Group Title: Citrus Station mimeo report - Florida Citrus Experiment Station ; 54-10
Title: Use of diacetyl test for detection of microbiological spoilage in citrus juices
Full Citation
Permanent Link:
 Material Information
Title: Use of diacetyl test for detection of microbiological spoilage in citrus juices
Series Title: Citrus Station mimeo report
Physical Description: 7 leaves : ; 28 cm.
Language: English
Creator: Hill, E. C
Wenzel, F. W
Citrus Experiment Station (Lake Alfred, Fla.)
Florida Citrus Commission
Publisher: Florida Citrus Experiment Station :
Florida Citrus Commission
Place of Publication: Lake Alfred FL
Publication Date: 1954
Subject: Concentrated fruit juices -- Preservation -- Florida   ( lcsh )
Citrus juices -- Microbiology -- Florida   ( lcsh )
Microorganisms -- Florida   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Includes bibliographical references (leaf 3).
Statement of Responsibility: E.C. Hill and F.W. Wenzel.
General Note: Caption title.
 Record Information
Bibliographic ID: UF00072357
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 74322889

Full Text

Use of Diacetyl Test for Detection of Microbiological
Spoilage in Citrus Juices
E. C. Hill and F. W. Wenzel

The spoilage of citrus juice by microorganisms during the evaporation pro-
cess used in the manufacture of concentrate has almost entirely been of the kind
termed "buttermilk". The off-flavor associated with this type of spoilage is due
to the production of diacetyl, generally by members of the genera Leuconostoc and

The use of a chemical method, the diacetyl test, for the detection of the pro-
duction of diacetyl and acetylmethylcarbinol and therefore the coincident growth of
the contributory bacteria during concentration of orange juice has been previously
reported (1), (2), and (3). Further investigations were carried out during the
past year using both the diacetyl test and direct microscopic count as methods for
the detection of microbial activity and the results obtained are included in this

Chemical method diacetyl test. This is a quantitative colorimetric test
for acetylmethylcarbinol and diacetyl in citrus juices. It is based on the color
reaction originally observed by Voges and Proskauer and is similar to the quanti-
tative test for blood acetoin of Westerfeld (4). The diacetyl test can be applied
in various ways. The test may be made on reconstituted concentrate as previously
reported (1). The condensate from the evaporator may be checked by this method
but this is feasible only when the condensate is not excessively diluted with water.
An adaptation and modification of the method originally reported (3) was first
used as a control procedure in a commercial plant by Mr. Ellis M. Byer, Florence
Foods, Inc., Florence Villa, Florida. The modification consisted of using the dis-
tillate from reconstituted concentrate rather than making the test directly on the
reconstituted juice. Description of the diacetyl test, including the distillation
procedure, is attached to this report.

Direct microscopic count. Slides for microscopic examination were prepared
in the following manner as previously reported (1). The citrus concentrate was
reconstituted to 12OBrix + 0.20 and then 5 ml. transferred to a clean test tube.
Five ml. of a filtered 0.075% aqueous solution of crystal violet, color idex.681,
were added to the juice. The tube was stoppered and shaken. A,.0ii ml. 6f-the\
juice-dye mixture was transferred by means of a 0.01 ml. pipettb'eto a thoroughly
clean microscope slide. The material was distributed evenly oi, a 1 sq, ,c-." area
on the slide by means of a straight nichrome wire. The slide was then laced
under a heat lamp until dry. When the smear had dried, it was sprayed liberally/
with Spraint plastic spray and then dried. This plastic coating'does not inter-'
fere with clear microscopic vision and produces a slide which isIpermanent aid'can
be filed. No other means of fixing is necessary and there is no destaining-or
washing. This technique of using a plastic spray prevents the juice r-the organ-
isms therein from being lost during the preparation of the slide, a difficulty
previously encountered in making smears of citrus juices for microscopic examination.
Using this procedure the microorganisms are stained a dark violet and are easily
visible on the faint to light violet background. The number of microorganisms in
each of 25 fields was counted and the average count per field calculated.

Citrus Station Mimeo Report 54-10. Florida Citrus Experiment Station and Florida
Citrus Commission, Lake Alfred, Florida. 465 10/6/53 ECH

Experimental results. Before discussing the results obtained from several
pilot plant runs made during the past year, it should be pointed out that the
distillate from freshly extracted citrus juices will give a positive test for
diacetyl, thereby indicating the presence of acetylmethylcarbinol, diacetyl, or
similar compounds in the juices. When 15 samples of fresh orange juice from fruit
of different varieties and different degrees of maturity were distilled, the dis-
tillates were found-to contain diacetyl ranging from 1.2 to 4.8 p.p.m. resulting
in an average of 2.4 p.p.m. There is some indication that the acetylmethylcarbi-
nol or diacetly in fresh juice increases with maturity. A range from none to 1.2
p.p.m. of diacetyl was obtained in the distillates from 18 samples of reconstitu-
ted orange, 2 samples of tangerine, and 1 sample of grapefruit concentrate. A
loss of some substances during evaporation was indicated since the diacetyl con-
tent of the concentrated products was usually smaller than that found in the
freshly extracted juices. These facts should be kept in mind when interpreting
results obtained using the diacetyl test.

A comparison between the diacetyl test made directly on reconstituted con-
centrate (Procedure A) and the test made on the distillate from reconstituted con-
centrate (Procedure B) is shown in Table 1. In both Runs 1 and 2, 200Brix concen-
trate was inoculated with Lactobacillus plantarum and placed in a single-stage
pilot plant evaporator. During evaporation the Brix of the concentrate was held
constant at 200 b$ the continuous addition of pasteurized deionized water. Con-
centrate was withdrawn from the evaporator every hour for testing by both proce-

Procedure B requires the recovery of 25 ml. of distillate from 300 ml. of re-
constituted concentrate. Theoretically, this concentrates the diacetyl twelve
times in the distillate. Therefore, the p.p.m, of diacetyl found in the distil-
late was divided by 12 to obtain the corresponding p.p.m. of diacetyl in the re-
constituted juice as shown in Table 1, column 4, However, this gave smaller
vales than. those obtained by Procedure A in which the test was made directly on
the reconstituted juice. This is explainable on the basis that when the test is
performed on the juice itself all of the acetylmethylcarbinol and diacetyl is
available for color formation, whereas when the reconstituted juice is distilled
most of the diacetyl is recovered in the distillate, but only about 8% of the
acetylmethylcarbinol is distilled over and recovered. The advantages of the dis-
tillation procedure are great enough to render the incomplete recovery of acetyl-
methylcarbinol insignificant. The absence of interfering substances in the clear
distillate obtained permit very small changes in diacetyl concentration in the
juice to be readily detectable.

The data in Table 2 were obtained by using a two-stage evaporator. The
connecting line between the two stages was closed so that the juice in one stage
would not come in contact with that in the other stage. Pasteurized 20Brix Va-
lencia orange concentrate was placed in both stages of the evaporator. The con-
centrate in Stage 1 was not inoculated, while the concentrate in Stage 2 was in-
oculated with Lactobacillus plantarum. During processing the concentration of the
product in both stages was held constant at 20Brix by the addition of pasteurized
deionized water. The diacetyl in the distillate from the reconstituted uninocu-
lated concentrate (Stage 1) decreased from 2.7 p.p.m. initially to 0.2 p.pJm.
after one hour and remained at that level for the duration of the run. The dis-
tillate from the reconstituted concentrate (Stage 2), which had been inoculated
with Lactobacillus plantarum had an initial diacetyl content of 3.1 p.p.m. which
decreased after 1 hour to 1.1 p.p.m. This level was maintained for 3 hours after
which there was a gradual continuous increase to 7.3 p.p.m. at which time the run
was discontinued. Data in Table 2 indicated that the increase in microbial activ-
ity was detected at the same time by both the diacetyl test and the direct micro-
copic count.
Citrus Station Mimeo Report 54-10. Florida Citrus Experiment Station and Florida
Citrus Commission, Lake Alfred, Florida. 465a 10/6/53 ECH

The run tabulated in Table 3 simulated to some extent the conditions that
exist in the low stage of commercial evaporators. Inoculated 200Brix concentrate
was placed in a single-stage pilot plant evaporator. Evaporation was started
and freshly extracted juice was continuously added to keep the concentrate in the
evaporator at 20Brix and 20Brix concentrate was steadily removed to maintain
correct juice level. The first indication of impending spoilage was between the
second and third hour of evaporation when the diacetyl content in the distillate
increased from 4.9 to 6.3 p.p.m. The direct microscopic count showed an increase
from 0.6 to 1.1 organisms per field between the third and fourth hour.

During all of the experimental runs previously described, samples of concen-
trate from the evaporator were reconstituted and tasted by several persons famili-
ar with the type of off-flavor caused in orange juice by bacterial production of
diacetyl. Results of these organoleptic tests indicated chiefly the need for a
more reliable method for detecting this type of spoilage in its early stages. The
senses of smell and taste are quite sensitive to diacetyl, but the ability of an
individual to detect off-flavors in juice varies from person to person, and in the
case of an individual from day to day. It is the opinion of the authors that it
is better to use chemical or microbiological methods, such as the diacetyl test and
the direct microscopic count, than to rely on organoleptic tests.

In summation it should be stated that the diacetyl test described is a rapid,
sensitive, and reliable method for determining the progressive microbiological
activity of organisms capable of producing spoilage of the "buttermilk" type in
citrus juices and concentrates. The direct microscopic count was also found to be
a very satisfactory means for detecting increases in microbial activity. Proper
utilization of either of these methods will help processors of citrus concentrates
to prevent loss of product because of bacterial spoilage.
Literature Cited

1. Hill, E. C., Wenzel, F. W., and Barreto, A. Comparison of various methods for
detection of microorganisms which produce off-flavors in orange concentrate.
Report presented at Third Annual Citrus Processors Meeting, Lake Alfred,
Fla. October 1, 1952.

2. Hill, E. C., Wenzel, F. W., and Barreto, A.- Colorimetric method for detection
of microbiological spoilage in citrus juices. Paper presented at Thir-
teenth Annual Meeting of the Institute of Food Technologists, June 24, 1953,
Boston, Mass.

3. Wenzel, F. W., Hill, E. C., and Barreto, A. Detection of acetylmethylcarbinol
or diacetyl in orange concentrate as an indication of the growth of certain
bacteria which produce off-flavors. Report to citrus processors at Citrus
Experiment Station, Lake Alfred, Fla. May 15, 1952.

4. Westerfeld, W. W. A colorimetric determination of blood acetoin. J. Biol.
Chem., 161, 495 (1945).

Citrus Station Mimeo Report 54-10. Florida Citrus Experiment Station and Florida
Citrus Commission, Lake Alfred, Florida. 465b 10/6/53 ECH

Comparison of

two procedures for diacetyl test for microbiological
activity in concentrated orange juicea

Processing Procedure A Procedure B
time Reconstituted juice Distillate Reconstituted
hr. Diacetyl p.p.m. Diacetyl p.p.m.
Run No. 1






























Run No. 2



























a Pasteurized 20Brix orange concentrate was inoculated with Lactobacil-
lus plantarum and processed in evaporator. The concentration of the
product was held constant at 20Brix during processing by the addition
of pasteurized deionized water.

Citrus Station Mimeo Report 54-10. Florida Citrus Fxperiment Station and
Florida Citrus Commission, Lake Alfred, Florida. 465 .10/6/53 ECH

-- -

Diacetyl content and bacteriological counts in 20Brix orange concentrate
during processing

Stage 1 Stage 2
Concentrate Concentrate' inoculated
not inoculated
Processing Diacetyl Diacetyl Direct micro- Plate count
time in in scopic count
distillate distillate
hr. p.p.m, Average of 25 Orange serum
fields agar, pH 5.4
Initial 2.7 3.1 0.4 480,000

1 0.2 1.1 0.4

2 0.1 1.1 0.4 840,000

3 0.2 1.1 0.5

4 0.2 1.2 0.6 1,060,000

5 0.1 1.3 0.8
6 0.2 1.6 0.9 1,110,000

7 0.1 1.7 1.4

8 0.2 2,3 2,4 2,640,000

9 0.2 3.0 3.1

10 0.2 3.8 3.6 4,850,000

11 0.1 4.4 5.0

12 0.1 5.7 5.3 7,300,000

13 0.3 7.3 5.9 6,900,000
b Two stage evaporator used in which inter-connecting line was closed. Con-
centrate, pH 4.0, in both stages was identical with the exception that con-
centrate in stage 2 was inoculated with Lactobacillus plantarum. The con-
centration of the product in both stages was held constant at 20Brix during
processing by the addition of pasteurized deionized water.

Citrus Station Mimeo Report 54-10. Florida Citrus Experiment Station and Florida
Citrus Commission, Lake Alfred, Florida. 465d 10/6/53 ECH

TAK7.7 3
Diacetyl content and bacteriological counts in 20OBrix orange concentrate
during processing simulating commercial operations

Processing Diacetyl Direct micro- Plate count
time in scopic count
hr. p.p.m. Average of 25 Orange serum
fields agar, pH 5.4

Initial 4.9 0.7 920,000

1 4.9 0.6

2 4.9 0.6

3 6.3 0.6 1,190,000

4 7.1 1.1

5 9.9 1.7

6 16.3 2.4

7 31.4 5.5 2,700,000

8 42.0 8.8

9 66.4 15.0 3,800,000

c 200Brix orange concentrate, pH 3.7, inoculated with Lactobacillus plantarum
and processed in evaporator. The concentration of the product was held
constant at 20Brix during processing by the addition of freshly extracted

Citrus Station Mimeo Report 54-10. Florida Citrus Experiment Station and Florida
Citrus Commission, Lake Alfred, Florida. 465e 10/6/53 ECH

Method for the Detection of Acetyl-Methyl-Carbinol
or Diacetyl in Citrus Concentrate

Basis for test

Acetylmethylcarbinol in the presence of atmospheric oxygen and strong alkali
is oxidized to diacetyl. In the presence of alkali, creatine, and alpha-naphthol
diacetyl forms a compound that results in a red solution, the intensity of which
may be measured colorimetrically.


(1) Alpha-naphthol solution. Dissolve 5 g. of alpha-naphthol (Eastman
Grade) in 100 ml. of 95% ethyl alcohol.

(2) Concentrated alkali solution containing creatine. Dissolve 40 g. of
KOH in sufficient water to make 100 ml. of solution and then add 0.3 g, of

The reagents should be prepared daily or may be kept for two weeks if


.(1) Reconstitute sample of citrus concentrate to be tested to 12 Brix
i0.2 .

(2) Distill 25 ml. from 300 ml. of reconstituted juice using an efficient
condenser. Collect the distillate in a graduated cylinder. The distillate should
flow down the sides of the graduate to prevent the dispersion of peel oil which
will cause cloudiness.

(3) Pipette 10 ml. of distillate from the graduated cylinder, without
disturbing any floating peel oil, into a 25 mm. x 200 mm. test tube,

(4) Add 5 ml. of alpha-naphthol solution and 2 ml. of KOH solution contain-
ing creatine, stopper the test tube, shake vigorously for 15 sec. and allow to
stand for 10 min. Shake again vigorously for 15 sec. In the same manner and at
the same time prepare a reagent blank using distilled water in place of distillate
from the juice.

(5) Fill colorimeter tubes with test sample and reagent blank. Null
electrophotometer with reagent blank and measure color intensity in test sample.

(6) Determine concentration in p.p.m. of diacetyl in distillate by using
a calibration curve obtained by applying this procedure to distilled water to
which known amounts of diacetyl have been added.

Citrus Station Mimeo Report h5-10. Florida Citrus Experiment Station and
Florida Citrus Commission, Lake Alfred, Florida. L76-10/6/53 ECH

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs