/ Citrus Station Mimeo Report 60-1
September 16, 1959
Preliminary Studies of Volatile Flavor Components in Citrus Juices
Using Gas Chromatography
R. W. Wolford
The continued emphasis being placed on maintaining good flavor in Florida
citrus processed products has resulted in a basic research study of the volatile
components responsible for the natural flavor and occasional off-flavor in these
products. The development and acquisition of the facilities and equipment for
this investigation has been underway for approximately 2 years. However, the
project has been actively engaged in since July, 1958.
The evaluation of maturity of citrus fruits and the determination of the
general quality of processed citrus products are based to a large extent on
standard chemical and physical methods. However, the flavor evaluation of such
products is still of an empirical nature. Many workers have made excellent
contributions to the knowledge of the volatile flavoring constituents of citrus
juices. Orange juice has been investigated extensively, especially with regard
to changes in composition of the volatile fractions, both oil- and water-soluble,
which may result from heating and storage (1, 2, 3, 4, 5, 6). It is felt that
there is a tremendous need for a better knowledge of the interrelation between
various volatile flavor components and the natural flavor of the juice. From
the standpoint of an overall characterization of fresh orange juice and its
products, canned juice or frozen concentrate, we must not only have qualitative
and quantitative information about chemical constituents, but also the knowledge
that certain volatile flavor components are outstanding in influencing the
particular flavor which we recognize as most acceptable. Moreover, if such
knowledge can be obtained we will be much closer to the development of physical
and chemical methods for the actual determination of flavor. Subsequently, a
correlation between flavor constituents and the accepted maturity standards,
which is considered to be a basic need, may be attainable.
It has been the primary aims in this work to develop methods for isolating
the volatile flavor components in a fresh state at as low a temperature as
possible; also, to be in a position to analyze the volatile constituents in a
rapid manner before chemical changes in these substances occur. Many of the
natural flavor components are in extremely low concentrations, therefore, these
must either be increased in concentration by some special means or be analyzed
in ultra-micro quantities.
The significant increase in the amount of research being conducted at the
present time on many fruits, their juices, fresh vegetables, and other food
products can be attributed to the advent of a relatively new and highly sensi-
tive method of analysis, gas chromatography. In conjunction with infrared and
mass spectrometry a powerful set of analytical tools are now available for
nearly complete identification and characterization of volatile flavor constit-
uents. Also, the analysis of ultra-micro quantities of material is the accepted
rule. Gas chromatography lends itself to this type of research since one of
its principal requirements is that the material to be analyzed must be volatiliz-
able at the temperature of operation of the instrument.
Florida Citrus Experiment Station
and Florida Citrus Commission,
Lake Alfred, Florida.
938 9/16/59 RWW
This discussion is presented to give a brief outline of the approach and
the magnitude of this project which is now being actively pursued. It is not
the intent of this report to provide at this time any conclusions regarding the
flavor constituents of citrus juices. An attempt will be made at the time of
presentation to show some of the evidences and indications of the identity of a
number of flavor components isolated from orange juice. A large amount of data
has been accumulated and considerable effort is being made to improve the facil-
ities for extraction and analysis to provide some answers to the many problems
one might consider.
Experimental Recovery of Volatile Flavor Components_
The recovery of volatile flavor components, oil- and water-soluble, has
been obtained as follows. A Buckeye stainless steel spray chamber (9 gal.
capacity) with an orifice spray and an upper baffle packed with stainless steel
rachig rings to provide partial fractionation was used. The vapors from orange
juice were passed through a series of seven traps. The first two at ice water
temperature (30 C.), one at brine temperature, and three traps at the tempera-
ture of dry ice and alcohol (approx. -730 C.). A final trap contained 2,
4-dinitrophenylhydrazene to catch the aldehydes and ketones. The spray chamber
was run initially at a jacket temperature of 1600 F. The temperature of the
vapor stream to the first trap was 1250 F. In two runs made with this system
the approximate evaporation rate was 1.6 gal. water/hour. The pilot plant
evaporator vacuum system was used to maintain a vacuum of 27-28 inches of
mercury throughout the runs. This system was employed in the recovery of
volatile flavor components from Hamlin juice. A total of 3.4 grams of oil
was obtained from each of the runs. The water phases from the first three
traps and from the three low temperature traps were combined separately, ex-
tracted with ether and each solution was analyzed by gas chromatography.
An all glass circulating evaporator, the design for which was obtained
from Dr. W. L. Stanley, U.S.D.A. Fruit and Vegetable Chemistry Laboratory,
Pasadena, California, was also employed. The initial runs made on fresh
orange juice and on reconstituted frozen orange concentrate were carried out
at a pressure of 0.5 mm. mercury. A very high velocity through the system was
observed, while the temperature of the juice after passing through the heat
exchanger never rose above 100 C. The vapors were subjected to three traps,
cooled by dry ice and alcohol. No oil separated in these runs. Obviously
large losses of volatile components were occurring through the vacuum system.
Later work with this separation system was carried out at 50 mm. pressure. The
juice passing through the heat exchanger reached a temperature of 400 C. and
water was condensed in the second cyclone separator. An oil layer was obtained
and the combined aqueous phases were extracted with ether for subsequent analy-
sis by gas chromatography.
The third source of volatile flavor components of orange juice was a
commercial essence recovery unit. Sizeable quantities of essence and oil were
obtained from this unit for further laboratory extraction of individual flavor
components. Most of the analytical separations on recovered essence were made
during the Valencia season. The fact that a virtually oil-free material was
available simplified to some extent the ether extraction of the volatile, water-
soluble components. The oil fraction was analyzed directly by gas chromatography.
1 Samples of orange essence supplied through cooperation of E. J. Kelly and
Associates, Inc., and Libby, McNeill and Libby.
Florida Citrus Experiment Station and Florida Citrus Commission,
Lake Alfred, Florida. 938 a 9/16/59 RWW
In this study of volatile flavor components the oil-soluble fraction ob-
tained by either centrifugation, collected at reduced pressure in low temperature
traps or separated by dilution of recovered concentrated essence was analyzed
directly by gas chromatography. The water-soluble components have been re-
covered for analysis by continuous solvent extraction with diethyl ether. By
low temperature evaporation of the resulting ether solution a relatively high
concentration of the water-soluble volatile components was obtained and by re-
taining some ether in the mixture the sample was injected into the instrument
in a volatilizable form.
The liquid stationary phases employed in the Perkin-Elmer 154-C Vapor
Fractometer were didecyl phthlate, D. C. 200 silicone oil, polyethylene glycol
(Carbowax 1500), and silicone grease, all fixed on Celite; also Ucon polar #9,
Ucon nonpolar, and Carbowax 600 which were fixed on either C-22Firebrick or
Chromosorb regular. The maximum column length used was 2 meters. The tempera-
ture of operation of the column was generally 1500 C. or 1700 C. and occasion-
ally 2000 C. Generally, it was found that the best separation and resolution of
individual components was obtained on 2 meters of Carbowax 600 at a temperature
of 1500 C. and 9 p.s.i. gas pressure with the carrier gas (helium) flow rate at
35 ml./min. A 5 my Leeds and Northrup recorder was used in conjunction with the
Perkin-Elmer instrument. Quantities of 50 microliters of either the oil fraction
or the ether extract, containing the volatile water-soluble components, was in-
jected into the instrument in those analyses designed for sample collections.
The individual fractions were condensed directly into a collection tube at the
temperature of ice and salt (-210 C.). Each condensed fraction was then trans-
ferred with a capillary needle to an ultra-micro liquid cell having 0.05 mm.
spacers. The infrared analyses were run using the micro liquid cell in con-
junction with a beam condenser on a Beckman IR-4.
That there was a possibility of causing some change in the volatile oil-
soluble constituents, in particular, was not precluded. Various fractions were
rechromatographed and the recordings showed single components in each case at
the sensitivity employed.
(1) Blair, J. S., Godar, E. M., Masters, J. E., and Reister, D. W., Food
Research 17, 235 (1952).
(2) Hall, J. A., and Wilson, C. P., J. Am. Chem. Soc. 47, 2575 (1925).
(3) Kirchner, J. G., and Miller, J. M., J. Agr. Food Chem. 5, 283 (1957).
(4) Kirchner, J. G., Miller, J. M., Rice, R. G., Keller, G. J., and Fox M..M.,
J. Agr. Food Chem. 1, 510 (1953).
(5) Kirchner, J. G., and Miller, J. M., J. Agr. Food Chem. 1, 512 (1953).
(6) Naves, Y. R., Parfums France 9, 60 (1931).
Florida Citrus Experiment Station
and Florida Citrus Commission,
Lake Alfred, Florida.
938 b 9/16/59 RWW