
determined
determined by Fourier
by Fourierof transform
transform of infrared (FTIR) spectroscopy on range of 400–4000 wavenumber.
by Fourier transform infrared of infrared
(FTIR) (FTIR) spectroscopy
spectroscopy on range of on400–4000
range of wavenumber.
400–4000 wavenumber.
The main
Themain
The mainbands
bandsidentified
identifiedand
andorganic
organicfunctions
functionsassociated
associatedwere:
were:O–H
O–H(3350
(3350
− 1 and3010
and 3010cm cm
−1; ;
−1
bands identified and organic functions associated were: O–H (3350 and 3010 cm ; alcohols); C–H
alcohols);
alcohols); C–H (2930 and 2856 cm −1; alkanes); C=O (1739 cm−1; carboxylic acids and esters); C–O (1650
(2930 andC–H 2856 (2930
cm−1 ;and 2856 cm
C=O; alkanes);
(1739 cm−C=O (1739 cmacids
−1; carboxylic acids and esters);−
and esters); C–O (1650 cm C–O (1650
−1 1 ; carboxylic 1 ; alcohols
alkanes);
cmcm −1; alcohols or phenols).
−1; alcohols or phenols).
or phenols).

Thermoanalytical methods,
Thermoanalytical methods, especially
especially thermogravimetry/differential
thermogravimetry/differential thermogravimetry
thermogravimetry
Thermoanalytical methods, especially thermogravimetry/differential thermogravimetry
(TG/dTG), are already being utilized to measure the physical and chemical properties ofofcoffee
(TG/dTG), are already being utilized to measure the physical and chemical properties coffee
(TG/dTG), are already
samplesasasa afunction being utilized
functionofoftemperature to
temperatureorortime.measure
time.For the
Forthe physical
thecoffee and
coffeesamples chemical
samplesanalyzed properties
analyzedininthis of
thisstudy,coffee
study,the
the
samples
samples
thermal as a function of temperature
decomposition occurred in or time.
four For between
stages the coffee90 samples
and 390analyzed
°C in this
(Figure 5). study,
A the thermal
significant mass
thermal decomposition occurred in four stages between 90◦and 390 °C (Figure 5). A significant mass
decomposition
loss(68%)
(68%)can
canbeoccurred
beobservedin four
observed afterstages between
thesecond
second 90 and
stage 390°C.CAt
(Figure 5). A significant mass loss (68%)
loss after the stage
◦ atat190
190 temperatures
°C. At temperatures above420
above
◦
420 °Conly
°C only ashes
ashes
can be observed
remain in the after
final the second stage at 190 C. At temperatures above 420 C only ashes remain in
matter.
remain in the final matter.
the final matter.

Figure5.5.5.Thermostability
Figure
Figure Thermostabilityof
Thermostability offermented
of fermentedcoffee
fermented coffeecherries
coffee cherriessubmitted
cherries submittedto
submitted toan
to anoxygen
an oxygenatmosphere
oxygen atmosphereheated
atmosphere heatedup
heated up
up
to 800
to800 ◦°C.
800°C.C.dTG:dTG: profiles
dTG:profiles
profilesofof main
ofmain volatile
mainvolatile products;
volatileproducts;
products;TG:TG:
TG:massmass loss
massloss rate
lossrate curve.
rate curve.
curve.
to

Thefirst
The firsttwo
twothermal
thermaldegradation
degradationevents
eventsthat
thatoccurs
occursatatthe
thetemperatures
temperaturesofof9090andand138
138°C°Care
are
associated with loss of free water and volatiles compounds (i.e., alcohols, aldehydes
associated with loss of free water and volatiles compounds (i.e., alcohols, aldehydes and organic and organic
acids)and
acids) andabsorbed
absorbedwater,
water,respectively.
respectively.AAsmall
smallvariation
variation(9%)
(9%)ininthethetotal
totalweight
weightloss
losswithin
withinthis
this
Fermentation 2017, 3, 11 9 of 11

The first two thermal degradation events that occurs at the temperatures of 90 and 138 ◦ C are
associated with loss of free water and volatiles compounds (i.e., alcohols, aldehydes and organic
acids) and absorbed water, respectively. A small variation (9%) in the total weight loss within this
temperature range corresponds to the water content in fermented beans [33]. A study performed by
Yeretzian et al. [34] monitored the emission of volatile components during the roasting step of the
fermented beans which observed that during the endothermic phase the loss of water and volatile
compounds that are not derived from Maillard’s reaction or non-volatile precursors is prominent.
The latter thermal degradation events represent a significant loss in mass (68%) of the fermented beans
correspond to depolymerisation of hemicelluloses or pectin and cellulose decomposition, which occurs
at 240–315 ◦ C and 370–400 ◦ C, respectively [35].

4. Conclusions
The results of the present study indicated that Saccharomyces sp. is a dominant, well-adapted
yeast found in coffee fermentation at Brazilian Cerrado Mineiro region. In addition, this study is the
first to report the presence of aroma-producing yeast Hanseniaspora vineae in coffee bean fermentation.
Physicochemical analyses showed that different organic compounds present in coffee bean samples
may be derived from microbial metabolism during the fermentation process. Future studies should
focus on the dynamic of diffusion of these compounds into the beans, and to determine the actual role
of the microbial fermentation for coffee beverage quality. Our findings are relevant as a support for the
development of usual starter cultures and controlled batch processes.

Acknowledgments: The authors thank Conselho Nacional de Desenvolvimento Cientíıfico e Tecnológico do

Brasil (CNPq) for the research scholarship.
Author Contributions: Carlos Ricardo Soccol idealized wrote the research project and obtained research
scholarship; Dão Pedro de Carvalho Neto and Gilberto Vinícius de Melo Pereira designed the experiments;
Dão Pedro de Carvalho Neto and Vanete Thomaz Soccol conducted the yeast isolation and molecular
sequencing experiments; Cristine Rodrigues, Valcineide O. A. Tanobe and Bruno José G. da Silva conducted the
physicochemical characterization experiments; Dão Pedro de Carvalho Neto and Gilberto Vinícius de Melo Pereira
analyzed the results and wrote the paper.
Conflicts of Interest: The authors declare no conflict of interests.

References
1. Pereira, G.V.M.; Soccol, V.T.; Pandey, A.; Medeiros, A.B.P.; Lara, J.M.R.A.; Gollo, A.L.; Soccol, C.R. Isolation,
selection and evaluation of yeasts for use in fermentation of coffee beans by the wet process. Int. J.
Food Microbiol. 2014, 188, 60–66. [CrossRef] [PubMed]
2. Evangelista, S.R.; Silva, C.R.; Miguel, M.G.P.C.; Cordeiro, C.S.; Pinheiro, A.C.M.; Duarte, W.F.; Schwan, R.F.
Improvement of coffee beverage quality by using selected yeasts strains during the fermentation in dry
process. Food Res. Int. 2014, 61, 183–195. [CrossRef]
3. Evangelista, S.R.; Miguel, M.G.C.P.; Silva, C.P.; Pinheiro, A.C.M.; Schwan, R.F. Microbiological diversity
associated with the spontaneous wet method of coffee fermentation. Int. J. Food Microbiol. 2015, 210, 102–112.
[CrossRef]
4. Pereira, G.V.M.; Neto, E.; Soccol, V.T.; Medeiros, A.B.P.; Woiciechowski, A.L.; Soccol, C.R. Conducting starter
culture-controlled fermentations of coffee beans during on-farm wet processing: Growth, metabolic analyses
and sensorial effects. Food Res. Int. 2015, 75, 348–356. [CrossRef]
5. Pereira, G.V.M.; Carvalho-Neto, D.P.; Medeiros, A.B.P.; Soccol, V.T.; Neto, E.; Woiciechowski, A.L.; Soccol, C.R.
Potential of lactic acid bacteria to improve the fermentation of coffee during on-farm processing. Int. J. Food
Sci. Technol. 2016, 51, 1689–1695. [CrossRef]
6. Pereira, G.V.M.; Soccol, V.T.; Brar, S.K.; Neto, E.; Soccol, C.R. Microbial ecology and starter culture technology
in coffee processing. Crit. Rev. Food Sci. 2015. [CrossRef]
7. Wootton, A.E. The fermentation of coffee. Kenya Coffee 1963, 28, 239–249.
8. Sivetz, M. Coffee Processing Technology; AVI: Westport, CT, USA, 1963; Volume 2.
Fermentation 2017, 3, 11 10 of 11

9. Masoud, W.; Cesar, L.B.; Jespersen, L.; Jakobsen, M. Yeast involved in fermentation of Coffea arabica in East
Africa determined by genotyping and by direct denaturating gradient gel electrophoresis. Yeast 2004, 21,
549–556. [CrossRef]
10. Silva, C.F.; Batistam, L.R.; Abreu, L.M.; Dias, E.S.; Schwan, R.F. Succession of bacterial and fungal
communities during natural coffee (Coffea arabica) fermentation. Food Microbiol. 2008, 25, 951–957. [CrossRef]
[PubMed]
11. Vilela, D.M.; Pereira, G.V.M.; Silva, C.F.; Batista, L.R.; Schwan, R.F. Molecular ecology and polyphasic
characterization of the microbiota associated with semi-dry processed coffee (Coffea arabica L.). Food Microbiol.
2010, 27, 1128–1135. [CrossRef] [PubMed]
12. The Association of Brazilian Coffee Producing Regions. Available online: http://www.abopcafe.com.br/
cerrado-mineiro/ (accessed on 19 January 2017).
13. Avallone, S.; Guyot, B.; Brillouet, J.M.; Olguin, E.; Guiraud, J.P. Microbiological and biochemical study of
coffee fermentation. Curr. Microbiol. 2001, 42, 252–256. [CrossRef]
14. Velmourougane, K. Impact of natural fermentation on physicochemical, microbiological and cup quality
characteristics of Arabica and Robusta coffee. Proc. Natl. Acad. Sci. USA 2013, 83, 233–239. [CrossRef]
15. Lopez, C.I.; Bautista, E.; Moreno, E.; Dentan, E. Factors related to the formation of “overfermented coffee
beans” during the wet processing method and storage of coffee. In Proceedings of the 13th International
Scientific Colloquium on Coffee, Paipa, Colombia, 21–25 August 1989.
16. Silva, C.F.; Vilela, D.M.; de Souza, C.C.; Duarte, W.F.; Dias, D.R.; Schwan, R.F. Evaluation of a potential
starter culture for enhance quality of coffee fermentation. World J. Microbiol. Biotechnol. 2013, 29, 235–247.
[CrossRef]
17. Blomberg, A.; Alder, L. Physiology of osmotolerance in fungi. Adv. Microbiol. Phys. 1992, 33, 145–212.
18. Sauer, M.; Porro, D.; Mattanovich, D.; Branduardi, P. Microbial production of organic acids: Expanding the
markets. Trends Biotechnol. 2008, 26, 100–108. [CrossRef] [PubMed]
19. Ciani, M.; Beco, L.; Comitini, F. Fermentation behaviour and metabolic interactions of multistarter wine
yeast fermentations. Int. J. Food Microbiol. 2006, 108, 239–245. [CrossRef] [PubMed]
20. Viana, F.; Belloch, C.; Vallés, S.; Manzanares, P. Monitoring a mixed starter of Hanseniaspora
vineae-Saccharomyces cerevisiae in natural must: Impact on 2-phenylethyl acetate production. Int. J.
Food Microbiol. 2011, 151, 235–240. [CrossRef] [PubMed]
21. Medina, K.; Boido, E.; Fariña, L.; Gioia, O.; Gomez, M.E.; Barquet, M. Increased flavour diversity of
Chardonnay wines by spontaneous fermentation and co-fermentation with Hanseniaspora vineae. Food Chem.
2013, 141, 2513–2521. [CrossRef] [PubMed]
22. Lleixà, J.; Martín, V.; Portillo, M.C.; Carrau, F.; Beltran, G.; Mas, A. Comparison of fermentation and wines
produced by inoculation of Hanseniaspora vineae and Saccharomyces cerevisiae. Front. Microbiol. 2016, 7, 1–12.
[CrossRef] [PubMed]
23. Swiegers, J.H.; Pretorius, I.S. Yeast modulation of wine flavor. Adv. Appl. Microbiol. 2005, 57, 131–175.
[PubMed]
24. Damiani, P.; Gobbetti, M.; Cossignani, L.; Corsetti, A.; Simonetti, M.S.; Rossi, J. The sourdrough microflora.
Characterization of hetero- and homofermentative lactic acid bacteria, yeasts and their interactions on the
basis of the volatile compounds produced. LWT Food Sci. Technol. 1996, 29, 63–70. [CrossRef]
25. Owusu, M.; Petersen, M.A.; Heimdal, H. Effect of fermentation method, roasting and conching conditions
on the aroma volatiles of dark chocolate. J. Food Process. Preserv. 2012, 36, 446–456. [CrossRef]
26. Martín, M.J.; Pablosm, F.; González, A.G. Characterization of green coffee varieties according to their metal
content. Anal. Chim. Acta 1998, 358, 177–183.
27. Franková, A.; Drábekm, O.; Havlík, J.; Száková, J.; Vanek, A. The effect of beverage preparation method on
aluminum content in coffee infusions. J. Inorg. Biochem. 2009, 103, 1480–1485.
28. Sinija, V.; Mishra, H. FT-NIR spectroscopy for caffeine estimation in instant green tea powder and granules.
LWT Food Sci. Technol. 2009, 42, 998–1002. [CrossRef]
29. Paradkar, M.M.; Irudayaraj, J. A rapid FTIR spectroscopic method for estimation of caffeine in soft drinks
and total methylxanthines in tea and coffee. J. Food Sci. 2002, 67, 2507–2511. [CrossRef]
30. Giovenzana, V.; Beghi, R.; Guidetti, R. Rapid evaluation of craft beer quality during fermentation process by
Vis/NIR spectroscopy. J. Food Eng. 2014, 142, 80–86. [CrossRef]
Fermentation 2017, 3, 11 11 of 11

31. Krähmer, A.; Engel, A.; Kadow, D.; Ali, N.; Umaharan, P.; Kroh, L.W.; Schulz, H. Fast and
neat—Determination of biochemical quality parameters in cocoa using near infrared spectroscopy. Food Chem.
2015, 181, 152–159. [CrossRef] [PubMed]
32. Lyman, D.J.; Benck, R.; Dell, S.; Merle, S.; Murray-Wijelath, J. FTIR-ATR analysis of brewed coffee: Effect of
roasting conditions. J. Agric. Food Chem. 2003, 51, 3268–3272. [CrossRef]
33. Yeretzian, C.; Jordan, A.; Badoud, R.; Lindinger, W. From the green bean to cup of coffee: Investigating coffee
roasting by on-line monitoring of volatiles. Eur. Food Res. Technol. 2002, 214, 92–104. [CrossRef]
34. Jakab, E.; Faix, O.; Till, F. Thermal decomposition of milled wood lignins studied by thermogravimetry/mass
spectrometry. J. Anal. Appl. Pyrolysis 1997, 40–41, 171–186. [CrossRef]
35. Ouajai, S.; Shanks, R.A. Composition, structure and thermal degradation of hemp cellulose after chemical
treatments. Polym. Degrad. Stabil. 2005, 89, 237–335. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).