. ._ Ekatfiau
.éwfifiwfiwfi? a.
._ s. . Hagan.
«Jeans?
‘53:;
.V s. L

was“,

7 7 :
5.12:3”)
5. ,

.2. uh...

5...; s
3:! hum”...

. a:
‘ Inflammg‘ x
n... Irmuz.
.3.» :5 . I l .I. .1
.4. ”Hangar t 5H: .3. {6%
- 3 1. J...
, . dim?! “and...
58?; ~51 I
.p s... D, . unsung...
ninth» «3.6L» , £51.54; a x A,
i, . .. 22. s.
; ., .. 19%.»... 35%.}:
.. r X n I. ‘ .
155.313.»? a. a ,2
v.13 V 2. ‘

.24..

V
I z

Lu. ...

..

v .2. 1.5..

5... n i .i. 1......
2...: 3.2..

x..
.. flaws...“
h . .

fisflsflfl? .4 .

.4!

v2?!
.1... Q
.55.

a... 95.11...“

ska—Rf:

v .éfulttwnr: in.“

Wrflnfifglhnfllf
v.3wf~lt|‘

x.
when
:

 

 

 

\ . .
5:2; 2!. .25,
3.3.9133: Hr!
.3..:i1...3.

:3

.331! L

- .32.... .z.

.4 : :...f.x.h»:.... r

g?! 3...)..55... ..

. 2:23.215:
.1 1. .11..

.2 5.

.. .!f 5.1.
3.3.93.3.
{sift .2.

5:2...

 

2......

u
533‘... A
\nluzlti
«3.13.125
. :10

iii...

Riga. éfifig _ ,

. 2.:nuflr . .nxlfldfihflsw
‘ . . . 5‘ V“ . 3A....“
. : .5 ... . f¢$§fi “in. .n

 

3..“
r. 7?}?

”bursa
. pit-pin.
.__m£ . fig
, . .L. .
5...; . ‘

THESIS
l

DOOT

 

 

LIBRARY
wichlgan Sigig
University

._J-— -—.———

 

 

.M‘ '

This is to certify that the

thesis entitled

AN INVESTIGATION OF RELATIONSHIP BETWEEN
TRAY USAGE OF THE STILL AND CONGENER COMPOUND
CONCENTRATION IN DISTILLED FRUIT SPIRITS

presented by

Michael Joseph Claus

has been accepted towards fulfillment
of the requirements for

M. S . degree in Chemi Stf‘jl

Major professor

Date/é gig/£4 829/7063/

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

C...

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

 

DATE DUE

DATE DUE

DATE DUE

 

lll

ARMS

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DatoDue.p65-p.15

 

 

AN INVESTIGATION OF RELATIONSHIP BETWEEN TRAY USAGE
OF THE STILL AND CONGENER COMPOUND CONCENTRATION IN
DISTILLED FRUIT SPIRITS
By

Michael Joseph Claus

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

2001

ABSTRACT

AN INVESTIGATION OF RELATIONSHIP BETWEEN TRAY USAGE
OF THE STILL AND CONGENER COMPOUND CONCENTRATION IN
DISTILLED FRUIT SPIRITS

By

Michael Joseph Claus

The relationship between the operating parameters of batch column stills with
reflux and the congener (trace compounds that provide flavors and aromas)
concentrations in the resulting distilled beverage product is not fiilly understood. To
explore this relationship congener concentration were determined in three different
collection fraction, or “cuts,” of the batch distillation. Methanol was studied because the
distillate must adhere to governmental regulations that limit its concentration in the
product. Ethanol was studied as changes in the operating conditions of the still affect the
ethanol concentration and resulting yield of the process. Operating condition parameters
that were studied include the number of trays used in the distillation as well as the use of
the “catalytic converter,” a high surface copper packing material thought to catalyze
formation of cyanide containing compounds allowing them to be separated from the
distillate. The effect of the number of trays used in a distillation on the concentration of
ethanol and the highest concentration congeners, methanol, acetaldehyde, ethyl formate,
ethyl acetate, l-propanol, 2-methyl-2-propanol, isoamyl alcohol, and benzaldehyde in the

final distilled spirits product is presented.

ACKNOWLEDGEMENTS

First I would like the person who without which any of this would not have been
possible, Dr. Kris Berglund. His assistance and guidance throughout this project has
been a great support throughout this endeavor. I would also like to thank the members of
my committee whose support has been greatly appreciated. Another thanks is extended
to Dr. Klaus Weispfennig while we did not always see eye to eye, your guidance during
the early portion of the project was invaluable.

I would like to thanks the members of my group both past and present that have
added to the rich experience of being a graduate student here at Michigan State
University. I especially would like to acknowledge those group members who have
worked directly on this project at one time or another; Klaus, Anna, Eric, Dave, Heather
and Deirdre have all been influential in the completion of this project.

I would also like to thank the support of my family and friends for their words of
encouragement in good times and in bad. The list of friends I would like to thank is long
and for that I am blessed, however, I will keep the list short by only mentioning the
dearest ones of all; Joann, John, Mike, Holly, Pat, Amy, Gary, Jenny, Rob, Janet and
Brad...

I would like to thank the vintners and distillers who have shared their interest in
the production of fruit spirits. The growing number of stills at wineries has kept me
focused to completion of this thesis. Finally thank you for reading I hope it is useful and

enjoyable reading for you.

iii

TABLE OF CONTENTS

1. INTRODUCTION

1.1

1.2

1.3

1.4

1.5

1.6

1.7

Michigan Fruit Brandy Industry
Distillation Process

Batch Distillation of Fruit Spirits
Congener Compound Formation
Catalytic Converter

Objective

Literature Cited

2. BACKGROUND INFORMATION

2.1

2.2

2.3

2.4

2.5

Distillation History

Distilled Beverage Flavor Analysis
Methanol Health Hazards and Regulation
Making Quality Fruit Spirits

Literature Cited

3. MATERIALS AND METHODS

3.1

3.2

3.3

3.4

Distilled Spirits Production Process
3.1.1 Fermentation

3.1.2 Distillation

3.1.3 Storage

Analysis of Distilled Spirits

Tray Usage

Literature Cited

iv

11

12

13

14

14

15

18

19

22

24

24

24

27

31

32

33

36

4. RESULTS AND DISCUSSION
4.1 Composition of Fruit Brandy
4.2 Batch Reproducibility
4.3 Distillate Analysis
4.4 Effect of Tray Usage
4.4.1 Ethanol
4.4.2 Acetaldehyde
4.4.3 Ethyl formate
4.4.4 Ethyl acetate
4.4.5 l-Propanol
4.4.6 2-Methyl-2-propanol
4.4.7 Isoamyl alcohol
4.4.8 Benzaldehyde
4.4.9 Methanol
4.4.10 Bottles of fruit spirits
4.5 Catalytic Converter Use
4.6 Literature Cited
5. SUMMARY AND CONCLUSIONS

6. FUTURE WORK

37

37

38

41

43

43

44

47

50

50

53

53

56

58

6O

60

63

64

68

Table 3.1

Table 4.1

List of Tables

Average retention times from the chromatographic conditions 34
used for the most common congeners present in distilled fi’uit

spirits. The boiling point for each congener is also given.

Fermentation and distillation conditions. The conditions used 39
for each distillation and how they relate to the fermentation

batches of the fruit used. The distillation conditions used

include: (1) All trays used and catalytic converter engaged, (2) all
trays used with no catalytic converter, (3) two trays used with
catalytic converter engaged, (4) two trays used with no catalytic

converter.

vi

Figure 1.1

Figure 1.2

Figure 1.3

Figure 2.1

Figure 3.1

Figure 3.2

List of Figures

McCabe-Thiele diagram. Using more stages yields a higher
concentration of ethanol in the distillate. The values

demonstrate the effect and are not experimental.

The process of making distilled fruit spirits.

Hydrolysis of amygdalin. The process of the breakdown of
amygdalin to benzaldehyde, hydrogen cyanide, and glucose

in the fermentation mash.

Volatiles in distilled spirits. The amount of volatile
congeners in the distilled spirits per 100 mL of alcohol,
notice fruit spirits have higher amounts of volatiles than

other distilled beverages.
Schematic design of 200 gallon fermentation system.
Copper coil wound around a PVC trellis is used for the

cooling of the system.

Schematic of Christian Carl 165 L still.

vii

10

16

25

29

Figure 3.3

Figure 4.1

Figure 4.2

Figure 4.3

Sample Chromatagraph from a cherry eau-de-vie 35
distillation. A Restek 30m Stabilwax column with 0.25mm

i.d. was used for analysis. The initial conditions had the

column temperature at 40° C, injector port temperature at

240° C and detector temperature of 255° C. The analysis

included a ramp from 40° C to 190° C at 25° C/ min.

Ethanol concentration as a function of cumulative distillate 42
volume for cherry spirits distillations. All distillations were

from the same fermentation batch and were run with the same
operating conditions used. All distillations were carried out

with all the trays used and the catalytic converter used as well.

The average concentration of ethanol (% v/v) as a function of 45
distillate volume with varying distillation conditions. (1) All

trays used and catalytic converter engaged, (2) all trays used

with no catalytic converter, (3) two trays used with catalytic

converter engaged, (4) Two trays used with no catalytic converter.

Translation of the data in Figure 4.2 into bottles of 40% 46
ethanol, plum brandy at the four different operating

conditions. (1) All trays used and catalytic converter

engaged, (2) all trays used with no catalytic converter, (3)

two trays used with catalytic converter engaged, (4) Two

trays used with no catalytic converter.

viii

Figure 4.4

Figure 4.5

Figure 4.6

Acetaldehyde profile. Concentration of acetaldehyde in plum 48
distillate as a function of cumulative volume of distillate at

four different operating conditions. (1) All trays used and

catalytic converter engaged, (2) all trays used with no catalytic
converter, (3) two trays used with catalytic converter engaged,

(4) Two trays used with no catalytic converter.

Ethyl formate profile. Concentration of ethyl formate in plum 49
distillate as a function of cumulative volume of distillate at

four different operating conditions. (1) All trays used and

catalytic converter engaged, (2) all trays used with no catalytic
converter, (3) two trays used with catalytic converter engaged,

(4) Two trays used with no catalytic converter.

Ethyl acetate profile. Concentration of ethyl acetate in plum 51
distillate as a function of cumulative volume of distillate at

four difl‘erent operating conditions. (1) All trays used and

catalytic converter engaged, (2) all trays used with no catalytic
converter, (3) two trays used with catalytic converter engaged,

(4) Two trays used with no catalytic converter.

ix

Figure 4.7

Figure 4.8

Figure 4.9

Concentration of l-propanol in plum distillate as a function 52
of cumulative volume of distillate at four different operating
conditions. (1) All trays used and catalytic converter engaged,

(2) all trays used with no catalytic converter, (3) two trays

used with catalytic converter engaged, (4) Two trays used with

no catalytic converter.

Z-Methyl-Z-propanol profile. Concentration of 2-methyl-2- 54
propanol in plumdistillate as a function of cumulative volume

Of distillate at four different operating conditions. (1) All trays

used and catalytic converter engaged, (2) all trays used with no
catalytic converter, (3) two trays used with catalytic converter

engaged, (4) Two trays used with no catalytic converter.

Isoamyl alcohol profile. Concentration of isoamyl alcohol 55
in plum distillate as a fiJnction of cumulative volume of

distillate at four different operating conditions. (1) All trays

used and catalytic converter engaged, (2) all trays used with

no catalytic converter, (3) two trays used with catalytic

converter engaged, (4) Two trays used with no catalytic

converter.

Figure 4.10

Figure 4.11

Figure 4.12

Benzaldehyde profile. Concentration of Benzaldehyde in plum 57
distillate as a function of cumulative volume of distillate at

four different operating conditions. (1) All trays used and

catalytic converter engaged, (2) all trays used with no catalytic
converter, (3) two trays used with catalytic converter engaged,

(4) Two trays used with no catalytic converter.

The concentration of methanol in plum distillate as a fimction 59
of cumulative volume of distillate at four different operating
conditions. (1) All trays used and catalytic converter engaged,

(2) all trays used with no catalytic converter, (3) two trays used

with catalytic converter engaged, (4) Two trays used with no

catalytic converter.

Congener concentration in drinking strength eau-de-vies 61
(40% ethanol v/v) produced by blending and dilution of

hearts cuts. (1) All trays used and catalytic converter engaged,

(2) all trays used with no catalytic converter, (3) two trays used

with catalytic converter engaged, (4) two trays used with no

catalytic converter. The concentration range of ethyl acetate was

too small to be included with this data set.

1. INTRODUCTION

1.1 Michigan Fruit Brandy Industry

The Michigan fruit brandy industry has been growing since recent changes in the
state laws of Michigan regarding the licensing of fi’uit distilleries. The number of stills in
Michigan for the distillation of fruit spirits has grown from zero in 1996 to seven in the
year 2000. The distillation of fi'uit spirits has been practiced for centuries in Europe;
however, only recently has the distillation industry grown in the United States. The two
competing styles for fruit brandy distillation are alambic, i.e. French style, and the other
is batch distillation, i.e. German style. The French style, resulting in cognac style spirits,
involves distillation through a simple pot still, requiring multiple distillations in order to
obtain high proof spirits. The German style of fiuit distillation, which results in eau-de-
vie or schnaps, involves a single pass through a batch still with a reflux column to obtain
high proof spirits. These spirits are traditionally stored in glass and are served as water
clear brandies‘. The batch column still is designed to preserve the delicate fruit essences,
making it the desired distillation process for most fruit spirits.

The United States industry faces challenges that differ from those in Europe. The
process of distilling fruit spirits involves fermentation of the whole fruit. Utilization of
the whole fruit leads to a better sensory experience for the consumer of the fi'uit brandy;
however, by using the pectin of the huh, and the stones in stone fruit spirits, the
concentrations of many of the congener components (compounds other than ethanol and
water) can be greatly increased from other classes of distilled spirits. Many types of
congeners are present in the fruit spirits. Some compounds have higher concentration in

the early portion of the distillation process, and others are higher in the later portion of

the distillation process. Because the distillation process is a method of concentrating, or
separating the volatile compounds from the non-volatile compounds, higher
concentrations of these volatiles are found in the distillate. Some of these volatile
compounds, such as methanol, can have chronic health effects if relatively high
concentrations are consumed, and are therefore regulatedl’ 2. Compounds such as
urethane (ethyl carbamate) which is thought to be a carcinogen, and methanol, which can
cause many health risks, are regulated in many countries, but the regulations differ. The
United States has currently set the regulation of methanol in fruit spirits to be 0.35% v/v,
where in Europe the regulations for methanol have been set higher and therefore are not
as much of a concern there“ 2. The regulations on urethane in Canada are set at 400ppb3 .
Currently the United States has not set a regulatory limit, however, the fruit spirits
industry needs to minimize the concentration of the urethane in the fi'uit spirits to avoid
regulation.

It is widely known that the greater the number of trays used in a distillation
process the more separation of the components will be achievedl’ 4’ 5’ 6. An example of
this can be seen in Figure 1.1, which represents a sample McCabe—Thiele diagram, which
shows that the use of seven trays yields a higher concentration of ethanol in the distillate
than using two trays. However, most studies have included two or three component
mixtures, whereas fruit spirits have hundreds of different compounds present in the
distillate, many of which are positive flavor components“. The composition of the fruit
distillate must have a good aroma and flavor in order to sell at market, but it must also
meet limits for those compounds that are regulated. The compounds that are present in

the distillation, other than the two main components, ethanol and water, are called

2E. 6353:. 2: E .2550 .3 5:95.339 ..oswE a 3.8% Swen 0.8:. min: ...—Ema:— 202.73.692

.QHQU _5=0=—T—0n—fl0 «O: 0&6 nun—a ugh—0 05 OuEHmEGEQ—u mos-5r

3 2:5

 

 

 

 

 

 

 

:Qm >
3 ad ad to No o
A r _ _ moduuzx
ii! i. . N>> - n O
momma N - 6:: do .. ...... no on x. m
momma n- 6:: do .. .. .. _
> H X n w I N.O
\. m
6:: E25=Scmlll ...... .. ...}... ....... 4 .. L
.. it t ...... in. i v.0
. ...... ... h H h I ... . X
om ouEx .............. .... ....H......u...._ ............... m
... Isl l I I . H
ivu._ -oo
Eduax .. I.
., md

 

 

 

 

 

congeners. Some congeners are present in greater concentration in the early part of the
distillation and others are present in higher concentrations at the end of the distillation,

while still others have a more consistent concentration throughout the distillation process.

1.2 Distillation Process

The distillation of fruit brandies can be broken down into four basic steps: fiuit
processing (mashing), fermentation, distillation, and storage]. A representation of the
process involved in making fruit spirits can be seen in Figure 1.2. The fi'uit chosen for
distilled spirits can include fruit unsuitable for market such as that fruit which is overripe
and those with blemishes, making the fi‘uit from late harvest the ideal choice for use in
distillation. Fresh fruit donate better aromatic and flavor properties than that which has
been frozen.

The mashing of the fruit can take many forms, from mashing the fruit beneath
ones feet, to a modern rolling mill or processing pump. For large volumes of fruit where
manual mashing of the fiuit is not an economical process, rolling mills and progressive
cavity pumps are utilized. Both the rolling mill and pump will crush the pulp of the fi'uit
while keeping most of (~90°/o) of the pits, in the case of stone fruit, from being cracked.
The cracked pits from stone fruit are thought to increase the amount of urethane and
benzaldehyde in the distilled spiritsl’ 3. The compound amygdalin is found in the pits of
stone fruits and one mole can be hydrolyzed to two moles of glucose, and one mole each
of cyanide and benzaldehyde. Benzaldehyde is a positive flavor component and is

commonly known as bitter almond oil, but urethane is a suspect carcinogen.

«0—>£z—>

Crushed or Mashed

Ripe Fruit Fruit

 

Fermentation
1-2 Weeks

 
  

\

 

Distillation
Storage
1-3 Months
. Cut
g with Water
0
Consumption

Figure 1.2. The process of making distilled fruit spirits.

The resulting mash from the processed whole fruit is then transported to a
fermenter where yeast is added to begin the fermentation. The yeast added must first be
inoculated in warm water before being added to the mash. By adding excess yeast, the
added strain can effectively compete with the wild strains of yeast present in the mash
and prevent their growth. A method of stirring the fermentation is useful for enhanced
heat transfer, where a temperature control is essential to the fermenter used. The
fermentations of fruit mashes should be kept within the temperature range from between
15° and 20° C. If the temperature becomes too high the yeast will produce more
compounds such as ethyl acetate, and [Use] alcohols. If the temperature drops too far
below 15° C the yeast will be unable to efficiently ferment the fruit mash. Measurement
of the brix (% sugar, weight basis) is an ideal way to examine the progress of the
fermentation. Brix can be measured using refractive index, and a more detailed
observation of the fermentation can be done through I-IPLC with either refractive index
detection, or UV/V is detection. The pH of the fermentation should also be controlled.

The yeast survive in a pH range greater than 2.8 and less than 5.2. A pH range of 3.5 to
4.2 is In commonly usedl. The pH of the mash can be changed with a number of
different acids or bases, however, phosphoric acid or ammonia are considered the favored
solutions to change the pH of the solution because each provide either nitrogen or
phosphorus nutrients for the yeast. If a fermentation becomes " stuck, " or appears to have
a high concentration of sugar dissolved in solution, but is not producing the C02 side
PFOduct or not reducing the sugar concentration, the reasoning is often because the yeast
does not have enough nutrients. Dibasic ammonium phosphate is the best choice for

adding nutrients to a stuck fermentation.

The average fermentation lasts for ten to fifteen days. The fermented mash is
distilled in a batch column still with reflux in a single pass. The distillate is collected in
volumes referred to as heads (discarded first fraction), hearts (product), and tails
(discarded or redistilled last fraction). The clear spirits are then stored at high (>100°)
proof. The storage of the spirits lasts for as little as three months prior to proof

adjustment with distilled water. The product is then bottled and is ready for

consumption.

1.3 Batch Distillation of Fruit Spirits

From an engineering standpoint the mass balances of batch distillation are
somewhat different from those for continuous distillation. In batch distillation we are
more interested in the total amounts of bottoms and distillate collected than in the ratesé.
This correlates to the German farmer who wishes to maximize distillate per volume of
fiuit mash than they are about the time it takes to run the still.

The distillation of fruit brandies is more complicated than the simple binary
distillation of ethanol and water. There are many compounds present in varying
concentrations throughout the process of distillation of fruit spirits. Fruit distillates differ
widely in quality and composition, dependent on the raw material used and the
processing procedures3 . These various parameters lead to poor reproducibility of the
distilled spirits form one batch to the next. In comparison to other spirit groups (e. g.
whiskey, vodka, gin, ...) they contain the highest amounts of alcohols and esters. The
Sum of volatiles in fruit brandies amount to an average of nearly 2000 mg/ 100 ml ethanol

With exceptions of Calvados (apples) and raspberry brandy. This can be compared with a

Scotch whiskey, which has an average sum of the volatiles of about 200 mg/ 100ml

ethanol3 .

1.4 Congener Compound Formation

Various acids, esters, and fusel alcohols form the main body of the aroma
compounds in the distilled spirits7. Some other compounds are also present as congeners,
however, the majority of the congeners in the spirits are these aroma compounds. The
formation of most of the congeners in the distillation occurs in the fermentation of the
mash in the influence of yeast7. Fusel alcohols, defined as alcohols with more than two
carbons, compose the largest group of aroma compounds in alcoholic beverages. Isoamyl
alcohol is the main fusel alcohol synthesized during fermentation by yeast. Other
important fusel alcohols include n-propanol, isobutyl alcohol, and optically active amyl
alcohol7. Formation of these fiisel alcohols is thought to be independent of the raw
materials used in the mash in that the formation of these longer chain alcohols can occur
in whiskeys as easily as in tequila, gin, or fruit spirits. N—propanol, and branched C4 and
C6 alcohols are formed from valine, leucine, and isoleucine in the presence of yeast. on-
keytobuteric acid is thought to act as in intermediary in formation of n-propyl alcohol and
optically active amyl alcohol7’ 8. Fusel alcohols are thought to form in fermentation under
both catholic conditions from amino acids and anabolic conditions from sugars7.

Yeast synthesizes fatty acids irrespective of the nature of the raw materials
involved in the fermentation]. The acids formed in the fermentation mash are thought to

be bacteria products as well as byproducts of the formation of alcohol by the yeast. It has

been shown that with semi-aerobic conditions in the fermentation, lower temperatures
result in an increased amount of organic acids in the distilled spirits7.

Carbonyl compounds, specifically aldehydes, are thought to be intermediates in
the formation of fusel alcohols, and acids by the yeast. Aldehydes can also be oxidized to
form more acids when the fermenter is not sealed properly creating aerobic conditions
rather than the anaerobic conditions found in most fermentations]. Most esters found in
the fermentation are ethyl esters due to the high concentration of ethanol relative to all
the other alcohols in the fermentation mash. It has been shown that distillation in the
presence of yeast will increase the amount of ethyl esters present in the distillate when
compared to the distilled spirits with no yeast present in the distillation pot7.

Two other compounds of importance in the distillation process can be traced to
the pits of stone fruit. Amygdalin contains two moles of glucose for every one mole of
benzaldehyde and one mole of cyanide. The hydrolysis of amygdalin in the fermentation
produces the benzaldehyde, which is desired, and the prussic acid, which is undesired, as
it is thought to be a reagent in the formation of ethyl carbamate3. Figure 1.3 shows the
hydrolysis of amygdalin.

Methanol is also formed through fermentation of the fruit sugars; however, a
prominent source of the methanol is an enzymatic reaction of pectinesterase with the
pectin of the fiuit. Although methanol is a desired compound for flavor and aroma, and

at the same time a regulated compound for the health hazards it possesses, a method for

limiting the amount of methanol produced has not been discovered to product.

CH 2OH

 

 

 

o_CH20 CN
O_CH Water
H H OH © Catalyst
Amygdalin
CHZOH HOCH;O
Water 0
OH +H O_C'H
Catalyst H OH
H
Glucose Prunasin©
CIIZOH CN
0 I Dissociation
OH HO—CH
2 H OH ©
H
Glucose Mandelonitrile
CH 2OH

2 H<)E:C:QOHO +© CH+ HCN
Hydrogen cyanide

Benzaldehyde

Glucose

Figure 1.3

Hydrolysis of amygdalin. The process of the breakdown of amygdalin

to benzaldehyde, hydrogen cyanide, and glucose in the fermentation

mash3.

10

Overall most of the congeners in the distillation of fruit spirits are products of the
fermentation of the fruit by the yeast7. The distillation of these compounds also makes
more of the congeners especially in the case of ethyl esters. Further synthesis of the
flavor and aroma compounds occurs in the storage of the spirits; however, these are

mostly oxidative processes between the congeners already present in the spirits].

1.5 Catalytic Converter
The catalytic converter of the still contains a high surface area copper packing

material in a stainless steel vessel. The distillate vapor is introduced through the bottom
of the packing material and fromthe top of the vessel on its way directly to the
condensing column. The catalytic converter has valves located in the distillation stream
such that it can be bypassed if desired. The catalytic converter has been thought to
remove urethane from the distilled fruit spirits. The theory behind the catalytic converter
involves the cyanide containing compounds binding to the copper metal surface in the
catalytic converter, and thus being effectively removed from the distillateg. These
cyanide-containing compounds, which can be traced back to the amygdalin in the pits of
the fiuit, would form cyanic acid in the distillate, and then during storage of the distillate,
the cyanic acid would react with the ethanol to form ethyl carbamate. It is not well know
if the catalytic converter will function to effectively remove enough cyanide containing
compounds to keep the spirits under the 400 ppb limit the Canadian government has set
regulating ethyl carbamate in distilled spirits3 . The focus of this thesis does not deal with
the effectiveness of the catalytic converter as a tool for removal of the cyanide derived

compounds from the distillate. Instead, the catalytic converter was studied with respect

ll

to its ability to act as an additional tray in the multistage distillation process. The
catalytic converter it should be noted is not a true definition of a catalytic converter,
however, the German still manufacturers have called it a catalytic converter and the name

persists9.

1.6 Objective

The objectives of this study are to determine a characteristic profile for
some of the congeners present in distilled fruit spirits and how changing the operating
conditions used in the distillation process affects the concentrations of these components.
Another objective is to test the usefulness of the catalytic converter portion of the still as
to how it affects the congeners of the distilled spirits. A further aim includes developing
a profile for the ethanol concentration in the batch distillation with reflux of hit spirits
and how the number of trays used affect this concentration. The final purpose of these
experiments is to develop a method for determining the approximate concentration of the
congener components in the bottled fruit spirits from the cuts taken of distillate during the

distillation process.

12

1.7

Literature Cited.

Tanner, H., and Brunner, H.R., Fruit Distillation Today 3rd ed., (English

Translation) Heller Chemical and Administration Society, Germany, 1982.

United States Code of Federal Regulations, Title 27—Alcohol, Tobacco Products
and Firearms, Chapter I, Part 5, Subpart C—Standards of Identity for Distilled

Spirits. <www.accessgpogov/nara.cfr/cfr-table-search.html>.

Zimmerli, B., and Schlatter, J ., Ethyl Carbamate: Analytical Methodology,
Occurrence, Formation, Biologic Activity, and Risk Assesment. Mutation

Research, 259, 1991.

Piggott, ”JR, and Patterson, A. ed., Distilled Beverage Flavour: Recent
Developments. Ellis Horwood Ltd, 1989.

 

Leaute, R., Distillation in Alambic. Am. J. Enol. Vitic., 41, (1) 1990.
Wankat, P.C., Equilibrium Staged Separations. Prentice Hall, New Jersey, 1998.

Suomalainen, H. and Lehtonen, M., The Production of Aroma Compounds by

Yeast. J. Inst. Brew., 85, 1979.

Guymon, J .F ., Higher Alcohols in Beverage Brandy: Feasibility of Control of
Levels. Wines and Vines, 53, (1) 1972.

Plank, Alexandr, Personal correspondence with author. 1998-2001.

13

2. BACKGROUND INFORMATION

2.1 Distillation History

The history of distillation is thought to have stared around 800 BC in China and
India with the concentrating of the alcohol from rice and barley beer. Distillation was
also used in the Middle East and Mediterranean for the concentrating essences for
perfumes. Around 1100 AD in Salerno doctors began using distillation to extract alcohol
from wine because it was thought that alcohol had medicinal effects. In 1300 Catalan
physicians gave the name aqua vitae, "water of life, " to the distillate of wines. This term

can be found translated into many different Ianguagesl’ 2.

The addition of herbs, spices, or fruit was used to mask the harsh taste of the
distillate and was thought to add to the medicinal effect of the spirits. The distillation
process used in antiquity involve only a pot and simple one tray column with a
condenser. One pass of the wine would yield a harsh tasting distillate, and thus need to
be masked. This style of distillation is still common today in the Cognac region of
France and the whisky distillers in Scotland. These distillers use multiple passes through
their alambic type stills, but storage of these spirits must be done in wooden barrels. The
alcohol will extract flavors from the barrels giving the classic flavors to the whisky and
cognac. The alambic style still, developed in the middle ages, involves a pot, alembic
head, which functions as the distillation column, and a condenser. Multiple passes
through this type of still must be made in order to concentrate the ethanol out of whatever

. - - 2
mash is used to the concentration desrredl’ .

14

In the 19‘h century the continuous still was developed. This type of still allows
the passing of vapor through multiple stages on one distillation. This method avoids the
need to send the distillate through the still more than once which leads to losses of the
total volume that could be collected. The modern multistage still can be made with over
a hundred trays to remove most flavor components for the production of vodkas. The
production of fiuit spirits by European farmers has been going on for hundreds of years,
and uses these multistage stills. The market for these types of fruit spirits, which has
begun in the United States in the last decade, is attempting to follow the example of the
European farmer in the production of their fruit spirits. While distillation has come a
long way, the understanding of the interaction of the hundreds of compounds found in the
distillate and how their interaction effect the overall quality of the distillate needs to be

studied morel’ 2.

2.2 Distilled Beverage Flavor Analysis

Fruit distillates differ widely in quality and composition, dependent upon the raw
materials used and the processing procedures. In comparison with other distilled spirits
fruit distillates have more volatiles, as well as the highest amount of esters and alcohols.
The sum of the volatiles in fruit spirits in comparison to other distilled beverages can be
seen in Figure 2.1. The concentration of volatile compounds in fruit distillates can be as
much as one hundred times as much as could be found in gin or vodka3.

Methanol is quantitatively the main congener of stone and pome fi'uit brandies.
Methanol poses a health hazard, as methanol is associated with acute and chronic health

effects, however, it is also a desired component for flavor and aroma. Because a certain

15

 

 

I Methanol

El Higher alcohols
Carbonyl compounds
Ethyl acetate

I Other esters

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2.1 Volatiles in distilled spirits. The amount of volatile congeners in the
distilled spirits per 100 mL of alcohol, notice fruit spirits have higher

amounts of volatiles than other distilled beverages“.

l6

 

minimum amount of methanol is formed by enzymatic cleavage from the pectin of the
fruit, methanol has been used for the evaluation of the authenticity and possible
adulterations such as addition of neutral alcohol to the distillate, or addition of sugar to
the mash3.

Fruit spirits contain high amounts of 1-propanol, l-butanol, 2-butanol, and 1-
hexanol when compared with other types of spirits. Isoamyl alcohol and optically active
isoamyl alcohol are also main components of fruit spirits, however, they are in a range
comparable to other types of spirits. Other alcohols are present in smaller amounts, and
of these benzyl alcohol is the most important, as it can be formed from the benzaldehyde
which forms from the hydrolysis of amygdalin in the pits of the fruit3.

Acetaldehyde is the most important of the carbonyl compounds. Even though it
has a low boiling point, acetaldehyde can dissolve in the ethanol water matrix and stay in
solution. Benzaldehyde, which comes from the pits is also an important carbonyl
compound in stone fruit spirits, as it can give an indication as to how much cyanide is in
the distillate3’ 5.

The main ester component in fruit spirits is ethyl acetate and ethyl lactate, which
composes 80% of the total ester concentration in the fruit spirits. Ethyl acetate is also
present in the distillates of fruit brandies, and ethyl benzoate and benzyl acetate are
present in higher concentration in stone fruit spirits than in other fruit spirits. The
distillation procedure can affect the concentration of esters in the distillate. Heads cutting
during the distillation results in lower concentrations of acetaldehyde, ethyl acetate and

higher fatty acid esters. On the other hand ethyl lactate and diethyl succinate are extreme

l7

tails components and can be reduced by making the tails cut earlier in the distillation
procedure3.

Ethyl carbamate (urethane) is thought to form from cyanic acid reacting in a light
catalyzed reaction with ethanol in the storage of distilled spirits. For reasons of health the
Canadian government has established a limit on the concentration of ethyl carbamate in

fruit spirits of 400 pg / L. The cyanic acid is thought to come from the cyanide that is

formed in the hydrolysis of amygdalin in the fermentation of the spirits3’ 5 .

2.3 Methanol Health Hazards and Regulation

Methanol has many acute health hazards associated with it. Methanol can cause
damage to a number of organs in the human body, including the liver, kidney, and
nervous system. In some cases the ingestion of methanol has caused temporary or
permanent blindness and even death. Methanol can not be made non-poisonous,
although, the treatment for ingestion of methanol involves giving the patent ethanol.
Ethanol can act as an inhibitor to the methanol effects on the body6’ 7.

The United States Environmental Protection Agency recommends a minimum
acute toxicity concentration of methanol in drinking water to 3.9 parts per million. The
EPA has also set the permissible exposure limit in air as 200 parts per million for an eight
hour time weighted averages.

Methanol has been widely regulated in its production in fruit spirits throughout
Europe. The German government regulates the maximum methanol content in fruit
brandies at different levels based upon the fruit used. In cherry spirits methanol is

regulated at 400 mg/ 100 mL absolute alcohol, while Bartlett pears have a regulatory limit

18

of 790 mg/ 100 mL absolute alcohol. Italy and Austria have limits set at 800 mg/ 100 mL
absolute alcohol, and 1000 mg/IOO mL absolute alcohol, respectively9. The United
States regulates the amount of methanol in distilled fiuit spirits at 0.35% v/v for all fruit
spirits sold in the United States. This corresponds to a value of 700 mg/ 100 mL absolute
alcohol").

Although methanol is heavily regulated in the United States, the laws were
written more to prevent the cutting of distilled spirits with methanol, than they were as a
concern for having too high a concentration of methanol in the distillate. As ethanol is
the antidote for methanol poisoning and the concentration of ethanol is 40% compared to
the methanol concentration of 0.35%, one would have to assume that the ethanol would

be lethal before the methanol would. However, a low concentration of methanol is still

desired to avoid any toxicological effects3.

2.4 Making Quality Fruit Spirits

Many factors effect the quality of distilled spirits. The choice of fi'uit is
important, in that the higher quality fruit will yield higher quality distillate. The ripest
fruit is desired, because it has more sugar, which in turn would mean more alcohol in the
fermentation. Fruit without visible blemishes is preferred, as fiuit with imperfections in
the skin are also avoided. Sticks and stems should be removed from the fruit to keep
them out of the fermentation. Overall, however, the fiuit chosen does not need to be as
perfect as that shipped to market by the farmer. This will allow the farmer to sell the fruit

to the distiller that he would not be able to sell at marketz’ 9.

l9

The fermentation conditions also are important. Temperature control is the most
important condition in the fermentation. The amount and quality of yeast is another
factor that should be considered important in the distillation. Bakers yeast will not give
the same quality to the distillate as Champagne type yeast. The amount of yeast used
should be large enough that the chosen yeast type will be able to compete with undesired
bacteria and yeast during the fermentation?

Distillation conditions are also important to the overall distillate quality. Heating
the mash too quickly can result in baking the mash to the pot of the still. If the mash is
allowed to foam, which can happen with fiuit high in pectin, the trays can become dirty
and loose their effectiveness“. Care must also be taken as to where to take the heads,
hearts and tails cuts from the distillate as the hearts out the desired to be as large as
possible but yet contain as little of the compounds which cause negative aroma and flavor
characteristics as possible.

The dilution of the spirits does not effect the quality of the distillate very much.
However, it is possible to determine if there are imperfections in the distillate. The
addition of water, which should be as pure and free of calcium as possible, can cause the
distillate to become cloudy. A cloudy distillate is not desired, and can occur for many
reasons including terpenes oiling out of solution, or hard water causing the cloudinessg.
The dilution of the spirits must not drop the concentration of the distillate below 40% v/v,
as by law, the fruit spirits must be above this value in order to be called a fruit brandy").

In general the rule for making quality fiuit spirits would be, "low-quality in,
equals low-quality out." In other words if you use quality ingredients and take care of the

fermentation and distillation you might be able to produce good spirits. However, if you

20

use poor ingredients or do not take care in cleaning the still in between runs than you will

not produce desirable fruit spiritsz.

21

2.5

10.

Literature Cited

Walton, S., The New Guide to Spfiirits and Liqueurs, Lorenz Books, New York,
1998.

Berglund, K., CEM 924: Class Notes. Fall Semester 1998, Department of
Chemistry, Michigan State University, East Lansing, MI.

Piggott, J .R.. and Patterson, A. ed, Distilled Beverage Flavour: Recent
Developments. Ellis Horwood Ltd., 1989.

Addapted fi'om: Piggott, J .R.. and Patterson, A. ed., Distilled BeveragFlavour:
Recent Developments. Ellis Horwood Ltd., 1989.

Zimmerli, B., and Schlatter, J., Ethyl Carbamate: Analytical Methodology,
Occurrence, Formation, Biologic Activity, and Risk Assesment. Mutation
Research, 259, 1991.

Methanol, Material Safety and Data Sheet.

The Merck Index, 11th ed. Rahway, New Jersey, 1989.

Cleland, J .G., and Kingsbury, G.l., Multimedia Environmental Goals for
Environmental Assessment. U. S. Environmental Protection Agency: EPA-

600/7-77-136b, E-28, November 1977.

Tanner, H., and Brunner, H.R., Fruit Distillation Today 3rd ed., (English

Translation) Heller Chemical and Administration Society, Germany, 1982.
United States Code of Federal Regulations, Title 27—Alcohol, Tobacco Products
and Firearms, Chapter I, Part 5, Subpart C—Standards of Identity for Distilled

Spirits. <http://www.access.gpo.gov/nara/cfr/cfr-table-search.html>.

22

1 1. Plank, Alexandr, Personal Correspondence with Author. 1998-2001.

23

3. MATERIALS AND METHODS

3.1 Experimental Process
3.1.1 Fermentation

Fermentation of the fruit was carried out in ZOO-gallon homemade fermenters. A
ZOO—gallon HDPE tank was used with a valve was inserted into the side housing of the
tank. A trellis was fashioned out of PVC piping and sealed to keep the mash out of it.
Copper tubing that is used for the cooling water was wound around this trellis in an
attempt for even cooling of the mash. The trellis with copper tubing wound around it was
made to fit inside the ZOO-gallon fermenter. A layer of water-heater insulation was
placed around the outside of the ZOO-gallon tank to decrease heating effects from the
environment. A diagram of the ZOO-gallon fermenter is shown in Figure 3.1.

The top of the fermenter had a notch cut out of it~for the stirring shaft as well as
holes for cooling water flow in and out of the copper coils around the trellis. The notch
was covered with a thin piece of plastic with a hole cut for the stirring shaft. This piece
was taped to the fermenter to prevent insects from getting into the fermentation. A piece
of piping insulation was taped around the stirring shaft also to prevent bugs from getting
in. The mass of the top worked to keep the insects from getting in through the seal
between the top of the fermenter and the top of the ZOO-gallon drum. There were no gas-
lock channels installed in the fermenter as the C02 produced never seemed build up to a
noticeable pressure, and always seemed to release by itself.

The stirring motor had a variable speed control. The shaft had two, three blade

propellers. These were set to mix the mash as much as possible at as low speed as

24

.5893 2: he mam—o8 o...

.5.— comz m_ £6... 2:: U>m a ...—:93 e583 :8 .3:er .5893 5535—50.. :23» :3 .3 quoc otafioaom fin 95w;

    

       

cocooccoo
955
.8 o>_e>

03"; Paco!)-
. {a}
it

.
\S‘. . freakistior... (I (I
I.) o .

 

 

 

(...-5...:K
. . . .\A\.\.AAA§\\ 1:: .i
. $5.15... ....q .
5....) .
£111.13:- SSNI .
.

I I . s ‘ .
. . .. .....(grzrim . .. . \3?)...Annfin .r ..

                 
    

 

 

. \i.
\\ oint- u.\\‘ c \r .

 

; x .S\
. i . “FIE-g3, fie...- .
... . (...-)3 .
\ (I‘VE-r
{..er

. I . 1...... .
- «cot-RN .Sz.

 

 

\

as»:
oaamaom

mzoo wcsooo coaaoo

 

 

x5: ”.753 55% com
Sane.“ Hob—CW

cots“ auton— men.

.328 Each

 

 

 

25

possible. The range of speeds used varied from 60 rpm to 210 rpm. The working range
desired for the better fermentations was 60 to 110 rpm.

With these fermentations a temperature of 15-20° C was desired]. The flow rate
of the cooling water through the fermenters was kept constant at about 50 ml per second.
The water chiller was set to maintain a temperature in the waterbath of 12° C, which
accounted for the initial heat spike associated with fermentations.

For this study, frozen fruit was used. The use of frozen fi'uit does not give the
quality of the distillate that is associated with the use of fresh fiuit, however, the ability to
perform these experiments in the winter and spring months allowed us to use the still
outside of the growing season. The frozen fruit was thawed for 48 hours to room
temperature. A cavity displacement pump was used for the mashing of fruit with small or
no pits (cherries, pears), and a rolling mill was used for the mashing of the filth with
larger stones (peaches, plums). After mashing the fruit was transported to the ZOO-gallon
fermenters, and allowed to. sit for at least an hour such that the initial temperature of the
mash was 17°C. This would allow the cooling water to accommodate the initial heat
spike from the fermentation without exceeding 21°C. These parameters were never
calculated to provide adequate compensation for the heat spike, however, in practice they
did work.

For small fermentations (~10 L) a ratio of 5 g of dry champagne type yeast for 10
L of fruit mash is recommended. Scaling the amount of mash to 200 gallons (756 L)
does not require the same amount of yeast. Only 300g of dry yeast were added to the
mash, where a much larger amount could be expected. The yeast is activated in warm

water for fifteen minutes before addition to the mash. The yeast used in these

26

experiments was Pries-de-mousse, Champagne type yeast (saccharomyces cerevisiae)
purchased from a commercial vendor, Lavlin Inc., in California.

The progress of the fermentation was observed through the measurement of the
sugar concentration in the mash. Using a portable refractometer, the percent solid of the
mash could be measured. This value, also known as the Brix of the mash, gives an
approximate amount of sugar left in the mash. To obtain the results, a sample of mash is
first centrifuged, or filtered through a 0.5 pm filter to remove the non dissolved solids.
The sample is then dripped onto the faceplate of the refractometer. The readout of the
refractometer yields the approximate dissolved solids in the mash. A common value at
the start of fermentation for cherries is 13.5 brix and at the end of fermentation the value
is below 4 brix.

After fermentation, the fiuit mash was pumped in batches into the still such that a
volume of mash in the still was about 165 L. An average of five charges for the still was
taken per fermentation in the ZOO-gallon fermenters. For those types of fiuit that have
stones too large to fit through the cavity of the pump we use, a wet/dry Hoover shop

vacuum was used to transfer the mash from the fermenter to the still.

3.1.2 Distillation

The distillation of fi'uit spirits using the Germanic batch distillation method,
involves many steps. First the construction of the still is an important. The design may
Vary a bit as to the number of trays and the placement of the column of trays, however the
underlying importance is that the still be made out of copper. There is an unknown

Catalytic effect found in copper stills that is not found in stills made of other metals. The

27

basic design to the still we use, constructed by the Christian Carl Ing. GmbH. in
Germany, includes steam heating of the pot, one tray over the pot, and three trays, which
can be allowed to be open or closed, in a column next to the pot, a “catalytic converter”
next to that column, and a condensing column next to that. Two partial condensers are
located in the still. The first is at the top of the first tray, and the second is at the top of
the second column of trays. Figure 3.2 is a schematic of the Christian Carl still utilized in
these experiments.

Two other parts of the still are important in the distillation process. First, a
stirring motor is utilized during the distillation to stir the mash and prevent foaming, as
well as baking to the surface of the still. For more viscous mashes such as apples or
pears, baking can be a problem as there is not uniform heating of the mash occurring.

The second piece of equipment in the still is the “puking tube.” This tube is for the
condensate to return to the pot of the still. It prevents mash from foaming, as well as
allowing the condensate to mix with the mash rather than being immediately vaporized in
contact with the pot of the still.

The cooling water used is cold tap water. There is a cooling water flow regulator,
which is linked to a temperature sensor at the top of the condensing column. The
temperature of the distillate is maintained at 72° C, at the top of the column by regulating
the flow of cooling water through the flow control device7. A manual adjustment can be
made as well in an attempt to increase or decrease the rate of distillate fi'om the end of the
still. The cooling water at the top of the condensing column, which has been effectively

Preheated, then feeds the partial condensers at the top of the trays.

28

55 5

           

owcasofla
03:2me

 

 

«0.:—

1.33
a

3.530

5:. 3 mam—coo

s

.2250 ASH

homeom

6.3809th

3351—

  

/ . ---.

 

 

‘I

 

..oCo>=oO \
och—Sm“

 

 

 

...:m A m3 taU 5332—0 .3 omuafionum

   
            

tom
:30 Box

mEozcm

.25
anam

.......
..........

uohumam

 

 

iiiil‘m amazvscoficom

 

magnum cameo

 

 

 

 

 

 

5am 33>» wE—ooo
£8 5a; Seazzao

u

 

fl ..om coo—.00 Each

.2 8:09.00 333m

IDDDI

«8.0 as 835

 

 

 

Na 9...»:—

29

The distillation begins by filling the pot with the fermented mash. During the fill,
the cooling water is allowed to fill the condensing column and the two partial condensers.
The decision of using the catalytic converter, or opening some of the trays can be made at
this time as well. When the pot is full and the still is ready for operation, the stirring
motor is turned on. The stirring motor prevents baking of the fruit on the copper surface
and can help to prevent foaming of the distillation bottoms up to the trays.

The valve for the cooling water is closed to allow the temperature sensor at the
top of the condensing column that regulates the flow of water into the condensing column
to take over water flow control. At this point the steam is allowed into the steam jacket
surrounding the pot of the still. The steam is slowly brought to a pressure of 0.3 to 0.4
bar. This allows for good heating of the mash, and brings it to a boil in a relatively quick
time of about a half hour7. The distillation is then allowed to proceed inside the still
without interruption until the distillate begins to exit the condenser of the still.

The distillate is mostly ethanol and water. The congeners present in the distillate
come out at different times in the distillation process. There are three basic cuts of
distillate taken in the distillation process. The first cut is the heads portion of the
distillate, which contain congeners with relatively low boiling points. The second cut is
the hearts cut, which is the portion that will eventually be consumed. The last out is the
tails, which contains all the heavier congeners and those with relatively higher boiling
points. Most notably in the tails portion of the distillate are the fusel alcohols, which are
known for their unpleasant aromal.

For the experiments in this thesis, the amount of distillate collected per cut was

kept consistent. Three 400 ml cuts were taken for the heads portion of the distillate, and

30

increments of 1000 ml were taken for the hearts portion until the concentration was
below 60 % A.B.V. The final cut of hearts varied in volume based upon the distiller’s
analysis of the spirits at the time of distillation. The tails cut was taken at a volume of
1000 ml. Any further alcohol collected went into a communal tails carboy for

redistillation, which yields a product with poorer aromatic qualities.

3.1.3 Storage

The spirits are stored in glass jars at high proof. They can be blended before
storage, however, the practice was to save each cut separately through the storage process
of the fruits and blend them three to six months after distillation. As it is theorized that
the formation of ethyl carbamate is a light catalyzed process, the distillate is stored in the
dark.

Reactions that occur in the aging process include oxidation, esterification, and
actealization. Oxidative processes involve the conversion of aldehydes to acids, and
alcohols to aldehydes and water. Esterification reactions involve the addition of alcohols
and acids to form esters and water. The most common esterification reaction involves the
addition of ethanol and acetic acid to form ethyl acetate. Acetalization reactions involve
the addition of an aldehyde and alcohol to form diacetals and water. All of the stoage
reactions are dependent on the concentration of the reactants and the temperature and pH
of the stored distillate. Most of these reactions are reversible and do not lead to complete

yield conversion of any of the reactants to the products‘.

31

3.2 Analysis of Distilled Spirits

Ethyl alcohol concentrations were measured by a tralles (% alcohol as well as
proof) hydrdmeter with an internal thermometer to perform instantaneous adjustments to
the actual alcohol concentration]. A sample from each cut was analyzed by gas
chromatography. A Shimadzu GC-17A gas chromatograph with flame ionization
detection was used for analysis of all samples. A Stabilwax 30m column with an inner
diameter of 0.25mm from Restek was used for all measurements. The initial conditions
for the chromatographic analysis were column temperature at 40°C, injector port
temperature of 240°C and detector temperature at 255°C. The temperature program used
for the analysis of the components in the eau-de-vie spirits was initially at 40°C and
ramped at 2.5°C/min until a temperature of 190°C was reached. The temperature was
held at 190°C for 5 minutes. A Shimadzu autosampler (AOC-20i) with a twelve vial
capacity was used to obtain reproducibility of the 0.2 LIL injection volume. Each sample
took 75 minutes to run due to sampling time, cooling time and run time; therefore, one
distillation containing eight different cuts would take 10 hours for complete analysis.
Figure 3.3, is a sample chromatograph and Table 3.1 shows the relative retention times of
the most common compounds in the fruit spirits, as well as their respective boiling points.

Concentrations of each congener were compared to ethanol, which was used as an
internal standard. The concentrations of these congeners were also compared to a
calibration curve for the concentration of the congener compared to the concentration of
ethanol. Use of the autosampler allowed for retention times for each standard to be
reproducible with slight variance due to the concentration difference of the ethanol water

mixture from one sample to the next. As the concentration of the ethanol water matrix

32

changes, the retention time of the lower boiling congeners varied by less than half a

minute, however the higher boiling compounds could vary by as much as a minute.

3.3 Tray Usage

The number of trays used during the distillation of the fi‘uit spirits as well
as the use of the catalytic converter was varied throughout the experiment. The
distillations were not run in any set pattern, however, it should be noted that an attempt
was made to get in one distillation with each distillation condition per fermentation batch
in an attempt to minimize the amount of deviation due to starting material. The operating
conditions for the distillations were; all trays closed with the catalytic converter utilized,
all trays closed with the catalytic converter bypassed, one tray open with the catalytic
converter utilized, one tray open with the catalytic converter bypassed, and two trays
open with the catalytic converter bypassed. Because not every fermentation gave five or
six distillations, the last operating condition, two trays open with catalytic converter
bypassed, was not run very often. Instead an attempt to replicate one of the other

operating conditions was often chosen instead of the last condition.

33

Table 3.1. Average retention times, from the chromatographic conditions used,
for the most common congeners present in distilled fruit spirits. The

Boiling points for each congener were also given3.

 

 

 

Chemical Name RetentionTime (min) Boiling point (deg. C)
Acetaldehyde 2.5 20.8
Ethyl Fonnate 3.6 54.0
Ethyl Acetate 4.5 77.0
Methanol 4.7 64.7
Ethanol 5.8 78.0
1-Propanol 8.7 97.0
2-MethyI-2-propanol 10.8 82.4
lsopentanol

(isoamyl alcohol) 16.3 132.0
Benzaldehyde 33.0 179.0

 

34

 

0.005

0.004
0.003

0.002

0.001

(I)

0.000
-0.001

-0.002

Figure 3.3.

 

 

 

 

 

 

 

 

 

 

F

O

is

‘O

'1
b.
In
'1
9'

0

09

F“.

O

H
O
2?
. O b
N an 23 E2 hflflh
n ‘n '0 C 30538
g 2? or: a; '1.
'_l_w,_2 " ‘1 8888
6th
no:
raw.
nos

25. 50
Minutes

Sample Chromatagraph from a cherry eau-de-vie distillation. A
Restek 30m Stabilwax column with 0.25mm i.d. was used for analysis.
The initial conditions had the column temperature at 40°C, injector

port temperature at 240°C, and detector temperature of 255°C. The

analysis included a ramp from 40°C to 190°C at 2.5°C/min.

35

3.4

Literature Cited

Tanner, H., and Brunner, H.R., Fruit Distillation Today 3rd ed., (English

Translation) Heller Chemical and Administration Society, Germany, 1982.
Plank, Alexandr. Personal Correspondence with Author. 1998-2001.

The Merck Index, 11th ed. Rahway, New Jersey, 1989.

36

4. RESULTS AND DISCUSSION

4.1 Composition of Fruit Brandy

The composition of fruit brandy is complex with hundreds of compounds
included in the makeup of a good fi'uit brandy. With the high number of congeners
involved in the distillation process, it is difficult to choose which ones to study. For the
purposes of this experiment, the compounds chosen were those present in the highest
concentration ranges. The compounds studied involve those present in higher
concentration in the heads, those in higher concentration in the hearts, and those
compounds that have profiles that are not higher in the heads, hearts or tails. The
compounds studied in this experiment include; ethanol, acetaldehyde, ethyl formate, ethyl
acetate, l-propanol, 2-methyl-2-propanol, isoamyl alcohol, benzaldehyde, and methanol.
A final section will detail how the concentration of these compounds would relate to a
bottle of spirits at consumable strength (40% ethanol).

The amount of each congener in the distillate will vary widely from fiuit to fruit
and can vary based upon fermentation cbnditions as well. Each congener is profiled to
show how the concentration varies at the distillation progresses. The amount of each
congener that is found in each distillate is converted from the concentration as found in
the high proof distillate to the concentration that would be found in the drinkable strength
distillate. The dilution of the ethanol is the only effect taken into account, and the effect
of storage reactions on the increase or decrease of the concentration of the congener is

ignored.

37

Finally when looking at the concentration of distillate in each bottle of spirits, the
hearts portion is the only portion of the distillate used in calculating the amount of
congener present in the distillate.

The compounds that have lower boiling point than ethanol includes; acetaldehyde,
ethyl formate, ethyl acetate, and methanol. The compounds that have higher boiling
points than ethanol include, l-propanol, 2-methyl-2-propanol, isoamyl alcohol, and
benzaldehyde. These compounds represent the congeners with the highest concentration

throughout the fruit used in these experiments.

4.2 Batch Reproducibility

A large variance is introduced in to the concentration of each congener from one
distillation to the next. One batch of fermentations will not have exactly the same
amount of each compound present in it with variability coming in fruit type, fruit quality,
processing methods, yeast used, temperature and pH effects]. The distillations would
also have an effect not just from the number of trays used but also in which distillation
was being performed. The first distillation in the morning would take a little longer to
get to temperature than the last distillation of the day.

In an attempt to minimize the differences from the many distillations and
fermentations, the data shown will include only those distillations involving plums. The
profiles for each of the compounds studied is similar in the distillate of each fi'uit we have
studied; cherry, peach, pear, plum, with differences in concentration, but not overall
trends. The plum fermentations involved two separate fermentations with eleven total

distillations. Table 4.1 illustrates the distillations involved in the collection of the data

38

 

Table 4.1 Fermentation and distillation conditions. The conditions used for
each distillation and how they relate to the fermentation batches of
the fruit used. The distillation conditions used include: (1) All trays
used and catalytic converter engaged, (2) all trays used with no
catalytic converter, (3) two trays used with catalytic converter

engaged, (4) two trays used with no catalytic converter.

 

 

 

 

 

 

 

 

 

Fruit Fermentation # Run # Trays Used CC Used?
Cherry 1 1 All Yes
Cherry 1 2 All Yes
Cherry 1 3 All Yes
Cherry 1 4 All Yes
Cherry 1 5 All Yes
Peach 1 1 All Yes
Peach 1 2 All No
Peach 1 3 2 No
Peach 1 4 2 Yes

Plum 1 1 All - No

Plum 1 2 2 No

Plum 1 3 2 Yes

Plum 1 4 All Yes

Plum 2 1 2 Yes
Plum 2 2 All No

Plum 2 3 2 No

Plum 2 4 All Yes

Plum 4 1 All No
Plum 4 2 All Yes
Plum 4 3 2 Yes

Note:
Plum fermetnation (3) was lost due to pump
failure in the cooling water system

 

39

for these experiments. As the profile of each congener was desired, and the concentration
range for each distillation could be greatly different fiom distillation to distillation, the
runs were averaged where possible to show the trend. Other cases where the
concentration ranges deviated, the profile that showed the general trend for a particular
congener throughout all the fruit distilled in this study, was used. Error bars were not
included in the figures because there was a wide variety in concentration from one
distillation to the next.

Attempts were made to run as many distillations as possible from one batch in a
day. This was to minimize the amount of extra fermentation time the last batch distilled
would have. Rinsing the sides of the fermenter after charging the still in order to remove
any of the mash on to the sides diluted the rest of the mash, and was kept to a minimum.
Upon completion of fermentation, the fermenter was completely cleaned with a dilute
bleach solution to remove any excess yeast or mash stuck to the trellis or sides of the still.

Cleaning the still between distillations was completed with a water wash. The
cleaning of the trays involved the use of nozzles inside the still at each tray position.
Changing the flow of cooling water through the piping around the still allowed the
cooling water to flow to these nozzles where they rinsed the inside of the still. The
cooling water flowed to the pot where it mixed with the hot mash and cooled it as well.
Where visible fouling or excess mash adhered to the inside of the still was present, a
damp piece of cheesecloth was used to remove the dirt. The cleaning of the pot of the still
involves use of cold water on the inside of the pot of the still, after the bottoms of the

distillate were removed.

40

Care was taken to be as reproducible as possible with the distillations. One batch
of cherry distillations shows how unpredictable the distillation reproducibility can be.
Five distillations run under the same conditions produce five different ethanol profiles.
Figure 4.1 illustrates this data. The five distillations were all run on the same day, from
the same fermentation batch under the same distillation conditions. The amount of
ethanol in each distillate varied through uncontrollable parameters and each run was kept
as constant as possible. Variety of this sort is even more common in the congeners,

although every attempt was made to prevent variability from distillation to distillation.

4.3 Distillate Analysis

Analysis of the distillate was carried out by GC-F ID. The retention times of the
peaks we compared to the retention times of standards for each congener. Standards
were also used to develop a calibration curve for each congener. The amount of ethanol
in each distillate was measured by hydrometer, and corrected for temperature effects. As
the distillate is diluted after storage and before consumption, the concentration was set to
the ratio of ethanol in distillate and a 40% ethanol distillate. All data shown for each
congener were calculated to show the concentration as they would appear in a 40%
ethanol brandy. The effect of storage on the concentration of the congeners was not
taken into account. The concentration of esters would probably increase during storage,
however, because the storage reactions are equilibrium reactions, there is no way of
knowing if the concentration of each congener would increase or decrease during the

storage.

41

...o? as no»: gov—958 9:253 2: .28 com: 9?... 2: =a .53 :8 62....8 9.8.3 2.33.52:
=< 63.: 2859.8 5:32... 2:3 2: 5.3 E... 0.83 .23 :22— :emuauaogou 2:3 2: =8...— ouoB 26:255.

=< 85:255. 35% E23 ..8 25:?» 2.358% 023353 ..e 5:25.— a ma 5:538:89 3:2:H :3 0.5!:

3.5 eE=_o> 32:85
88 88 88 88 88 88 88 o

p _ _ b F _ i_i

 

0
L0

 

I
/

A
O
(O

 

 

 

O 0
co Ix
(II/A (yo) uorrenuaouoo roueqtg

 

 

 

 

 

 

 

l
l
l
l
8

42

The number of trays used in the distillation process were varied from run to run,
as was the use of the catalytic converter. The four types of distillations that were run
include; all trays engaged (in use) with the catalytic converter also used, all trays engaged
with the catalytic converter bypassed, one tray open (two engaged) with the catalytic
converter used, and two trays engaged (one left open) with the catalytic converter
bypassed. The effect of tray number on ethanol and the congeners as well as the effect
the catalytic converter would have on the compounds studied in the distillate were

studied.

4.4 Effect of Tray Usage
4. 4. 1 Ethanol

The end goal in the distillation of fruit spirits is to maximize the number of bottles
of 40% ethanol that can be made with the highest quality. These spirits are usually
bottled in 375 mL bottles. As each distillation contained the same amount of mash, and
hopefiilly the same amount of ethanol (if the yeast did its job correctly). Overhead
product was collected with 1200 mL of heads, (in three 400 mL volumes) and volumes of
1000 mL of hearts until a concentration of < 65% alcohol was achieved. This varied the
amount of hearts collected in each distillation as some batches gave 3000 mL of hearts
and others gave 5000 ml of hearts. A final 1000 mL was collected as tails.
The profile of ethanol concentration throughout the distillation of fiuit spirits, begins at a
concentration of about 84% ethanol. The concentration increases until a concentration
between 87% and 90% is reached. The concentration then slowly decreases for the next

1000 to 4000 mL. This portion constituted the largest portion of the hearts of the

43

distillate. The concentration will then decrease at a faster rate until the distillation is
complete. The amount of distillate collected is dependent upon the parameters the
distiller wants to follow.

Figure 4.2 shows the average concentration of ethanol per distillation volume,
with varying distillation conditions. Figure 4.3 shows how the average runs in Figure 4.2
would translate into bottles of 40% fruit spirits. Figure 4.3 shows that the more trays that
are engaged, the higher amount of 40% spirits will be collected.

The distillate collected with all the trays engaged and with the catalytic converter
used also, had the largest portion of distillate collected. The amount of distillate collected
with three trays but with the catalytic converter bypassed included the second largest
volume of distillate. This trend follows as would be expected, and corroborates the
theory that the greater number of trays will produce a larger volume of higher

concentration ethanol distillate.

4. 4. 2 Acetaldehyde

Acetaldehyde is the main component emerging before methanol in the distillation
of cherry spirits. It has a distinctive aroma characteristic, which can cause an eau-de-vie
to have a poor aromatic characteristic when present in too high of a concentrationz. As
acetaldehyde has a low boiling point, its largest concentration is distilled into the heads
portion of the distillate. However, when fewer trays are used a larger concentration
proceeds into the hearts portion of the distillate. Acetaldehyde is completely soluble in
both water and ethanol; therefore it is not easily separated from the spirits. The

acetaldehyde does stay in the distillate even though it has such a low boiling point.

44

82.923 6:233 2. .:_B :3... 9?... e3: 9; 609%.... ..2..o>..8 omen—33 .:_3 com: 93.: 95 A3 £23.58

3925.8 2. .53 we... 9?... =a AS ...owawg .2338 2.5.3.8 E... to»: 9?: =< A: 8:23—58 5.358%

wit: .:_3 2.5—E. 2259—. 3:23.59 .... 5:25.— .. 2. At.» .XL .328 .... 5:32.359 «was: 2E.

 

TV in... mlal w ror 2+4

 

3.5 2.2.5 e.e___.m_oe>_.e_=§_o

0000 000..» OOON

r _

 

 

Yi.ii iriiil

 

 

 

 

 

om

O O
I\ (0
(MA %) 'auoo 101191113

0
(D

om

hesifi

45

...o...o>..8 23.5.8 2. ......» to... 9?... 9s. A... 609%..» ..o...o>..8 firm—«.8 ......» can: 9?... 25.
Am. ...o...o>..8 2.25.8 2. .53 to... 2?... =a A". .uowaws 8.82.8 3.258 ...... new... 9?... =< A: 5.83:...8

witty? 2.0.5.5.. ...... 9:. ... .35....— Eai €5.59 $3. .... 8......— o.:_ N... 9...»:— ... 5a.. 2.. ..o Eta—mafia.

 

$5.0m .>.m.< .xév. .0 ..mnEsz
om mm ON 9. or m o

 

 

 

 

 

...... 9...»...

suompuog BuneJado

46

A plot of the acetaldehyde profile can be found in Figure 4.4. It should be noted
that unlike other congeners that have a boiling point less than ethanol, there is
acetaldehyde present in all parts of the distillation. The amount of acetaldehyde does
decrease as the distillation continues, however, the concentration does not reach zero in

any portion of the distillation.

4. 4. 3 Ethyl formate

Unlike acetaldehyde, ethyl formate does decrease to a point where we were
unable to measure it in the distillate. Again, the largest portion of ethyl formate was
present in the heads portion of the distillate, which would be expected for those
compounds whose boiling points are less than that of ethanol. Figure 4.5 shows the
profile of ethyl formate in the distillations. When all the trays are engaged and the
catalytic converter is used, the amount of ethyl formate in the distillate is the lowest
amount, when compared to a distillation using only two trays in which a greater amount
ethyl formate could be found.

Ethyl formate is an ester that can be expected to form during the storage of the
distillate. It is not the most common ester in the distillate, which would be ethyl acetate,
however it is present in high enough concentration that we were able to detect it in the
heads and a small part of the hearts of the distillation. Ethyl acetate, like most esters, can
be considered a positive flavor component in the distillate at lower concentrations,
however, if the concentration of ethyl acetate gets to be too large, the quality of the aroma

of the fruit spirits will decrease.

47

.8.8>..8 8.2.8.8
.... ......» .88 9?... ...... A... .8898 8.8.5.8 8.3.38 ....3 .82. was... 95 Am. .8.8>..8 8.3.58 .... ....B .8...
m8... .... AN. .889... 8.82.8 8.3.3.3 ...... .88 .8... ..< A: 68......8. 8.88.... 288:... ......— ... 82...“...

.8 2.5.2. 3.8.58.8 .... 8.8.8.. a .... 9......8... =5... ... 8.5.82.8... .... 8.8.2.880 2.8.... .....Eo...a.8<

 

v ...ui mlal N no: ..Iol

 

 

3.5 ..E:_o> 22:55 «2.22.50

 

 

 

l
C”.
o

 

 

(D
o'
(A[A%) 'uouoo apllqapmaov

 

 

 

 

0’.
o

...... 2......

48

28.828. 3.28....
.... ....2 .82. 9...... 82. A... ..8waw8 8.82.8 3.258 ....2 .82. 9?... 82. G. 28828. 3.28.8 .... ......» .82.
9...... ..a An. .8828 8.82.8 3.2.2.... .28 .82. 9...... ..< A: 68.3.2.8 w...2.8.... .88....... 2.8.. 2. o.2...2..

.8 88......» «2.2.2.28 .8 8.8.2... 8 .... 8.2.2.... 2...... ... 82.88.. .2... .... 8.8.2.8....U 2.8.... 82.88.. .2...— m... 82%....

 

..-u-m|+lw-.-.++n

 

 

...E. 25...; 32.85 82.235
comm come 00 .N o

E h } 3L
‘ t‘ ‘

 

 

 

 

 

wood

 

 

 

 

 

 

(IVA %) 'uouoo aneuuo; Mu};

 

N5...

49

4. 4.4 Ethyl acetate

After acetaldehyde, ethyl acetate is the most abundant of the minor components of
cherry distillates boiling before methanolz. The ethyl acetate distillation profile acts in a
similar manner as the acetaldehyde and ethyl formate. There is a higher concentration of
ethyl acetate in the heads and early hearts portion of the distillate, than other portions of
the distillate. The amount of ethyl acetate in the late hearts and tails increases only in
those distillations using the catalytic converters. The catalytic converter may act as a
surface for esterification reactions to occur as the concentration of ethanol decreases.

Figure 4.6 illustrates this point, and shows the profile of ethyl acetate in the distillate.

4. 4. 5 I-Propanol

Fusel alcohols compose the largest group of aroma compounds in alcoholic
beverages?“ 4. l-Propanol is the simplest of the fusel alcohols, and it has the lowest
boiling point of all the fusel alcohols. The amount of 1-propanol and isoamyl alcohol can
be considered as an indicator of the total amount of fiJsel alcohols in the distillate.
Because Fusel alcohols donate an unpleasant, damp cloth aroma to the distillate,
minimizing the amount of filsel alcohols in the distillate is a goal of every distillation.

The concentration of l-propanol in the distillate increases gradually, reaches a
maximum, a little past half way through the total hearts collected, and then decreases.
Figure 4.7 illustrates this profile of l-propanol. The amount of l-propanol in the cuts
reaches a maximum in the hearts portion of the distillate. The only ways to effectively
remove the higher concentrations of 1-propanol from the distillate would be to make the

tails cut earlier in the distillation, or redistill the collected distillate. Redistillation, would

50

......82..8 3.28.8 .... ....2

.82. 99.... ...... A... .8898 8.82.8 3.28.8 ....2 .82. 9...... ...... .9 .8.82.8 3.28.8 .... ......» .82. 9......

.... AN. ..8w..w.8 8.82.8 3.25.8 ...... .82. 9...... ..< ... 28.3.2.8 8.2.8.... 288...... ...... 8...... 82...»...

.... 88......» 82.2.5.8 .... 8.8.2... a 2. 8.2.2.... 8:... ... 8.38.. .2... .... 8.222888 2.... .8.......... 82.8.... ......H

oomo

 

 

 

winnmlslmnou.l+l

 

 

3.... ......._o> 32.29.. 3:2..an
oomv 005

I I
I ‘
l< - I
<

 

 

J?
I

‘f
".
O

‘3‘!
0

0°.
0

 

V
,4.
o

(A/A%)'uouoo atmaov Mina

 

 

‘0.
o

.3. 2......

51

...882.8 3.288 .... ....2 .82. 9.8... ...... A... ..8waw8 8.82.8 3.288 ....2 .82. 9...... ...... A... 2882.8
3.288 .... ....2 .82. 9.8. .... .8 .8828 8.82.8 3.28.8 ...... .82. 9...... ..< ... 2.8.3.2.... 2388....

288...... ...... 82:8... .... 8.2.8.. 82.2.2.8. .... 8.8:... a 2. 8.2.2.... 8:... ... 8.2.8.888 8.8.8.... ...... ......w...

 

vuuumlal mic: .lol

 

 

 

3.... 2.2.3 32...»... 9:23....

Comm oomv 00 .N o
. . moo

 

 

 

 

mNo

 

 

 

(A/A%) °uouoo |ouedord-|.

 

 

 

mod

52

reduce the concentration of the other flavor components and cause a loss of yield, making

the l-propanol a difficult compound to reduce in the distillate.

4. 4. 6 2-Methyl—2-pr0panol

The profile of 2-methyl-2-propanol, which can be found in Figure 4.8, is similar
to the concentration profile of the other fusel alcohols, 1-propanol and isoamyl alcohol.
The amount of 2-methyl-2-propanol is less than one third of the concentration of 1-
propanol, or isoamyl alcohol in the spirits however. The catalytic converter seems to
bring the maximum concentration point of the profile to a fraction earlier than that of the
distillations bypassing the catalytic converter. The 2-methyl-2-propanol has a retention
time between l-propanol and isoamyl alcohol. The amount of 2-methyl-2-propanol was

greater in the plum distillates than cherry or peach distillates.

4. 4. 7 Isoamyl alcohol (isopentanol, 3-methyl-1-butanol)

Isoamyl alcohol (3 -methyl-l-butanol) is the main fusel alcohol synthesized during
fermentation by yeast, with 1-propanol considered another important fusel alcohol.
However, these two compounds distill earlier than most of the other fusel alcohols and
therefore they are considered to be early indicators of the balance of the remaining fusel
alcohols. The fiJsel alcohols produce a “damp cloth” aroma that is undesirable in the
production of spiritss. 1-propanol, and isoamyl alcohol are present in the hearts in small
quantities, and increase in concentration at the end of the hearts cut and begin to decrease
in the tails. Most fusel alcohols will be present in greater concentrations in the tails than

in the hearts. As these two compounds are precursors to the other fiisel alcohols

53

.8.82..8 3.2...3 .... ......» .82. .2... ..2.
.9. 2893.8 8.82.3 3.2.33 ......» .82. «2.... 92. Am. .8.82..8 3.2.33 .... ....2 .82. m2... .... AN. ..8w..w.8
8.82:8 3.233 ...... .82. 2...... =4 ... 2.33.2.3 2......er 388...... ...... ... 8......8... .... 2.5....» 82.2.5.3

.... 538...... a 2. o........... :5... ... _o=a..e...-m-_.m._.o...-~ .... ......3..83:.U ...—......— _o:.........-~.._2.82..u

 

Tlnlmlalmlol .+_

 

...E. «52.3 82...»... 2.5353

comm OONV 00 ..N o
r F o

 

 

 

 

(It/A
%) 'uauoo IouedOJd-z-Mutaw-z

 

 

 

 

 

 

To

a... 9...»...

54

 

.8.82.8 3.233 .... :32
.82. .2... e2. .... 3.83.8 8.82.8 3.233 ......» .82. 2...... ...... .8 2382.8 3.233 .... ......» .82. 2.3.
.... AN. ...owaw... 8.82.8 3.23.3. ...... .82. 2...... =< ... 2.33.3.8 8.8.3.2.. .883... ...... 8...... 3.3—......

.... 2.5....» 82......an .... 238...... a as 3.3—...... 2...... ... 53.3.88... .....8... .288. ...—3...... .....8... .2333

m IDI mulcnl N I?! will;

 

 

3.5 8.2.5 32:8... «2.2:an

comm OONV oo rm 0
. L O

 

 

 

mod

(A/A %)
'uouoo |oqoo|v Mtueosl

 

 

 

l 0.20

 

 

 

 

VNd

a... 25w...

55

measuring the presence of l-propanol or isoamyl alcohol on line would allow for an
easier determination of where to make the hearts/tails cut.

A plot of the concentration of isoamyl alcohol can be seen in Figure 4.9. Again
the catalytic converter shifts the maximum of the concentration to an earlier volume than
when the catalytic converter is bypassed. As this trend can also be seen in 2-methyl-2-
propanol, and to a lesser extent in l-propanol, the catalytic converter may be
concentrating the filsel alcohols in the hearts, rather than the tails of the spirits. As the
catalytic converter acts as an additional tray causing the fiisel alcohols with lower boiling
points to emerge earlier in the distillation due to the increased separation of all the

components present.

4. 4. 8 Benzaldehyde

Benzaldehyde is a positive flavor component of fruit distillate. Benzaldehyde has
many trade names including, "the essence of cherry, " and "bitter almond oil." The
amount of benzaldehyde may be loosely related to the concentration of ethyl carbamate
in the spirits, asthe molar amounts of cyanide and benzaldehyde produced in the
hydrolysis of amygdalin in the fermentation are equivalent. The profile of the
benzaldehyde in fruit spirits can be seen in Figure 4.10. Benzaldehyde does not appear in
similar concentrations in pear spirits, as there is no stone with amygdalin to produce the
benzaldehyde. The higher number of trays in the distillation pushed the benzaldehyde
fithher towards the tails portion of the distillate. A noticeable shift in concentration
maximum occurs between two trays with catalytic converter and the three trays. The

catalytic converter seemed to have little effect between the run with all trays with the

56

«3202.8 3333 e: .53
tom: 93.: 95 A3 609%.; 32258 gum—53 :33 tom: 9?: 95 A3 ...3..o>=8 aux—33 e: .53 tom: 93.:

=a AS .uowawfi. 58.5.5.3 9.95.33 ES can: 9?... :41: £32.38 cows—=25 Engage .58 .33.: 35:25

.3 25:?» 9.52555 ..e 552:: a an 82:35 :5... 5 3:55:35» vague—annom 2:9:— ocmsoEanaom 3.9 9:.»5

 

flwlmlmlll.mlolrlél.

 

 

3.5 ¢E=.o> 32:65 955355
comm oowv 005 o

 

 

 

 

 

 

 

 

 

00.0

(NA %) 'uouoo apliuaplezuaa

57

catalytic converter, and the distillation with only three trays used without the catalytic

converter. Currently there is not a demand for the regulation of benzaldehyde, however,
if a relationship between the concentration of benzaldehyde and the concentration of the
ethyl carbamate in the spirits can be found, it may be possible to use that relationship to

regulate the ethyl carbamate through analysis of benzaldehyde.

4. 4.9 Methanol

Methanol is a health hazard that is regulated by the FDA at 0.35% (v/v) in
distilled spiritsé. The amount of methanol we produce in the hearts of our spirits is just
above that level. The methanol profile exists such that the methanol concentration is
higher in the heads, drops in the hearts to a lower concentration, and then increases again
in the tails. Of the compounds examined, methanol has a unique distillation profile.
Figure 4.11 shows the methanol concentration curves per cumulative distillation volume
for the operating conditions used. The distillation profile of methanol in this type of
multistage batch distillation is different than it would be for an alambic style distillation7.
It is important to test each cut to minimize the methanol concentration in the hearts cut.
Methanol is considered by many to be a positive flavor constituent in distilled spirits. It
is difficult to separate the methanol from the ethanol water mixture further without loss of
many of the other flavor components from the distillate. Treating the fermentation mash,
with some methanol inhibitor is probably the easiest way to minimize the amount of
methanol in the distillate. The concentration must be below 0.35% (v/v) but it is
desirable to have the concentration close to regulatory limit, as it is a positive flavor

component.

58

£83289 £3.83 2. 5.3 can: 93.: 95 A3 .uowaws 3202.8 £3.58 .53 can: 93.: 95 A3 cor—92.8

33.53 c: :33 to»: was... =a ANV .3330 3292.3 cum—83 ...—a com: 93.: =< A: 6.5528 news—=25

Each—G .82.: 32:35 be 25:?» «22:25.9 .3 .3525.— a us «82:25. :5..— E .2352: .3 53333.89 2E. .— fie 25w:—

00mm

 

 

Vlnlmlfilmlolrl+lfi

3.5 aE=_o> 32:55 352385

 

 

 

 

 

 

 

 

89. SE 0
_ . 8.0
W
a
1.-
U-
- l. . w
80 w
0
O
u
0
.u
l - m3 )
%
M
M.

 

 

 

mmd

59

4.4.10 Bottles of fruit spirits

Figure 4.12 shows the concentration of acetaldehyde, ethyl acetate, methanol, 1-
propanol, isoamyl alcohol, and benzaldehyde, as they would exist when the distillate was
diluted to 40% ethanol. A higher number of trays produce a lower concentration of
acetaldehyde in the distillate, although the catalytic converter seems to increase the
amount of acetaldehyde in the hearts of the distillate. The isoamyl alcohol shows the
effect of the trays on the congeners with boiling points greater than ethanol. The more
trays that are used results in a reduced concentration of congener. This is usefill for

controlling the filsel alcohols as they have an undesired aroma characteristic.

4.5 Catalytic Converter Use

The use of the catalytic converter as studied in this experiment, had a number of
different trends. First, in the earlier stages of the distillation, the catalytic converter,
acting as an additional stage, decreased the amount of esters that would be found in the
hearts by increasing the amount in the heads of the distillate. The effect on aldehydes in
the earlier portion of the distillation was not very clear, and there seems to be no effect or
trend that was apparent. As methanol is the only alcohol that is present before the
ethanol, the catalytic converter had the effect of reducing the concentration present in the
distillate.

Later in the distillation the catalytic converter had the effect of shifting the
maximum concentration range of the fusel alcohols to an earlier volume than that of the
distillation run with similar conditions without the catalytic converter. The catalytic

converter also caused increased ester formation in the tails of the distillate.

6O

..8 3.... £5 ......» 8.5.2.. 2. 8 :55 2.. was» 888.. ...—.8 .... own...— 8.22.838
2.... £88.58 9.3.58 2. ......» .82. as... 95 A... .83?» 8.82.8 firm—38 5.3 can: 9?... 95 An.
£2838 9.2238 c: 5.3 .83 99.... .... Am. .8395 8292.8 9.3.58 ...... .8»: 99.... =< A: .38 3.32. ....

8.2:... ...... 9:68... .3 .8269... e<> .2.an $3. 83.2.58 Emit... use—...... :. 5.2.2.838 .68ng .2... 25w...—

 

 

 

 

vl m mg _.l
..mcmmcoo
6:38..

02.5...ch 6:8... _>Emom_ -N- _ocm50_2& 6:889. 85.522 0.38.. 35m «35230.2

 

 

 

l
".
O

 

 

 

 

 

 

or; N
O 0
(MA °/,,) 'uouoa Jauafiuoo

.....
.........

. . . .
.....n. .....
......

ll r v.0

 

 

iii ii ii

61

The concentration of ethanol was also effected by the catalytic converter. The
ethanol concentration, and volume of USCfiJl alcohol were greater with all the trays
engaged and the catalytic converter used. The amount of alcohol, both volume and
concentration, were lowest in the run without the catalytic converter and only two trays
used. The catalytic converter seems to resemble an additional tray in the distillation,
however, the amount of increased separation caused by the catalytic converter is not
equal to that of an additional tray. The run with all trays used, and no catalytic converter
had more separation of the alcohol from the congeners, than the run with only two trays

used while still using the catalytic converter.

62

4.6

Literature Cited

Tanner, H., and Brunner, H.R., Fruit Distillation Today 3rd ed.. (English

Translation) Heller Chemical and Administration Society, Germany, 1982.

Stinson, E.E., et al., Composition of Montmorency Cherry Essence. 1. Low-

boiling Components. Journal of Food Science, 34, 1969.

Stinson, E.E., et al., Composition of Montmorency Cherry Essence. 2. High-
boiling Components. Journal of Food Science, 34, 1969.

Guymon, J .F., Higher Alcohols in Beverage Brandy: Feasibility of Control of
Levels. Wines and Vines, 53, (l) 1972.

Berglund, K., CEM 924: Class Notes. Fall Semester 1998, Department of
Chemistry, Michigan State University, East Lansing, MI.

United States Code of Federal Regulations, Title 27—Alcohol, Tobacco Products
and Firearms, Chapter I, Part 5, Subpart C—Standards of Identity for Distilled

Spirits. <http://www.access.gpo.gov/nara/cfr/cfr-table-search.html>

Suomalainen, H. and Lehtonen, M., The Production of Aroma Compounds by
Yeast. J. Inst. Brew., 85, 1979.

63

5. SUMMARY AND CONCLUSIONS

The distillation of fruit beverages has been occurring since ancient times, and
there is still a section of the process that is not fiilly understood. The number of trays
used in distillation has been known to cause better separation of the components in the
distillate with more trays used, however, the effect on the congener compounds has not
been fully studied. If any attempt is to be made to develop a process for in situ analysis
of the distillation of fruit spirits, the happenings of the congeners in the distillate must be
more fully understood.

The variation of the distillation conditions involved in this study, shows that each
congener is affected by the number of trays used in the distillation, as well as the catalytic
converter. Changing the number of trays can increase or decrease the amount of
congener in the hearts portion of the distillate. This study has shown that using the
maximum allowable number of trays in out still, the spirits will be less likely to contain
high concentrations of the undesired congeners, while not removing too much of the
desired congeners.

Maximizing the number of trays used on the still has yielded a larger volume of
higher concentration ethanol. This will translate into more bottles of spirits for sale at
market and therefore an increase in profits for the distiller. As mentioned earlier,
maximizing ethanol production is just one goal of the distiller, because, if the ethanol is
not of a good quality, the consumer will be less likely to purchase the spirits. To see how
the quality of the fruit spirits could be improved, the concentration variance of the

congener compounds was explored.

64

Low boiling compounds such as acetaldehyde, ethyl formate, and ethyl acetate,
which are present in the distillate as products of fermentation conditions, and
esterification reactions in the distillation process, are present in higher concentration in
the earlier volumes of the distillate. Acetaldehyde is present throughout the distillate,
however, using more trays in the distillation has the effect of having lower concentrations
of acetaldehyde in the distillate. While acetaldehyde is thought to contribute to the
headache associated with hangover, reducing the concentration of acetaldehyde can make
for a more enjoyable drinking experience. The catalytic converter, however, while it at
times acts as an additional tray, tended to increase the concentration of the distillate in the
fiuit spirits when compared with using the trays without the catalytic converter. The
same effect is seen with benzaldehyde, so the catalytic converter can be thought to
increase the reactions producing aldehydes in the distillation process.

The esters studied, ethyl formate, and ethyl acetate, were also low boiling
compounds present in higher concentration in the earlier volume of the distillation
process. The amount of these esters in the distillate decreased greatly over the heads and
first hearts out of the distillate. The amount of distillate collected as hearts, and where the
heads/hearts cut is made will have a large effect on the overall concentration of the esters
in the distillate. The ester concentration will also vary in the storage of the fi'uit spirits,
making this study a preliminary guess at the concentration of the esters in the final
product of the spirits. However, the general trend of more trays making for a lower
concentration of the ester in the distillate can be seen.

F usel alcohols present in the mash as a side product of the yeast in the

fermentation are known to have an undesired aroma in higher concentrations. The

65

amount of these filSCl alcohols in the distillate should be minimized in an attempt to make
quality distilled fruit spirits. The fusel alcohols studied included; l-propanol, 2-methyl-
2-propanol, and isoamyl alcohol. The higher number of trays used in the distillation
shifted the concentration maxima for the fusel alcohols towards the tails portion of the
distillate, thus making it possible to collect more quality distillate, before the unpleasant
fiJsel alcohol aroma would ruin the quality of the distillate.

Benzaldehyde is present in fruit distillate as a fermentation product, as well as a
product from the amygdalin in the stones of the fruit, can be considered a positive flavor
component. By varying the number of trays used in the distillation the concentration can
be shifted to be greater at a later distillate volume when more trays are used. The use of
more trays also led to a slight increase in the overall concentration in the benzaldehyde in
the distillate, this trend could be due to variation in the fermentation mash, however.

Methanol is the compound most interesting in the distillation of these fruit sprits.
The concentration of methanol in the distillate looks like the other low boiling
components in the heads and early hearts portion of the distillate, decreasing in
concentration as the distillate volume collected increases. However, methanol also has a
similar profile as the high boiling components in the later volumes of the hearts distillate
into the tails of the distillation. The minimum of the concentration of methanol occurs in
the middle of the hearts portion, making the concentration fall near the regulatory limits,
making it possible to sell the distillate legally. No distinct change in concentration can be
seen through the change in the number of trays used in the distillation, and the catalytic

converter, while it changed the distillation profile, did not have an effect that could be

66

easily distinguished. Regulation of methanol through changing distillation conditions
seems to be an unlikely prospect.

Overall the profile of the more common congeners found in fruit spirits were
studied for how the distillation conditions affect their concentration in the distilled spirits.
The analysis of these effects can lead to better tasting, and smelling fiuit spirits. The
amount of quality distillate produced per fermentation can be increased by deciding on
where to make the heads/hears cut on order to minimize acetaldehyde, methanol, and
ester concentrations in the hearts, as well as deciding on the hearts/tails cut to minimize
the fuse] alcohol content of the hearts of the spirits. The production of quality fruit spirits

should benefit fiom this study.

67

6. FUTURE WORK

Analysis of the congeners in fruit spirits can be difficult over a short time frames
as they have similar properties in being low molecular mass organic compounds, while
having a broad range of boiling points. The advent of High Speed Gas Chromatography
(HSGC) or a GC accelerator may allow the separation and analysis of the congener
compounds in the distillate over a quick time frame (2 minutes vs. over an hour). This
could allow for the distillate to be a sampled at the top of the condensing column and
analyzed before the bulk of the distillate is condensed at the bottom of the condensing
column. This could open a broad avenue for better control of the distilled spirits market,
and prevent high concentrations of harmful compounds such as ethyl carbamate, from
being in spirits sold to consumers.

Other work must be done to study the effectiveness of the catalytic converter for
the removal of the cyanide containing compounds from the distillate. As the catalytic
converter seems to promote ester formation at relatively lower concentrations of ethanol
in the distillation process, the inclusion of the catalytic converter in future still sales may

need further review.

68