Simula ion and fla o compound analysis o dealcoholized bee ia
one-s ep acuum dis illa ion
C is ina And és-Iglesias
a
, Juan Ga cía-Se na
b
, Olimpio Mon e o
c
, Ca los A. Blanco
a,
⁎
a
Depa amen o de Ingenie ía Ag ícola y Fo es al (Á ea de Tecnología de los Alimen os), ETS Ingenie ías Ag a ias (Uni e sidad de Valladolid), A da. de Mad id 44, 34004 Palencia, Spain
b
High P essu e P ocesses G oup, Depa men o Chemical Enginee ing and En i onmen al Tech., Uni e si y o Valladolid, 47011 Valladolid, Spain
c
Cen e o Bio echnology De elopmen (CDB), CSIC. Boecillo Technological Pa k, A . F ancisco Vallés 8, 47151 Boecillo, Valladolid, Spain
abs ac a icle in o
A icle his o y:
Recei ed 27 May 2015
Recei ed in e ised o m 9 July 2015
Accep ed 12 July 2015
A ailable online xxxx
Keywo ds:
Alcohol- ee bee
Aspen HYSYS simula ion
Dealcoholiza ion
Vola ile compounds
Fla o pe cep ion
HS-SPME
The coupled ope a ion o acuum dis illa ion p ocess o p oduce alcohol ee bee a labo a o y scale and Aspen
HYSYS simula ion so wa e was s udied o define he chemical changes du ing he dealcoholiza ion p ocess in
he a oma p ofiles o 2 di e en lage bee s.
A he lab-scale p ocess, 2 di e en pa ame e s we e chosen o dealcoholize bee samples, 102 mba a 50 °C and
200 mba a 67 °C. Samples aken a di e en s eps o he p ocess we e analyzed by HS-SPME–GC–MS ocusing
on he concen a ion o 7 fla o compounds, 5 alcohols and 2 es e s. Fo simula ion p ocess, he EoS pa ame e s
o he Wilson-2 p ope y package we e adjus ed o he expe imen al da a and one mo e p essu e was es ed
(60 mba ).
Simula ion me hods ep esen a iable al e na i e o p edic esul s o he ola ile compound composi ion o a
final dealcoholized bee .
© 2015 Published by Else ie L d.
1. In oduc ion
The ma ke o non-alcoholic b ews has expe ienced a significan
imp o emen du ing he pas yea s ha becomes mo i a ed mainly
by highly compe i i e ma ke s, d i ing/d inking ules, heal h
condi ions incompa ible wi h alcohol consump ion and/o eligious
easons (And és-Iglesias, Mon e o, Sancho, & Blanco, 2014; Blanco,
And és-Iglesias, & Mon e o, 2014; Ca a ino & Mendes, 2011). Simila ly,
i is well-known ha bee has posi i e e ec s and a whole ange o
p ope ies, such as no a o choles e ol con en , ee suga con en ,
high an ioxidan , magnesium and soluble fibe con en (B ányik e al.,
2012), plus i p o ides essen ial i amins and mine als con ibu ing o
a heal hy balanced die (And és-Iglesias, Blanco, Blanco, & Mon e o,
2014; Bam o h, 2001).
Bee a oma p ofile is made by many ola ile o ganic compounds a
e y low concen a ion (ppm le el), which a e esponsible o i s
unique fla o (Ca a ino, Mendes, Madei a, & Fe ei a, 2007). Le els o
di e en chemical compounds, such as alcohols, es e s, aldehydes,
ke ones, o ganic acids and phenols, can be ound on bee composi ion,
gi ing a specificfla o ha con ibu es o he o e all o ganolep ic p op-
e ies o he final bee (Ka lsson & T ägå dh, 1997). Among hem, es e s
and alcohols a e he main g oups o a oma compounds. Es e s a e
esponsible o he swee and ui y fla o s o bee , while alcohols
con e i an alcoholic, ui y and imma u e fla o (And és-Iglesias,
Blanco e al., 2014; And és-Iglesias, Mon e o e al., 2014; Ca a ino,
Fe ei a, & Mendes, 2009).
In low-alcohol and/o alcohol- ee bee p oduc ion, he di e en
echniques used ha e o be able o each he maximum alcohol by
olume (ABV) es ablished by he di e en coun ies legal egula ions.
In he majo i y o EU coun ies bee s wi hlow alcohol con en a e di id-
ed in o alcohol- ee bee s (≤0.5% ABV) and low-alcohol bee s (≤1.2%
ABV). In Spain, alcohol ee bee s a e di ided in non-alcohol bee s
(≤1.0% ABV) and ‘0.0%’bee s (≤0.1% ABV). Howe e , in he Uni ed
S a es he e should no be alcohol p esen in alcohol- ee bee s, while
0.5% ABV co esponds o he uppe limi o non-alcoholic bee s o
‘nea -bee s’(Olmo, Blanco, Palacio, P ádanos, & He nández, 2014).
A p esen , he e a e se e al me hods o low alcohol bee p oduc-
ion (Blanco e al., 2014). The s a egies can be di ided in o wo main
g oups: biological and physical me hods (B ányik e al., 2012;
Mon ana i, Ma coni, Maye , & Fan ozzi, 2009; Olmo e al., 2014).
While physical me hods wi hd aw he e hanol om a e men ed bee ,
biological me hods aim a con olling he alcohol p oduc ion du ing
he e men a ion p ocess (Zü che , Jakob, & Back, 2005).
Biological me hods can be achie ed by ei he es ic ing e hanol
o ma ion o sho ening he e men a ion p ocess. Ob aining low alco-
hol con en ia in e up ed e men a ion is accompanied by low con-
en s o a oma and fla o compounds, and hei p oduc s a e o en
Food Resea ch In e na ional xxx (2015) xxx–xxx
⁎Co esponding au ho .
E-mail add ess: cblanco@ia .u a.es (C.A. Blanco).
FRIN-05933; No o Pages 10
h p://dx.doi.o g/10.1016/j. ood es.2015.07.017
0963-9969/© 2015 Published by Else ie L d.
Con en s lis s a ailable a ScienceDi ec
Food Resea ch In e na ional
jou nal homepage: www.else ie .com/loca e/ ood es
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017
cha ac e ized by wo y o -fla o s. They a e usually pe o med using
adi ional b ewe y equipmen and hence do no equi e addi ional
in es men s (B ányik e al., 2012; Ca a ino & Mendes, 2011).
O he p ocesses o a oid hese limi a ions include he use o special
o immobilized yeas s as well as he use o low suga aw ma e ials
(Ca a ino & Mendes, 2011; Picke ing, 2000). The use o special yeas s
o a low alcohol bee p oduc ion p ocess inc eases he cos s wi h he
need o yeas selec ion, o gene ic modifica ion o he p oduc ion o gan-
isms. Howe e , sui able selec ed yeas s can con ibu e significan ly o
he p oduc senso ial quali y imp o emen . Alcohol ee bee p oduc-
ion p ocesses by con inuous e men a ion wi h immobilized yeas ,
whose objec i eis limi ed alcohol o ma ion,andwhich equi es special
equipmen and ma e ial. In his la e case, high in es men cos s a e e-
qui ed bu a e jus ified by a highe p oduc i i y o he con inuous p o-
cesses. In gene al, p oducing alcohol- ee bee by biological me hods
makes impossible hep oduc ion o alcohol- eebee s wi h alcoholcon-
en close o ze o (B ányik e al., 2012).
Physical me hods equi e conside able in es men s in o he special
equipmen o alcohol emo al (B ányik e al., 2012). The mos common
sepa a ion p ocesses used o bee dealcoholiza ion a e memb ane-
based p ocesses and hea ea men (Ca a ino e al., 2007).
Memb ane-based p ocesses include e e se osmosis, nanofil a ion, di-
alysis and pe apo a ion (Labanda, Vichi, Llo ens, & López-Tamames,
2009). Hea ea men p ocesses comp ise e apo a ion and dis illa ion;
hey a e bo h unde acuum condi ions o p ese e he o ganolep ic
p ope ies by a oiding undesi ed seconda y eac ions (Belisa io-
Sánchez, Taboada-Rod iguez, Ma in-Inies a, & López-Gómez, 2009).
Fu he mo e, he mal p ocesses o emo e alcohol om egula bee s
can cause he loss o he o iginal a oma (Blanco e al., 2014; Ca a ino
e al., 2009) bu hei ad an age is ha hey can emo e e hanol om
bee s o le els close o ze o (B ányik e al., 2012).
Among hese physical me hods, o la ge scale dealcoholiza ion, he
acuum e apo a ion is he mos economic p ocess (Zü che e al.,
2005). Dis illa ion is a sepa a ion ope a ion based on di e ences in ol-
a ili y. I a mix u e con aining subs ances ha di e in hei ola ili y is
b ough o ebulli ion, he composi ion o he apo s eleased will be di -
e en om ha o he boiling liquid. A e condensa ion, he apo s
cons i u e he “dis illa e”. The emaining liquid is called “ esidue”
(Be k, 2013). The applica ion o acuum o dis illa ion p ocess enables
o educe he e apo a ion empe a u e and hus he he mal s ess o
bee (Zü che e al., 2005). I he p essu e is educed, alcohol can be
d awn o a much lowe empe a u e (B ányik e al., 2012). The mal
p ocesses o p oduce alcohol ee bee s a e pe o med a empe a u es
be ween 30 and 60 °C a p essu es o 60 o 200 mba (Soh ab andi,
Mousa i, Raza i, Mo aza ian, & Rezaei, 2010; Zü che e al., 2005).
The de e io a ion o bee quali y by he mal dealcoholiza ion depends
mainly on he e apo a ion empe a u e and he pe iod o exposu e
(B ányik e al., 2012).
I is well known ha mos o he a oma compounds a e los in
alcohol ee bee s du ing p oduc ion by he mal p ocesses. The a oma
p ofile is clea ly damaged and o he , less pleasan fla o s, like b eady,
wo y o ca amel no es can appea (Blanco e al., 2014; Ca a ino e al.,
2009; Lehne e al., 2009; Soh ab andi e al., 2010). To compensa e
hese disad an ages many b ewe ies use a modified b ewing echnolo-
gy o he p oduc ion o a mo e a oma ic o iginal bee . Ano he a emp
o compensa e senso y disad an ages is by blendingdealcoholized bee
wi h a small quan i y o o iginal bee o a bee a oma ex ac ha can be
eco e ed in e apo a ion plan s wi h ec ifica ion columns. Since hese
a emp s a e no ye sa is ac o y u he possibili ies o imp o e he
quali y o hese bee s ha e been in es iga ed (Zü che e al., 2005).
Owing o bee chemical compound cha ac e iza ion has imp o ed
no ably nowadays, analysis o bee fla o compounds has been
cons an ly op imized o ob ain be e esul s in ela ion o sensi i i y
and specifici y (And és-Iglesias, Blanco e al., 2014; And és-Iglesias,
Mon e o e al., 2014). Gas ch oma og aphy–mass spec ome y
(GC–MS) is cu en ly used o measu e ola ile compound concen a ions
in bee . E he s, es e s, acids, aldehydes, ke ones, alcohols, sul u com-
pounds, hyd oca bon compounds, alicyclic compounds, he e ocyclic
compounds and a oma ic compounds can be measu ed simul aneously
by using GC–MS me hods (And és-Iglesias, Blanco e al., 2014;
And és-Iglesias, Mon e o e al., 2014). The combina ion o solid phase
mic oex ac ion (SPME) wi h gas ch oma og aphy (GC) o gas
ch oma og aphy–mass spec ome y (GC–MS) has p o en o be a sensi-
i e and p ecise me hod o he analysis o di e en classes o ola ile
compounds (Dong e al., 2013).
Bee dealcoholiza ion ia acuum dis illa ion in a ba ch sys em can
be assumed o esemble di e en ial dis illa ion a educed p essu e.
The p inciples o di e en ial dis illa ion a e well es ablished since he
beginning o chemical enginee ing knowledge. Thus, his ype o dis il-
la ion is o en known as “Rayleigh dis illa ion”. Lo d Rayleigh's law is
based on a dynamic ma e ial balance o he ola ile compound o a
wo componen mix u e coupled o he global mass balance (Be k,
2013). Ex ending hebalance o a mul icomponen mix u e wass udied
in deep by se e al au ho s such as Lang e al. (1994) and, Ya im e al.
(1993) who modified he p ocess o he addi ion o an ex ac i e
agen , o including sie es. An in e es ing compa a i e s udy was
conduc ed by Zü che e al. (2005) using lab scale ba ch and con inuous
dis illa ion as well as an indus ial scale plan . They in es iga ed he
bee dealcoholiza ion a 60 and 150 mba by ollowing changes in a
numbe o compounds, e.g., e hanol, 1-p opanol, e hyl ace a e, 2-
me hylp opanol, 3-me hylp opanol and se e al es e s. Howe e , hey
did no simula e he p ocess.
In addi ion, se e al au ho s ha e in es iga ed he simula ion o
spi i s p oduc ion by his p ocess. Claus and Be glund s udied ui
b andy dis illa ion using a ba ch column dis illa ion. They simula ed
he p ocess using CHEMCAD wi h good esul s using NRTL (Non-
Random Two Liquids) equa ion o s a e (EoS) oge he wi h
UNIFAC pa ame e s (Claus & Be glund, 2005, 2009). On he o he
hand, Gaise e al. simula ed he whisky s ill dis illa ion p ocess
using Aspen Plus so wa e and selec ing he NRTL-2 p ope y pack-
age o ha so wa e; hey claimed ha his EoS p o ides a good ap-
p oxima ion o e hanol–wa e azeo ope (Gaise e al., 2002).
Low alcohol and alcohol ee bee consump ion is inc easing yea by
yea , and o en, hese ypes o be e ages a e known o ha e a poo fla-
o p ofile in compa ison o he o iginal bee . In his sense, i becomes
impo an o adjus he fla o o non-alcoholic bee s o ha o egula
ones, which needs unde s anding o how he dealcoholiza ion p ocess
modifies i , an issue o which he scien ific in o is sca ce.
In his wo k, we ha e combined lab scale di e en ial acuum dis il-
la ion,a oma compoundanalysis and simula ion o he dealcoholiza ion
p ocess o shed ligh on his p ocess. The main objec i e is o es a
simula ion en i onmen ha can explain he lab esul s, so ha , i can
be ex apola ed o a simila p ocess a indus ial scale. Fo his, we
ha e selec ed wo model bee s, one om Spain and one om
Ge many, and adjus ed he in e ac ion pa ame e s o a he modynamic
model. To ou knowledge, his is he fi s ime ha i is done o bee s.
2. Ma e ial and me hods
2.1. Samples and acuum dis illa ion dealcoholiza ion expe imen s
Two di e en big-scale lage bee b ands we e chosen o he s udy,
one om Spain (S) and ano he one om Ge many (G). Bo h o hem
we e lage alcoholic bee s con aining 5.5 and 4.8% alcohol by olume
(ABV) espec i ely, and we e ob ained as esh as possible om he
local ma ke . Bee bo les we e s o ed a 4 °C un il dealcoholiza ion
p ocess. 400 mL o bee we e weigh ed and placed in 1 L flask o he
acuum dis illa ion sys em o each expe imen ; he flask was co e ed
wi h a black plas ic ma e ial o a oid he ligh oxida ion o he sample
componen s. Subsequen ly, 10 μL o an i oam emulsion (E-900, AFCA)
we e added o educe he oam and CO
2
con en .
2C. And és-Iglesias e al. / Food Resea ch In e na ional xxx (2015) xxx–xxx
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017
The expe imen s o bee dealcoholiza ion by labo a o y scale acu-
um dis illa ion we e done a wo di e en acuum p essu es and
wa e ba h empe a u es. The empe a u e needed in he wa e ba h
is di ec ly ela ed o he o al p essu e by he phase equilib ium o he
sys em, bu has o be se o a sligh ly highe alue o assu e enough
hea ans e . Thus, he fi s se o expe imen s was conduc ed a
102 mba and 50 °C (co esponding o a sa u a ion empe a u e o
pu e wa e , 46.2 °C) and he second se a 200 mba and 67 °C (co e-
sponding o a sa u a ion empe a u e o pu e wa e , 60.1 °C), A
Ro a apo R-215 wi h acuum pump V-700, acuum con olle V-850
and diagonal condense (BÜCHI Labo echnik AG, Swi ze land) was
used. The flask o a ion was fixed a 20 pm and emained cons an in
all expe imen s. Each dealcoholiza ion p ocess was s opped a 15, 30,
45 and 60 min a e he onse o sampling in o de o analyze he
di e en ola ile compounds e apo a ed along wi h he e hanol. A
he end o he dis illa ion p ocess, he esidual bee was cooled in
glass bo les and weigh ed o he ma e ial balance calcula ion.
Fo all expe imen s he same s eps we e done. A he beginning o
each expe imen he wa e ba ch was efilledun il hesame olumei
necessa y, once he ba ch eached he empe a u e, he expe imen
s a ed a he pm indica ed abo e, he p essu e was eached immedi-
a ely and emained cons an (±1) o e he whole expe imen and
con olled by he acuum con olle .
Fo he GC–MS analysis 15 mL da k ials sealed wi h PTFE–silicone
sep a (Supelco, USA) we e used o sample p epa a ion. Vials con ained
2 g o NaCl (Scha lau, Scha lab S.L., Spain) and 5 mL o bee we e added
and s i ed o sol e he NaCl and homogenize he sample. A o al o 60
samples we e aken and analyzed om he o iginal bee s, and om e-
sidual bee s a each ime and dealcoholiza ion p ocess expe imen s.
2.2. Gas ch oma og aphy–mass spec ome y (GC–MS) equipmen
Vola ile compounds we e sepa a ed and de ec ed by gas ch oma og-
aphy using an Agilen GC 6890N (Agilen Technologies, USA) equipped
wi h an Agilen 5973 single quad upole mass spec ome e (, Agilen
Technologies, USA) as de ec o . A headspace solid phase mic oex ac ion
(HS-SPME) manual equipmen (Supelco, USA) was used o he
ex ac ion and concen a ion o he ola ile compounds, which was
ca ied ou wi h 100 μm polydime hylsiloxan (PDMS) fibe (Sulpeco,
USA). P io o use, he SPME fibe was condi ioned a 250 °C o 30 min
in he GC injec o , acco ding o he manu ac u e 's ins uc ions. Blank
uns o he fibe we e comple ed, be o e sampling, each day o ensu e
no ca y-o e o analy es. The ch oma og aphic sepa a ions we e
accomplished using a BP-1 30 m × 0.32 mm × 1 μm capilla y column
(SGE Analy ical Science, Aus alia).
2.3. Analysis o ola ile compounds
The ola ile composi ion o bee samples wasmeasu ed by iplica e.
Solid phase mic oex ac ion o compounds was pe o med a 30 °C o
45 min. The deso p ion was achie ed in he injec o o he GC
ch oma og aph in spli less mode o 15 min, and he empe a u e was
se a 250 °C as indica ed by he manu ac u e o PDMS fibe . Ca ie
gas was helium a a cons an flow o 1.2 mL/min.
The o en empe a u e was p og ammed as ollows: ini ial empe a-
u e was se a 35 °C and kep o 7 min, his was ollowed by 2 amps in
which empe a u e was isen a 8 °C/min o 200 °C and kep his
empe a u e o 5 min, and hen empe a u e was isen a 10 °C/min
o 250 °C, his empe a u e being kep o 10 min.
The ioniza ion ene gy was 70 eV, and de ec ion and da a acquisi ion
we e pe o med in scan mode om 37 o 350 Da. Fo iden ifica ion
da a ob ained in he GC–MS analysis we e compa ed wi h m/z
alues compiled in he spec um lib a y WILEY. Valida ion o com-
pound iden ifica ion was ca ied ou by compa ison o MS spec a
and e en ion imes wi h hose o comme cials anda ds. Quan ifica ion
was ca ied ou by using s anda d calib a ion cu es o 2-me hylbu anol
(≥99.0%), 3-me hylbu anol (≥99.0%), 2-phenyle hanol (≥99.0%),
e hyl ace a e (≥99.5%), isobu anol (≥99.0%) ( hese om Sigma,
USA). 1-P opanol ≥99.5% (Fluka, Sigma-Ald ich, USA) and isoamyl
ace a e ≥99.0%(Fishe ,UK).Because1-p opanolco-elu edwi h
e hanol, he ex ac ed ion ch oma og am (EIC) o he ion wi h m/z
60.05 and e en ion ime o 3.10 min was used o quan ifica ion o
his compound.
2.4. HYSYS simula ion and pa ame e s
In o de o simula e he sys em unde s udy o he ba ch dis illa ion
o bee he ollowing assump ions we e conside ed:
1) The acuum is done almos ins an ly and a = 0 hesys em is a he
cons an desi ed acuum p essu e.
2) Liquid composi ion is homogeneous and hea is uni o mly dis ibu ed.
3) The flask has been simula ed by a cylinde o simpli y le el calcula ion.
4) The hea flux o each da a poin is de e mined o ma ch he ime e-
qui ed o a ce ain apo iza ion olume. This is because he Ro a apo
sys em can p o ide di e en hea flux depending on a numbe o
a iables (wa e le el, flask loca ion, ambien empe a u e, o a ion
speed, e c.).
5) No eac ion occu s in he bulk liquid.
The simula ions ha e been ca ied ou using HYSYS simula ion
so wa e (Aspen inc. p oduc ) as i has a powe ul non-s eady s a e sim-
ula ion ool.
Wilson-2 p ope y package waschosenin o de o simula e he non-
ideal beha io o he liquid phase, while ideal gas is conside ed o he
gas phase (as i was unde educed p essu e condi ions).
The main simula ion p ocess flow diag am is depic ed in Fig. 1.The
main dis illa ion essel (V-100) has one eed s eam-5 ( i ual o sim-
ula ion pu poses se a almos ze o flow), one hea sou ce (Q-100), one
liquid ou le s eam-2 ( i ual o simula ion pu poses se a almos
ze o flow) and one apo ou le s eam-3 (main dis illa ion ou le ).
Themain calcula ions we e ca ied ou using anExcel sp eadshee o
de e mine he con e sion be ween ppm and mola ac ion alues om
expe imen al condi ions o he simula ion and ice e sa.
The main componen s simula ed we e: suc ose, e hanol, e hyl
ace a e, 1-p opanol, isobu anol, isoamyl ace a e, 2-me hylbu anol, 3-
me hylbu anol, 2-phenyle hanol, wa e and ni ogen.
Suc ose was used as a simula ion ick o inc emen he densi y o
wa e a ge ing he eal alue o 1010 kg/m
3
, o ha pu pose a concen a-
ion o 3% w . was used in all simula ion expe imen s. Ni ogen was used
o simula ion pu poses mimicking he a mosphe e o he Ro a apo .
Ini ial alues o composi ions o he liquid we e inse ed in he
“hold-up” alues o he dis illa ion essel. The o al p essu e o
s eam-2 was fixed o he expe imen al absolu e p essu e, coinciding
wi h he essel ini ial p essu e (i.e., 102 and 200 mba ).
Fig. 1. HYSYS simula ion model o a di e en ial acuum dis illa ion.
3C. And és-Iglesias e al. / Food Resea ch In e na ional xxx (2015) xxx–xxx
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017
As indica ed in he assump ions, he hea flux was es ima ed o
ma ch he mass e apo a ed a each ime sample poin . This way, he
simula ion ime is no as impo an as he e apo a ed mass, ha is
used as he x-axis a iable as pe cen age o mass e apo a ed
(% apo ). Thus, all expe imen s we e ca ied ou un il 15, 30, 45 and
60 min, ime when he dealcoholiza ion p ocess was s opped and he
samples we e collec ed. The % o apo ac ion (% V ) was calcula ed
as he pe cen ageo ini ialmass o he bee minus hemassa he di e -
en poin s o he simula ion un il he las mass (a 60 min o simula ion)
di ided by he ini ial mass. Al hough he adi ional ASTM D-86 cu es
o pe oleum dis illa ion a e ca ied ou in olume, in his case, mass
was p e e ed o o e come densi y a ia ions (ASTM-In e na ional,
2012). Fu he mo e, he hea flux could ha e a ied along wi h he
expe imen . Fo his eason, we ha e conside ed his a iable mo e ac-
cu a e han expe imen al ime i sel . In addi ion o his, esul s could be
ans e ed o a eal acuum dis illa ion p ocess wi h be e scale-up
chances.
The de eloped so wa e is a ailable ee in he web page o he e-
sea ch g oup o High P essu e P ocesses o he Uni e si y o Valladolid
(h p://hpp.u a.es/so wa e/) in he sec ion o ‘Bee Dis illa ion’.
3. Resul s and discussion
Two lage bee s we e in es iga ed in his s udy, one sample om
Spain (S) and he o he sample om Ge many (G). Bo h samples we e
dealcoholized by acuum dis illa ion a labo a o y scale a 2 di e en
Fig. 2. Sample o TIC ch oma og am o S bee sample, alcohol bee on he op and bee dealcoholized by labo a o y acuum dis illa ion on he bo om. (1) 1-p opanol, (2) e hyl ace a e,
(3) isobu anol, (4) 3-me hylbu anol, (5) 2-me hylbu anol, (6) isopen yl ace a e, (7) 2-phenyle hanol.
Fig. 3. E hanol beha io agains he % apo ac ion on he le o S sample and o G sample on he igh .
4C. And és-Iglesias e al. / Food Resea ch In e na ional xxx (2015) xxx–xxx
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017
p essu es and empe a u es, 102 mba , 50 °C and 200 mba , 67 °C. A
o al o 45 compounds we e iden ified, and 7 o hem quan ified by
peak a ea. The p ofile o quan ified ola iles consis ed o 5 alcohols (1-
p opanol, 2-me hylp opanol, 2-me hylbu anol, 3-me hylbu anol and
2-phenyle hanol) and 2 es e s (e hyl ace a e and isoamyl ace a e).
These compounds a e conside ed as ones o he main alcohols and es-
e s in bee (Blanco e al., 2014; Lehne e al., 2009; Rod igues,
Caldei a, & Câma a, 2008;Willae & Nedo ic, 2006), he mos abundan
being e hyl ace a e, isoamyl ace a e, he amyl alcohols and isobu anol
(Piddocke, K eisz, Held -Hansen, Fog Nielsen, & Olsson, 2009). Also,
hese compounds a e hose cu en ly analyzed by o he au ho s be-
cause hey a e ele an fla o compounds in bee (Cha y-Pa a,
DeJesús-Eche a ia, & Pe ez, 2011; Kobayashi, Shimizu, & Shioya,
2008) and chosen as e e ence compounds when s udying indus ial
p ocesses o p oduc ion o non-alcoholic bee s (Ca a ino & Mendes,
2011; Mo a e al., 2011).
A ypical o al ion ch oma og am (TIC) o a egula bee sample and
i s dealcoholized bee by labo a o y scale acuum dis illa ion p ocess is
shown in Fig. 2.
3.1. Final e hanol con en calcula ed by ASPEN HYSYS simula ion
Du ing he di e en ial dis illa ion p ocess, he mos ola ile ac ion
(e hanolic ac ion) abandons he sys em in fi s place oge he wi h an
inc easing amoun o wa e . In his wo k, we ha e ocused on he anal-
ysis o he bee , a he han he e apo a ed ola ile ac ion (e hanolic
ac ion).
Table 1
Pe cen ageo he apo ac ions(%V )o SandG samplesandi sco esponden imes, o
bo h lab-scale acuum dis illa ion p ocesses and he a e ages (%).
Time, min 0 15 30 45 60
S 102 mba 0 7.46 9.55 13.40 15.76
S 200 mba 0 6.17 10.14 15.12 19.22
G 102 mba 0 5.70 9.00 14.40 17.60
G 200 mba 0 10.80 13.40 14.80 18.90
A e age (% V ) 0 7.53 10.52 14.43 17.87
Fig. 4. E hanol concen a ion i he final olume is dilu ed o non-dilu ed a he wo
expe imen al p essu es.
Table 2
Re en ion ime (R ), concen a ion o ola ile compounds (mg/L) delacoholized a 102 mba , 50 °C and 200 mba , 67 °C in he expe imen (EXP), in simula ions (SIM) and he s anda d
de ia ion o he expe imen al alue (S De ).
SIM EXP S De SIM EXP S De SIM EXP S De SIM EXP S De SIM EXP S De
102 mba , 50 °C R S, 0 min S, 15 min S, 30 min S, 45 min S, 60 min
1-P opanol 3.10 5.56 5.56 2.47 1.22 2.09 0.36 0.50 0.25 0.12 0.05 0.24 0.00 0.01 0.00 0.00
E hyl ace a e 4.13 17.82 17.82 1.11 1.52 1.07 0.10 0.67 1.23 0.41 0.15 0.86 0.30 0.07 0.54 0.09
Isobu anol 4.60 9.41 9.41 2.52 5.67 4.97 2.03 4.38 4.68 2.73 2.35 3.67 2.56 1.63 3.26 2.10
3-Me hylbu anol 9.11 40.99 40.99 2.73 25.01 17.44 1.45 19.37 17.15 7.14 10.43 11.51 1.54 7.24 7.33 1.05
2-Me hylbu anol 9.27 13.37 13.37 0.94 7.78 5.26 0.54 5.83 5.29 2.14 2.91 3.55 0.63 1.91 1.93 0.27
Isopen yl ace a e 14.09 1.92 1.92 0.33 0.36 0.11 0.08 0.15 0.10 0.09 0.02 0.09 0.08 0.00 0.09 0.08
2-Phenyl e hanol 19.82 34.01 34.01 1.39 37.07 40.46 6.26 37.95 48.76 17.74 39.51 62.04 9.92 40.16 85.28 3.90
102 mba , 50 °C R G, 0 min G, 15 min G, 30 min G, 45 min G, 60 min
1-P opanol 3.10 8.93 8.93 3.48 2.36 6.99 2.20 0.52 0.28 0.01 0.02 0.17 0.03 0.00 0.00 0.00
E hyl ace a e 4.13 26.54 26.54 0.74 3.85 3.65 1.08 1.08 2.93 0.47 0.20 1.06 0.17 0.07 0.45 0.06
Isobu anol 4.60 10.47 10.47 2.66 6.89 6.88 2.08 4.48 4.32 2.21 2.07 2.88 1.93 1.12 2.87 1.93
3-Me hylbu anol 9.11 43.77 43.77 2.12 28.46 27.85 2.79 18.58 13.69 2.02 8.58 4.92 0.82 4.65 5.13 0.88
2-Me hylbu anol 9.27 11.54 11.54 0.42 7.27 7.48 1.18 4.50 3.49 0.70 1.86 1.12 0.43 0.90 1.22 0.25
Isopen yl ace a e 14.09 2.58 2.58 0.25 0.67 0.27 0.15 0.15 0.16 0.10 0.01 0.13 0.09 0.00 0.11 0.08
2-Phenyl e hanol 19.82 37.69 37.69 4.68 40.00 53.95 5.40 41.46 56.92 5.71 43.17 69.57 6.81 43.71 75.17 4.58
200 mba , 67 °C R S, 0 min S, 15 min S, 30 min S, 45 min S, 60 min
1-P opanol 3.10 5.56 5.56 0.69 3.63 3.58 0.90 1.51 2.51 0.94 0.30 0.00 0.00 0.02 0.00 0.00
E hyl ace a e 4.13 17.82 17.82 1.79 2.99 4.09 1.00 0.79 1.19 0.07 0.22 0.86 0.14 0.07 0.46 0.11
Isobu anol 4.60 9.41 9.41 2.87 5.60 4.88 1.83 3.06 4.25 1.65 1.40 3.67 1.71 0.53 3.03 1.47
3-Me hylbu anol 9.11 40.99 40.99 4.05 26.58 20.44 4.01 15.70 16.90 0.83 7.88 11.51 2.29 3.34 7.35 0.70
2-Me hylbu anol 9.27 13.37 13.37 1.57 8.40 6.44 1.18 4.70 5.18 0.67 2.17 3.55 0.75 0.80 2.20 0.29
Isopen yl ace a e 14.09 1.92 1.92 0.10 0.63 0.20 0.07 0.17 0.12 0.04 0.02 0.09 0.05 0.00 0.07 0.04
2-Phenyl e hanol 19.82 34.01 34.01 2.59 36.63 43.36 7.14 38.28 53.41 3.85 39.66 62.04 6.35 40.17 70.65 3.61
200 mba , 67 °C R G, 0 min G, 15 min G, 30 min G, 45 min G, 60 min
1-P opanol 3.10 8.93 8.93 3.32 1.11 0.40 0.11 0.26 0.30 0.02 0.22 0.14 0.02 0.01 0.00 0.00
E hyl ace a e 4.13 26.54 26.54 3.91 0.90 5.18 0.48 0.35 1.24 0.28 0.34 2.91 0.53 0.10 2.73 0.25
Isobu anol 4.60 10.47 10.47 3.29 2.43 5.07 1.76 1.22 3.46 1.56 1.24 1.93 1.53 0.36 1.82 1.34
3-Me hylbu anol 9.11 43.77 43.77 6.60 12.11 19.95 0.57 6.50 9.48 1.31 6.64 5.04 0.88 2.18 2.94 0.76
2-Me hylbu anol 9.27 11.54 11.54 1.17 2.85 4.85 0.50 1.43 2.29 0.28 1.45 1.17 0.34 0.40 0.64 0.22
Isopen yl ace a e 14.09 2.58 2.58 0.37 0.11 0.49 0.05 0.02 0.16 0.06 0.02 0.09 0.05 0.00 0.08 0.07
2-Phenyl e hanol 19.82 37.69 37.69 5.40 42.27 54.97 2.93 43.47 62.58 10.36 43.27 59.41 8.68 43.98 59.97 1.01
5C. And és-Iglesias e al. / Food Resea ch In e na ional xxx (2015) xxx–xxx
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017

Ne e heless, he concen a ion o e hanol in he e hanolic ac ion
in alcohol by olume pe cen age (% ABV) was es ima ed by simula ion
a he wo expe imen al p essu es, 102 mba and 200 mba and an ad-
di ional educed p essu e o 60 mba .
The ini ial poin (IP) was he labeled alcohol con en o each bee
4.7% o G and 5.5% o S. The concen a ion o e hanol in he bee
phase exhibi ed an exponen ial-like decay agains he apo ac ion
(Fig. 3). The % o apo ac ions a hei co esponden imes in he ex-
pe imen is shown in Table 1.
In gene al, 1.0% ABV was ob ained a abou 15% o liquid apo iza-
ion. In his s udy we ha e analyzed and simula ed he composi ions
conside ing he ins an olume du ing he p ocess. So, we ha e no
co ec ed he alues conside ing a possible final dilu ion wi h wa e o
he ini ial olume. This means ha i he final esidue (dealcoholized
bee ) would be dilu ed o he ini ial olume (e.g., adding wa e ), he %
ABV achie ed would be lowe han 1% o e hanol ( ha was ob ained
a 200 mba o ins ance). This ac is illus a ed in Fig. 4, whe e we com-
pa e he % ABV o dilu ed dis illa e and o no dilu ed dis illa e.
3.2. Di e ences o he ola ile compound p ofile du ing he labo a o y scale
acuum dis illa ion p ocess
The ola ile compound ac ion in bee , apa om e hanol, is mainly
comp ised o highe alcohols o med du ing p ima y bee e men a ion
(Blanco e al., 2014). Highe alcohols con ibu e o he a oma o bee
and p oduce a wa m mou h eel (Willae & Nedo ic, 2006). The mos
significan con ibu ion is owed o p opanol, isobu anol and isoamyl
alcohols (2 and 3-me hylbu anol) (Blanco e al., 2014; B ányik,
Vicen e, Dos álek, & Teixei a, 2008). Highe alcohols a e he immedia e
p ecu so s o mos fla o ac i e es e s; hence, o ma ion o highe
alcohols needs o be con olled o ensu e op imal es e p oduc ion
(Gonçal es e al., 2014) because o es e s ha e e y low fla o
h esholds and a majo impac on he o e all fla o (Willae &
Nedo ic, 2006). Howe e , low con en o a oma compounds in
alcohol ee bee s could be a ibu ed o he dealcoholiza ion p ocess
(Riu-Auma ell, Mi ó, Se a-Cayuela, Buxade as, & López-Tamames,
2014).
When we analyzed bo h egula bee s, esul s showed (Table 2) ha
o all ola ile compounds he concen a ions we e highe o he G
sample han o he S sample wi h he excep ion o 2-me hylbu anol,
which was highe o he S sample (13.37 mg/L). Calcula ing he pe -
cen age o ola iles (di e en han e hanol) los in he dealcoholiza ion
p ocess a 102 mba and 200 mba a he end o he expe imen , i was
ound ha almos all ola ile compounds s udied we e e apo a ed
alongwi h e hanol wi h he excep ion o 2-phenyle anol.Fo he S sam-
ple, losses o 97% o es e s and 88% o alcohols we e obse ed a
102 mba and 76% o es e s and 95% o alcohols a 200 mba . Fo he G
sample losses o 96% o es e s and 92% o alcohols we e achie ed a
102mba , and 90% o es e s and 95% o alcohols o 200 mba . These ol-
a ile compound losses can be compa ed wi h hose epo ed in p e ious
s udies by o he au ho s using di e en dealcoholiza ion p ocesses
(Table 3).
F om ou esul s, we can conclude ha p essu e does no ha e a sub-
s an ial impac on he ela i e ola ili y be ween he e hanol and o he
a oma componen s; he e o e, we sugges ha he final ola ile p ofile
a e dealcoholiza ion canno be significan ly imp o ed by only modi y-
ing he p essu e. Thus, he main alcohols and es e s could be a ec ed by
he highe empe a u e applied a 200 mba . F om da a in Table 2 i can
be seen ha a 200 mba and 67 °C he ola ile compounds losses we e
highe o all compounds excep o he amyl alcohols in S sample and
e hyl ace a e in G sample. Looking a he se en ola ile compounds
quan ified in his s udy (Table 2), o he e hyl ace a e, he e apo a ion
was almos comple ed a he fi s 7.53% apo ac ion (V ), co espon-
den wi h he a e age o % V a 15 min o he p ocess (Table 1), in bo h
samples and p essu es ( om ini ial alues o 17.82 and 26.54 mg/L o
1.07 and 3.65 a 102 mba ; and o 4.09 and 5.18 mg/L a 200 mba , o
he samples S and G espec i ely), al hough o he 200 mba p essu e
he e apo a ion seems mo e g adually.
Table 3
Losses o o al es e s and alcohols in pe cen age (%) by di e en alcohol ee bee p oduc ion p ocesses: lab-scale acuum dis illa ion ( his wo k, p esen as he a e age o bo h samples
losses), osmo ic dis illa ion (Liguo i e al., 2015), acuum ec ifica ion (Mon ana i e al., 2009), alling film e apo a ion, dialysis (Liguo i e al., 2015) and e e se osmosis (S ein, 1993).
Lab-scale acuum dis illa ion Osmo ic dis illa ion Vacuum ec ifica ion Falling film e apo a ion Dialysis Re e se osmosis
To al es e s 97 (102 mba )
83 (200 mba )
99 100 95–100 99 78
To al alcohols 90 (102 mba )
95 (200 mba )
77 78 95–98 96 69
Table 4
Boiling poin s (°C) o he ola ile compounds a he di e en expe imen p essu es.
Boiling poin s (°C)
Compounds A mosphe ic p essu e 102 mba 200 mba
E hyl ace a e 77.1 13.7 32.3
1-P opanol 97 33.6 52.2
Isobu anol 107.9 44.5 63.1
Isopen yl ace a e 142 78.6 97.2
2-Me hylbu anol 127.5 64.1 82.7
3-Me hylbu anol 131.1 67.7 86.3
2-Phenyl e hanol 220 156.6 175.2
Table 5
Es ima ed pa ame e s o Elemen -1 o Wilson-2 equa ion in HYSYS.
Suc ose E hanol E hyl
ace a e
1-P opanol Isobu anol Isopen yl
ace a e
2-Me hylbu anol 3-Me hylbu anol 2-Phenyle hanol Wa e Ni ogen
Suc ose 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
E hanol 0.0000 0.0000 2.0000 −12.6000 −1.8000 9.5000 2.5000 2.5000 0.0000 −0.0503 0.0000
E hyl ace a e 0.0000 0.3000 0.0000 2.0000 0.0000 0.0000 0.0000 0.0000 0.0000 −1.8000 0.0000
1-P opanol 0.0000 12.6000 −0.8296 0.0000 0.9130 0.0000 3.0350 3.0350 0.0000 1.1919 0.0000
Isobu anol 0.0000 1.8000 0.0000 −0.7573 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Isopen yl ace a e 0.0000 −9.5000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 2.0000 0.0000
2-Me hylbu anol 0.0000 −0.3000 0.0000 −2.0368 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
3-Me hylbu anol 0.0000 −0.7256 0.0000 −2.0368 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
2-Phenyle hanol 0.0000 0 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Wa e 0.0000 −2.5035 −1.8000 −7.0000 0.0000 −2.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Ni ogen 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
6C. And és-Iglesias e al. / Food Resea ch In e na ional xxx (2015) xxx–xxx
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017
1-P opanol was almos comple ely gone wi hin he fi s o 10.52% V
in bo h expe imen s and bee s apa om he S sample a 200 mba , in
which i s los ex ended o a 14.43% V .
Isobu anol was e apo a ed g adually o e he p ocess ime cou se
unde bo h expe imen al condi ions bu , a he fi s 10.52% V mo e
han a hal o he concen a ion was emo ed ( om 9.41 and
10.47 mg/L o 4.68 and 4.32 a 102 mba ; and o 4.25 and 3.46 mg/L a
200 mba , o hesamples S and G, espec i ely). A simila end was ob-
se ed o isopen hyl ace a e, bu in his case mo e han a hal was e-
mo ed du ing he fi s 7.53% V .
Fo bo h expe imen s and samples du ing he fi s 7.53% V he
amoun o amyl alcohols (2-me hylbu anol and 3-me hylbu anol) was
educed by app oxima ely 50%, excep o he G sample a 102 mba .
A he end o he labo a o y dealcoholiza ion p ocess he amyl alcohols
we e in highe concen a ion o he S sample in bo h expe imen s
(102 mba , 50 °C and 200 mba , 67 °C).
A he end o bo h dealcoholiza ion p ocesses (17.87% V ) he con-
cen a ions o he majo i y o he ola ile compounds analyzed we e
highe o heSsample.
The a oma ic alcohol 2-phenyle hanol causes ‘swee ’o ‘ ose’fla o s
in bee (Šmog o ičo á & Dömény, 1999). Su p isingly, in his labo a o y
scale dealcoholiza ion p ocess 2-phenyle hanol was appa en ly p o-
duced du ing he expe imen al p ocess. This compound has a high boil-
ing poin (Table 4), and i was expec ed o sligh ly inc ease i s
concen a ion due o he apo iza ion p ocess ( ha educes he olume
o he emainingliquid). This ea u e wassimula ed using Aspen HYSYS,
ob aining ha 2-phenyle hanol should ha e inc eased i s concen a ion
by 3 o 5% maximum, as epo ed p e iously by Zü che e al. (2005).
Howe e , i s concen a ion inc eased a e he dis illa ion p ocess by
abou 30 o 50%, ha is om 37.69 ppm up o 59.97 ppm (G a
200 mba , 67 °C) and75.17 ppm (G a 102 ppm, 50 °C), and om an ini-
ial o 34.01 ppm up o 70.65 ppm (S a 200 mba ) and 85.28 ppm (S a
102 mba ).
I is well known ha , du ing e men a ion, 2-phenyle anol is o med
by phenylalanine ca abolism (Kobayashi e al., 2008). Highe alcohols
achie e maximum concen a ions du ing ba ch e men a ion a a ime
oughly coinciden wi h cell g ow h a es and minimum ee amino ni-
ogen (FAN) concen a ion. Thei o ma ion akes place by he so-
called anabolic and ca abolic ou es. In he anabolic ou e he 2-oxo
acids, a ising om ca bohyd a e me abolism, a e deca boxyla ed o
o m aldehydes, which a e educed o he co esponding alcohols. Si-
mul aneously, 2-oxo acids also de i ed om amino acid u iliza ion,
which is e med he ca abolic (Eh lich) ou e o highe alcohol o ma-
ion. The final concen a ion o highe alcohols is he e o e de e mined
by he up ake e ficiency o he co esponding amino acid and he suga
u iliza ion a e. The con ibu ion o each biosyn he ic pa hway is influ-
enced by he amino acid composi ion o he wo , e men a ion s age
and yeas s ain. In addi ion, some highe alcohols may o igina e om
he educ ion o aldehydes and ke ones ha a e p esen in he wo
(B ányik e al., 2008). Howe e , in he p esen s udy, he bee s used
we e comme cial bee s, so hey we e fil e ed and no e men a ion op-
ion is possible. The e o e, we explain his e ec by he po en ial deg a-
da ion and/o ans o ma ion o o he bee componen s due o a
combined e ec o empe a u e and esidence ime. I has been shown
ha , a indus ial scale, bee s ays only o a ew seconds in he
dealcoholiza ion p ocesses as i happens in hin film e apo a o s o
spinning cone columns (B ányik e al., 2012). Con e sely, in he expe -
imen al se up used in his s udy, he in e acial a ea o e apo a ion was
conside able lowe han ha in hin film e apo a o s, and, hence, he
ime equi ed o eaching he final e hanol con en (≤1%) was calcula -
ed o be nea ly 45 min, a esidence ime long enough o gi e ise o
ans o ma ion o o he po en ially ela ed compounds. In his sense,
one o he possible o ma ion ou es is om he deg ada ion o he
amino acid 2-phenylalanine, bu any o he componen om he same
me abolic ou e, e.g., phenyl py u a e, phenyl ace aldehyde o phenyl
ace ic acid can addi ionally lead o 2-phenye hanol in an acidic
Table 6
Es ima ed pa ame e s o Elemen -2 o Wilson-2 equa ion in HYSYS.
Suc ose E hanol E hyl ace a e 1-P opanol Isobu anol Isopen yl ace a e 2-Me hylbu anol 3-Me hylbu anol 2-Phenyle hanol Wa e Ni ogen
Suc ose 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
E hanol 0.0000 0.0000 −1000.0000 2066.8071 1159.7832 −3062.4265 −500.0000 −500.0000 0.0000 −69.6372 0.0000
E hyl ace a e 0.0000 −300.0000 0.0000 −500.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
1-P opanol 0.0000 51.4504 −500.0000 276.3888 −85.8744 0.0000 0.0000 0.0000 0.0000 −132.9251 0.0000
Isobu anol 0.0000 −1675.7465 0.0000 −29.2113 0.0000 0.0000 −442.0000 −442.0855 0.0000 −247.3062 0.0000
Isopen yl ace a e 0.0000 3724.3137 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 −1319.7350 0.0000
2-Me hylbu anol 0.0000 314.0000 0.0000 843.0000 243.0000 0.0000 0.0000 0.0000 0.0000 −462.0000 0.0000
3-Me hylbu anol 0.0000 314.4464 0.0000 843.8578 243.3993 0.0000 0.0000 0.0000 0.0000 −462.4493 0.0000
2-Phenyle hanol 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Wa e 0.0000 346.1512 0.0000 625.5155 −1633.2924 −1716.3821 −2102.0000 −2102.8264 0.0000 0.0000 0.0000
Ni ogen 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
7C. And és-Iglesias e al. / Food Resea ch In e na ional xxx (2015) xxx–xxx
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017
hyd ogen dono bulk liquid (i.e., wa e /e hanol) such as bee . When a
p olonged hea ing o bee is made, p obably he emained con en o
his amino acid o o he simila compound o ms he compound by e-
ac ion. Acco dingly, 2-phenyle hanol migh be used as a ma ke o
o e hea ing o o e iming in bee dealcoholiza ion p ocesses.
3.3. Simula ion esul s and he modynamic pa ame e s
In o de o demons a e he easibili y o a dynamic Aspen HYSYS
simula ion o he dealcoholiza ion p ocess, se e al he modynamic
packages we e conside ed. In he p esen s udy, i was necessa y o con-
side an EoS wi h in e ac ion in liquid phase, such as NRTL o Wilson.
Fo ou simula ion he bes esul s we e ound using Wilson-2 he mo-
dynamic package om HYSYS da abase.
Howe e , he de ia ions o he simula ion esul s agains he expe -
imen al esul s we e unaccep able using he de aul pa ame e s o he
so wa e. Thus, we ha e pe o med a fi o he selec ed bina y in e ac-
ion coe ficien s o he main measu ed compounds a 15 min, and
hen he simula ion was es ed o check whe he he sys em was able
o p edic o no he o he expe imen al da a poin s.
The bes fi pa ame e s o Wilson-2 Elemen -1 and Elemen -2 (i.e.,
in e ac ion pa ame e s acco ding o Aspen HYSYS nomencla u e) a e
lis ed in Tables 5 and 6 (see also Figs. 5 and 6 o componen concen a-
ion g aphs).
The p edic ions o he se en compounds analyzed we e e y ac-
cep able. A e age absolu e de ia ions (de e mined as he absolu e
alue o hesimula ed ins an concen a ion minus heexpe imen al in-
s an concen a ion, di ided by he ini ial alue o he concen a ion)
we e be ween 6.9 and 15.1% o bo h S and G bee s (excluding he
0.0
5.0
10.0
15.0
20.0
25.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
E hyl ace a e G
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
2.0
4.0
6.0
8.0
10.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
Isobu anol G
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
0.5
1.0
1.5
2.0
2.5
0% 5% 10% 15% 20%
ppm
% apo ac ion
Isopen yl ace a e G
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
2-Me hylbu anol G
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
10.0
20.0
30.0
40.0
50.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
3-Me hylbu anol G
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
2-Phenyle hanol G
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
2.0
4.0
6.0
8.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
1-P opanol G
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
Fig. 5. Concen a ion p ofiles o he main a oma compounds analyzed in he Ge man bee (G) a e he dealcoholiza ion p ocess.
8C. And és-Iglesias e al. / Food Resea ch In e na ional xxx (2015) xxx–xxx
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017
alues o 2-phenyle hanol ha beha es oddly). The alues ob ained by
simula ion (SIM) and expe imen a ion (EXP) a e lis ed in Table 2 (see
also Figs. 5 and 6).
Fo he case o 2-phenyle hanol i is clea ha he componen is gen-
e a ed by eac ion, so he simula ion canno p edic i as he assump ion
5(seeSec ion 2.4)isno ulfilled.
Conside ing he di ficul y o he analysis and he sys em i sel we
conside he simula ion alues o p edic ion accep able. This is he
fi s ime, o ou knowledge, ha bee is dealcoholized and he expe i-
men al alues a e fi o a simula ion and he modynamic model aimed
a c ea ing a p edic ion ool.
F om ou poin o iew, he p edic ion could be imp o ed by s udy-
ing he kine ics o o ma ion o 2-phenyle hanol and by s udying a pilo
scale plan using a sho - esidence ime equipmen (such as alling fill
e apo a o ), bu his is ou o he scope o his pape . Ne e heless, 2-
phenyle hanol appea ed om 15 min on, so his means ha he he -
modynamic app oach is alid o imes below ha ime ha indica es
ha i could be used o simula ion o sho esidence ime pieces o
equipmen .
4. Conclusions
Low alcohol and ee alcohol bee s om he mal dealcoholiza ion
(e.g., acuum dis illa ion) lack o he fla o and a oma compounds
ha he o iginal bee s possesses. Li e a u e da a on his is sca ce and,
so a , no simula ion ools o p edic he composi ions du ing he
dealcoholiza ion p ocess ha e been published.
In his s udy, we ha e obse ed how fla o compounds analyzed
anished o e y low concen a ion le els in he lab-scale acuum
0.0
5.0
10.0
15.0
20.0
25.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
E hyl ace a e S
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
2.0
4.0
6.0
8.0
10.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
Isobu anol S
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
0.5
1.0
1.5
2.0
2.5
0% 5% 10% 15% 20%
ppm
% apo ac ion
Isopen yl ac a e S
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
2-Me hylbu anol S
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
10.0
20.0
30.0
40.0
50.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
3-Me hylbu anol S
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
2-Phenyle hanol S
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
0.0
2.0
4.0
6.0
8.0
0% 5% 10% 15% 20%
ppm
% apo ac ion
1-P opanol S
SIM 60 SIM 102 EXP 102 SIM 200 EXP 200
Fig. 6. Concen a ion p ofiles o he main a oma compounds analyzed in he Spanish bee (S) a e he dealcoholiza ion p ocess.
9C. And és-Iglesias e al. / Food Resea ch In e na ional xxx (2015) xxx–xxx
Please ci e his a icle as: And és-Iglesias, C., e al., Simula ion and fla o compound analysis o dealcoholized bee ia one-s ep acuum
dis illa ion, Food Resea ch In e na ional (2015), h p://dx.doi.o g/10.1016/j. ood es.2015.07.017