quality improvement of bixin from annatto seed by modified extraction methods

Concentration of m-xylene in annatto extracts during extraction processes by sodium hydroxide solution, soybean oil, and acetone……….. Percentage of bixin extracted from annatto seed at

Hoang Van Chuyen

Directed by Professor Jong-Bang Eun

August 2011

Page

ABSTRACT……… 1

INTRODUCTION……… 3

MATERIALS AND METHODS……… 14

1 Materials………

2 Measurement of moisture content in annatto seeds………

3 Determination of total content of bixin in annatto seed………

4 Extraction of bixin………

5 Determination of bixin degradation………

6 Analysis of bixin content in annatto extracts………

7 Gas chromatography (GC) analysis of toluene and m-xylene………

8 Data analysis………

14 14 14 15 16 17 17 18 RESULTS AND DISCUSSION 1 Moisture content and total bixin content of annatto seeds………

2 Extraction yield of bixin from annatto seeds by three methods………

3 Degradation of bixin during three extraction processes………

4 Total amount of bixin extracted out from annatto seeds………

5 Amounts of toluene and m-xylene produced during extraction processes……

6 Effects of temperature and volume of sodium hydroxide solution on the extraction yield of bixin………

20

20

20

21

25

27

32

7 Effects of temperature, volume of soybean oil and light on the extraction yield

of bixin and level of toluene and m-xylene in oil extracts………

8 Effects of extraction method, volume of acetone and light on the extraction yield of bixin and content of volatile compounds in extracts………

9 Extraction of bixin from annatto seed by combinations of sodium hydroxide solution and soybean oil………

36 42 46 CONCLUSION……… 51

REFERENCES……… 52

ABSTRACT IN KOREAN……… 56

ACKNOWLEDGEMENTS……… 58

LIST OF TABLES

1 Estimated world production of annatto seed……… 6

2 Annatto products and their uses……… 7

3 Codes and synonyms of annatto pigment……… 8

4 Moisture content and total bixin content of annatto seeds during storage……… 22

5 Concentration of toluene in annatto extracts during extraction processes by

sodium hydroxide solution, soybean oil, and acetone……… 30

6 Concentration of m-xylene in annatto extracts during extraction processes by

sodium hydroxide solution, soybean oil, and acetone……… 31

7 Concentration of toluene and m-xylene in soybean oil extracts in daylight and in

darkness ……… 41

8 Concentration of toluene and m-xylene in acetone extracts in daylight and in

darkness……… 45

9 Toluene and m-xylene contents in the extracts of combination 1……… 49

10 Toluene and m-xylene contents in the extracts of combination 2 ……… 50

LIST OF FIGURES

1 Annatto tree, fruits, and seeds……… 5

2 The structural formula of cis- and trans-bixin……… 11

3 Sample preparation procedure for analysis of toluene and m-xylene in annatto

extracts by gas chromatography……… 19

4 Bixin yield during extraction processes by soybean oil, acetone, and sodium

hydroxide solution……… 23

5 Degradation of bixin during extraction processes with vegetable oil, acetone, and

sodium hydroxide solution……… 24

6 Factual percentage of bixin extracted out from annatto seeds by vegetable oil,

acetone, and sodium hydroxide solution………

26

7 GC-FID chromatogram of toluene (1), ethylbenzene (2), m-xylene (3), and o-

xylene (4) in annatto extract……… 29

8 Extraction yield of bixin at different temperature by sodium hydroxide solution 33

9 Extraction yield of bixin with different ratio of annatto seed weight/volume of

sodium hydroxide solution……… 35

10 Extraction yield of bixin by soybean oil at 80 and 100oC……… 37

11 Extraction yield of bixin with different ratio of annatto seed weight/volume of

soybean oil……… 39

12 Yield of bixin extracted by submerging in acetone at 50oC with different ratio of

annatto seed weight/volume of acetone and in the dark condition……… 43

13 Percentage of bixin extracted from annatto seed at stages of the extraction

combinations between sodium hydroxide solution and soybean oil……… 47

Quality Improvement of Bixin from Annatto Seed by

Hoang Van Chuyen

Department of Food Science and Technology Graduate School of Chonnam National University (Directed by Professor Jong-Bang Eun)

ABSTRACT

The extraction yield, degradation of bixin, and the contents of toluene and m-xylene in

annatto extracts during extractions by acetone, soybean oil and sodium hydroxide solution were

investigated Pigment from annatto seeds was extracted by immersing seeds in soybean oil at

120oC, by acetone in Soxhlet extractor, and by sodium hydroxide solution at 50oC The bixin

extraction yield by acetone was the highest (81.3%) and followed by the extraction by sodium

hydroxide (45.4%) The extraction by soybean oil had the lowest bixin extraction yield (24.7%),

and it decreased gradually after 40 min of extraction process The losses of bixin in the

extraction by acetone and sodium hydroxide increased slowly at quite low levels These were

only 12.3% and 14.5% after 120 min, respectively In contrast, the extraction by soybean oil

caused severe degradation of bixin, and it grew up rapidly during process to 63.4% All three

methods produced trace amounts of toluene, which was the highest in sodium hydroxide extract

at the level of 18.25 mg/L Concentration of m-xylene in acetone extracts was very high and

remained unchanged during extraction process in the range of 90-110 mg/L Amount of

m-xylene produced by soybean oil extraction also increased rapidly in a time-dependent manner to

80.88 mg/L at the end of extraction, whereas appearance of this compound in sodium hydroxide

extracts was very low

The replacement of extraction in Soxhlet extractor by immersion of annatto seeds in

acetone at 50oC with exclusion of light effects resulted in a great reduction of m-xylene content

in acetone extracts (15.74 mg/L) Moreover, it also achieved quite high extraction yield of bixin

(67.2%) after only a short time of extraction (40 min) without significant effects of acetone

volume to extraction yield Besides, this method also required simple equipment for extraction

process than the Soxhlet system The combination of 40-min extraction by sodium hydroxide

solution at 50oC and 20-min extraction by soybean oil at 100oC in darkness significantly lowered

concentration of volatile compounds compared to those of single extraction methods Total bixin

yield of this combined extraction (53.7% in 60 min) was significantly improved, compared to

45.4% in 120 min of sodium hydroxide extraction and maximum bixin yield (30.2%) of soybean

oil extraction Moreover, the reduction of solvent volume in this combination produced extracts

with high concentration of pigment, which provide benefit in cost saving for production of

annatto powder and distribution of oil-soluble annatto Therefore, it is suggested that this

combined extraction is a potential replacement for single extractions by sodium hydroxide

solution and soybean oil

I INTRODUCTION

Annatto is a natural colorant, which imparts a range of color from yellow to red, due to

content of color compounds in the solution This pigment is obtained from the seed coat of the

tropical shrub Bixa orellana L This tree is native to tropical South America, where it has been a

traditional part of some foods for centuries But it is also cultivated in many countries of Central

America (Mexico, Guatemala, Caribbean), Africa (Kenya, Tanzania, Côte d’Ivoire, Angola), and

South Asia (India, Sri Lanka, Philippines, Vietnam), especially in coffee-growing areas The

Bixa orellana L plant grows from two to five meters tall when mature The fruits of the tree Bixa orellana L consist of a pod covered with fleshy spines, varying in size and shape The inside of

the pod is usually divided into two valves containing between ten to fifty small seeds (Figure 1)

Fruits become mature after pollination 5–6 months Production of seed reaches a maximum in 4–

12-year-old plants, which can remain productive for more than 20 years Seed yield is reported to

be as high as 3-5 t/ha Annatto seeds that are properly harvested and dried to suitable moisture

content retain their quality for a long time provided they are stored in a cool, dark and dry place

Color content of the seed varies from1.5 to 4% due to variety, cultivation conditions, and

postharvest techniques (Bechtold and Mussak, 2009)

The principle export form of annatto is seed However, several suppliers have carried out

partial processing/refining of annatto seeds into concentrates to increase export values The main

market for annatto is the USA, Western Europe and Japan Besides, there is also considerable

inter-trade (in seeds) between the Central and South American suppliers The quantity of annatto

seed harvested in the producing countries is estimated at 14,500 metric tons Brazil, Peru,

Ecuador and Kenya contribute the major proportion of this production (Table 1) It is likely that

around 7,500 tons of annatto seed are used annually as a food color worldwide and, assuming an

average color content of 2%, this equates to 150 tones of bixin available for extraction The

remaining 7,000 metric tons of annatto seed is consumed locally in Brazil, Peru and Ecuador

mainly as a spice/condiment

Annatto seeds and extracts have been used for over 200 years to impart a range of color

from yellow to red Although commercial production occurred in South and Central America

from the beginning of 19th century, the production of commercial extracts in Europe and the

United States started until the 1870s, primarily to color butter and cheese In modern times,

annatto color is also used worldwide in other foods like meat and fish products, soft drinks, sugar

confectionery, margarine, ice cream, soups, etc (Table 2) Annatto plays a significant role in the

industry of food color It is considered that annatto extracts rank as the second most

economically important natural color in the world and it is the most frequently used natural color

in the food industry in the United Kingdom(FAO, 1995; Lauro & Francis, 2000; Cannon, 2003)

Variety of codes and synonyms is used for annatto pigment due to each region,

organization, and country (Table 3) In the European Union and the United States, annatto is

permitted to use in foods, drugs, and cosmetics

Figure 1 Annatto tree, fruits, seeds, and powder (http://davesgarden.com)

Table 1 Estimated world production of annatto seed (metric tons)

(either as seed or its equivalent in extract)

Europe Japan Others

3,000 2,500 1,500

(Smith, 2006)

Table 2 Annatto products and their uses

Formulation Product description Typical applications

Water soluble powder Annatto extract with potassium

carbonate and potassium hydroxide

Dry mixes, soups, drinks, cakes, cereals, desserts, tablets

Water soluble liquid Annatto extract in potassium

solution

Chesses, cereals, cakes, drinks, snack foods, ice cream cones, ice cream, sausage casings

Oil soluble liquid Annatto extract in vegetable oil Margarine, fats and oils,

shortenings, spreads, salad oils, butters, spice blends, pasta

Oil soluble suspension Annatto suspended in vegetable

oil

Processed cheeses, cake mixes, frostings, candy coatings, salad oil, margarines

Emulsion color Purified annatto, processed and

emulsified with propylene glycol, monoglycerides and potassium hydroxide

Beverage, candies, yogurts, sherbets, butters, margarine, cheeses, shortening, frosting, confectionary coatings, baked goods, snack foods, salad dressings, ice cream

(Satyanarayana, 2003)

Table 3 Codes and synonyms of annatto pigment

INS number 160b (annatto, bixin, norbixin)

Common synonyms CI Natural Orange 4; CI 75120

Achiote, Achiotl, Achote, Annotta, Arnatta, Arnatto, Arnotta, Bija, Rocou, Roucou, Roucouyer, Roucoyer, Orlean, Orleanstraugh, Terre orellana, Beni-No-Ki, Urucu, Urucum, L Orange

C.A.S Number Annatto: 1393-63-1

Bixin: 39937-23-0 (trans) 39937-79-5 (cis) Norbixin: 542-40-5 (trans) 626-76-6 (cis)

Pigment may be separated from annatto seeds in different ways, such as immersion of

seeds in hot vegetable oil or dilute alkaline aqueous solutions Solvent extraction techniques are

now also used widely as a mean of production of annatto powder

In the original and traditional food recipes of Latin America, annatto seeds were heated

with cooking oil, separated, and the colored oil used for preparing rice, soups, and tortillas A

similar direct-processing system can be applied industrially by immersion of the seeds in

vegetable oil to produce bixin slurry, which can then be heated and filtered The filtrate can be

marketed as a color for high-fat foods

Another separation process of annatto pigment from seeds is carried out by stirring the

annatto seeds in mildly alkaline water This separates the pigment layer from the seed into

alkaline solution, which is then acid precipitated, filtered off and dried to give a granular powder

of annatto

Solvent extraction techniques for production of annatto products of 80 - 97% purity have

been developed as a response to the need for more concentrated annatto extracts Extraction of

annatto seed with organic solvents and subsequent solvent removal yields the bixin crystals,

which are then processed by the manufacturers of food colors according to specific applications

Numerous patents and research reports cover a variety of organic solvents for producing

concentrates, such as chlorinated hydrocarbons, mixtures of ethanol and chloroform, acetone,

ethanol, ethyl acetate, hexane, methanol or alcoholic sodium hydroxide

Recently, a series of efforts have been carried out to use supercritical carbon dioxide fluid

for extraction of pigment of annatto pigment However, the results of these studies seemed to be

not satisfied with both economical and efficient aspects (Anderson, 1997; Nobre, 2006; Silva,

2008)

The principal color component of annatto pigment is bixin (methyl hydrogen 9’-cis-6,

6’-diapocarotene-6, 6’-dioate) The amount of this compound in annatto is more than 80% of color

content (Preston & Rickard, 1980) Bixin normally occurs in the cis- form, also present in

annatto seed, as minor constituents, are trans-bixin, cis-norbixin and trans-norbixin The cis-

form is redder than the trans- form and thus is the pigment used commercially

Not only providing high color intensity but biological properties of annatto pigment such

as antioxidant activity and cancer-preventing effects also attracted consideration of researchers in

many recent studies For examples, antioxidant property of annatto pigment was reported by the

ability of annatto pigment to inhibit hydroperoxide formation of triglycerides (Haila, 1996) The

study of Montenegro et al (2004) also showed that bixin is an efficient quencher of singlet

molecular oxygen (1O2) Júnior et al (2005) demonstrated the antigenotoxic potential of annatto

pigment by protecting DNA damage of E coli cells against UV radiation, hydrogen peroxide,

and superoxide anion (O2.-) and its antimutagenic activity based on the Salmonella mutagenicity

test For these reasons, annatto has a significantly economic importance and it becomes one of

the most frequently used natural colorants of the food industry

Figure 2 The structural formula of cis-bixin and trans-bixin

As almost carotenoids, bixin is very susceptible with processing and storage conditions

such as temperature, light, oxygen, and the factors of foods that contain annatto pigment This

characteristic might cause a significant color loss of annatto pigment As the result, many studies

have concentrated on the degradation of bixin during processing and storage, and in foods added

with annatto pigment Stability of annatto solution extracted by chloroform to light, air,

anti-oxidants and pro-anti-oxidants has been described by Najar et al (1988) They observed the most

destructive agent was light, and followed by the pro-oxidant, benzoyl peroxide Gloria et al

(1995) investigated effect of water activity on the stability of bixin in an annatto extract, in

which bixin was observed to be more stable at intermediate/higher water activities The effects of

processing conditions on the stability of annatto added in to cakes, chegodis, biscuits and fried

rice was reported in the survey of Prabhakara Rao et al (2005) This showed no effect of

microwaves on bixin in the preparation of biscuits and the maximum loss of bixin (65%) in the

processing of the deep fat fried snack (chegodis)

Beside the effect on color intensity of annatto pigment, the decomposition reaction of

bixin also produces some volatile compounds such as m-xylene and lesser amount of toluene,

toluic acid, etc (McKeown, 1965) Unfortunately, these compounds can cause some negative

effects to human health It was claimed that m-xylene exhibits neurological effects at certain

levels like headaches, lack of muscle coordination, dizziness, confusion, and alterations in body

balance Thus, m-xylene or products containing this compound should not be used indoors or

around food Similarly, if toluene is inhaled even at low level, it can also cause tiredness,

confusion, weakness, drunken-type actions, memory loss, and nausea, etc By these reasons,

some surveys have concentrated on the level of these volatile compounds in annatto products

Scotter et al (1995, 1998, 2000) carried out a series of studies on annatto pigment to survey the

formation and composition of its thermal degraded products in different annatto formulations by

using of high-performance liquid chromatography and gas chromatography

Although many studies investigated effects of the factors on stability of annatto pigment,

there have been no reports on the differences in the loss of bixin between processing methods

and the formation of undesirable volatile compounds during extraction processes Likewise,

some studies on volatile compounds in annatto have only concentrated on the commercial

products but not investigated the processing stages of this pigment Therefore, it is necessary to

carry out studies on the stability of bixin and levels of volatile compounds are produced in the

extraction processes to find methods for improving extraction quality of bixin from annatto seed

In the present study, we investigated the degradation, extraction yield of bixin, and formation of

toluene and m-xylene in annatto extracts during three extractions by soybean oil, acetone, and

sodium hydroxide solution Effects of some factors to bixin extraction yield and content of

volatile compounds were also investigated A modification of extraction by acetone and two

combinations of sodium hydroxide solution and soybean oil were carried out to improve

extraction yield and extraction quality of bixin

II MATERIALS AND METHODS

1 Materials

Annatto seeds were purchased from local market in Vietnam Soybean oil (Ottogi) was

purchased from local market in Korea Acetone and sodium hydroxide were supplied by Dea

Jung company; pentane was provided by Kanto Chemical Company; tetrahydrofuran (≥ 99%),

ethylbenzene (99.8%), toluene (99.8%), m-xylene (≥ 99%), and o-xylene (≥ 99%) were obtained

from Sigma-Adrich The other reagents were in analytical grade

2 Measurement of moisture content in annatto seeds

Moisture content in annatto seeds was analyzed by using of a moisture analyzer (MB45,

Ohaus, Greifensee, Switzerland) The program was set up at 120oC and manual operation About

2 grams annatto seeds were weighed exactly in the balance of the apparatus and the analyzing

procedure completed when there were no changes in moisture content observed in 2 min

3 Determination of total content of bixin in annatto seed

Bixin content in annatto seed was determined according to the method of Balaswamy et

al (2006) About 2 grams annatto seeds were weighed exactly in a beaker wrapped and covered

by aluminum foil to prevent the loss of bixin by effects of light Pigment in the seeds was then

extracted repeatedly by chloroform until total pigment is removed from the seeds (determined by

observation color of the extracts) During extraction process, the extracts were collected into a

glass bottle, which was wrapped by aluminum foil, capped, and stored in a refrigerator The total

extract was then diluted to a suitable concentration and bixin content was analyzed by a

spectrophotometer (Optizen 2120UV, Mecasys Co.Ltd, Daejeon, South Korea) The content of

bixin in annatto seed was calculated using the following formula:

= mM(100 − W) 100

C: Content of bixin in annatto seed (%)

m: Total weight of extracted bixin (mg)

M: Weight of annatto seed sample (mg)

W: Moisture content of annatto seeds (%)

4 Extraction of bixin

4.1 Extraction by soybean oil: 5 grams of annatto seeds was submerged into a beaker

containing 20 ml soybean oil at 120oC and stirred for 120 minutes (Preston & Rickard, 1980)

4.2 Extraction by acetone: Pigment in 2 grams of annatto seeds was extracted in a

continuous Soxhlet system with acetone as volatile solvent for 120 minutes (Balaswamy, 2006)

4.3 Extraction by sodium hydroxide solution: 5 grams of annatto seeds were

submerged and stirred in a beaker containing 20 ml of NaOH 0.1N solution at 50oC for 120

minutes (Shuhama, 2003) Bixin contents in the extracts of above methods were measured at

each 20 minutes

4.4 Extraction of bixin from annatto seed in darkness (exclusion of light)

Annatto seeds were submerged and stirred in a beaker containing acetone, sodium

hydroxide solution, or heated soybean oil The extraction processes were prevented from effect

of light by wrapping the beaker with aluminum foil and then the beaker was also covered by

aluminum foil

4.5 Combined extractions of sodium hydroxide solution and soybean oil

Two combined extraction processes between sodium hydroxide solution and soybean oil

were established:

Combination 1: 5 grams annatto seed was submerged in 10 ml sodium hydroxide solution

and stirred in a beaker, which was wrapped by and then covered by aluminum foil at 50oC for 20

min Subsequently, the extract was drained completely and the seeds were extracted continuously

in beaker (wrapped and then covered by aluminum foil) containing soybean oil at 100oC for 40

min

Combination 2: 5 grams annatto seed was submerged in 10 ml sodium hydroxide solution

and stirred in a beaker, which was wrapped by and then covered by aluminum foil at 50oC for 40

min Subsequently, the extract was drained completely and the seeds were extracted continuously

in beaker (wrapped and then covered by aluminum foil) containing soybean oil at 100oC for 20

min

5 Determination of bixin degradation

The degradation of bixin was investigated by measuring bixin content in the samples,

which were added a certain amount of bixin and undergone the same conditions of the three

methods To investigate the loss of bixin in the extraction by vegetable oil, 1 ml of vegetable oil

extract of bixin was added in 19 ml pure vegetable oil at 120oC in a stirring beaker for 120 min

Similarly, 1 ml of NaOH extract of bixin was added in a stirring beaker containing 19 ml sodium

hydroxide solution at 50oC for 120 min for measuring the degradation of bixin in the extraction

by sodium hydroxide solution The determination of bixin lost in the acetone extracted method

was carried out by adding 1 ml of acetone extract of bixin in Soxhlet extractor, which content

acetone, and then this apparatus was operated for 120 min Each solution of these processes was

measured every 20 min to determine the decrease of bixin content in solutions during heating

time

6 Analysis of bixin content in annatto extracts

The bixin content in the extracts was determined by measuring absorbance of bixin

solutions in a spectrophotometer at 487 nm (Smith, 2006) One milliliter of bixin extract was

dissolved in 10 ml tetrahyrofuran in 100ml-volumetric flask and then this solution was diluted by

acetone to achieve a solution with absorbance in the range from 0.2 to 1.0 Absorbance of the

diluted solutions was measured in 1-cm cuvet by a spectrophotometer (Optizen 2120UV,

Mecasys Co.Ltd, Daejeon, South Korea) at 487 nm The concentration of bixin in solution was

calculated base on the extinction coefficient (E1%1cm) of 3.090, according to following formula:

7 Gas chromatography (GC) analysis of toluene and m-xylene

7.1 Sample preparation

One milliliter of extract was accurately transferred into a Falcon tube fitted with a

screw-cap Methanolic KOH solution (10% (w/v), 2 mL) and o-xylene recovery standard (10 mg/L in

pentane, 2 mL) were added The Falcon tube was capped, wrapped in aluminum foil, and

roll-mixed for 4 to 6 h The pentane phase was transferred to a second Falcon tube and the

methanolic phase re-extracted with 2 x 1 mL portions of pentane by recapping the Falcon tube

and shaking for about 1 min The emulsions were broken by centrifugation at 4000 rpm for 10

min The pentane extracts were pooled in the second Falcon tube and washed with 2 x 1 mL

portions of water, which was removed via a Pasteur pipette The water washings were

back-Bixin concentration (%)

vial and dried by anhydrous sodium sulfate The combined pentane extract was added internal

standard solution (ethylbenzene in pentane) and diluted to achieve a solution containing 10 mg/L

of ethylbenzene This solution was then analyzed by a gas chromatography system (Scotter,

2000)

7.2 Gas chromatography system

The system comprised a HP 6890 series GC (Hewlett Packard Co.Ltd, California, USA)

fitted with a capillary column, Omegawax 320 (30 m length, 0.32 mm inner diameter, and 25 µm

film thickness) Helium carrier gas was maintained at 19 cm/s Split mode injection (20:1) was

used at a temperature of 250 °C, with an injection volume of 4 µL Flame ionization detector at

250 °C was used for the detection of the volatile compounds The column temperature was

programmed to increase from 60 - 75 °C at 15 °C/min and then raised to 90°C at 3°C/min

Calibration was established by injection of mixed standards prepared in pentane with the range

of concentration from 1 to 20 mg/L, containing a fixed amount of ethylbenzene internal standard

at 10 mg/L Detector response of each component was normalized against ethylbenzene, and

calibration lines were constructed using response ratios Detector response was linear for each

component over this range with the same R-squared value (R2 = 0.998)

8 Data analysis

All experiments were performed in triplicates, and the results were expressed as the mean

± standard deviation The statistical significance was analyzed by one-way analysis of variance

(ANOVA) on Statgraphics Plus 3.0 statistical software LSD tests were used for comparisons

amongst values Differences were considered to be significantly different at p<0.05

2 ml o-xylene (10

mg/L in pentane)

1ml of sample

Tube with screw cap

Wrapped in aluminum foil and roll-mixed for 4-6 hours

2ml methanolic KOH solution (10% w/v)

Pentane phase

Transferred to a second tube

Methanolic phase

Re-extracted with 2 x 1 ml portions of pentane

Centrifuged at 4000 rpm for

10 minutes Washe with 2 x 1 ml portions

of water

Water removed via a Pasteur

pipet Water washings

Back-extracted with a fresh 1ml portion of pentane

Combined pentane extracts

Added ethylbenzene in pentane

Diluted to a suitable concentration

Pentane extracts

Dried by anhydrous sodium sulfate

Figure 3 Sample preparation procedure for analysis of toluene and m-xylene in

annatto extracts by gas chromatography

III RESULTS AND DISCUSSION

1 Moisture content and total bixin content of annatto seeds

Moisture and total bixin contents of annatto seeds during storage are presented in Table 4

During storage in the dark and cool place (in refrigerator), both moisture and bixin contents of

seeds decreased slowly Within one year of storage, the loss of moisture content was only 4.3%

and this reduction was only significant in first six months, in the second half of the storage time,

the moisture content had no significant change The reduction of bixin was even smaller than the

reduction of moisture, total bixin loss in one year was only 0.15%, from 1.34 to 1.19%

Moreover, no significant decreases in bixin content of seed were observed every three months

during storage time This negligible loss of bixin in annatto seeds is corresponding to the results

of a study of Balaswamy et al (2006) This showed a slight decrease of bixin content in annatto

seeds stored at room temperature in jute bags from 1.69% to 1.43% for 360 days of storage

2 Extraction yield of bixin from annatto seeds by three methods

The extraction yield of bixin in three extraction methods is shown in Fig 4 The amount

of bixin extracted by acetone was very high (44.0%) after only 20 min, and then it increased

rapidly to 76.9% at 80 min of extraction time before reducing the speed and reached to 81.3% at

the end of extraction process We also carried out a further measurement to determine extraction

yield of bixin at 140 min of extraction time, which achieved only slightly higher yield of bixin

but not significantly different (p<0.05) as compared to that at 120 min (data not shown) The

extraction process by soybean oil also achieved a quite high yield of bixin at the beginning of the

process However, it only increased slightly from 28.3% at 20 min to 30.2% at 40 min of the

process, and then decreased gradually to 24.7% at 120 min of extraction These results are

consistent with the report of Barnett (1957), who showed that only less than 30% of coloring

material was removed from annatto seed in the extraction by hot vegetable oil Although

obtained bixin of the extraction by sodium hydroxide solution was only 23.6% after 20 min, it

grew steadily during the process and reached to 45.4% at the end of the extraction

3 Degradation of bixin during three extraction processes

Fig 5 shows the loss of bixin during the extraction processes by soybean oil, acetone, and

sodium hydroxide solution The loss of bixin in the extraction by soybean oil was 29.7% after

only 20 min of the process and rose sharply to 63.4% at the end of extraction The large degraded

amount of bixin in this method can be explained by the very high temperature using in this

method (120oC) This result is consistent with a report of Reith and Gielen (1971), which showed

the percentage of degraded yellow products (C17 species) up to 40% of commercial oil soluble

annatto products Whereas, degraded amounts of bixin in both extraction processes by acetone

and sodium hydroxide solution were very low and increased slowly during extractions Moreover,

the degradations of bixin in both methods did not show any significant difference These values

were 4.1% and 2.7% after 20 min, and 12.3% and 14.5% at the end of processes for the

extractions by acetone and sodium hydroxide solution, respectively Although the extraction by

NaOH solution was carried out at lower temperature than extraction by acetone but the loss of

bixin was not different and even higher at the ending period of the process This result indicates

the effect of water activity to the stability of bixin as shown in the study of Maria et al (1995)

Table 4 Moisture and total bixin contents (%) of annatto seeds stored in a refrigerator

Storage time (month)

Extraction time (min)

Extraction by soybean oil Extraction by acetone Extraction by NaOH

Figure 4 Bixin yield during extraction processes by soybean oil, acetone, and sodium

hydroxide solution

Figure 5 Degradation of bixin during extraction processes with vegetable oil, acetone, and

sodium hydroxide solution

Extraction by soybean oil Extraction by acetone Extraction by NaOH

Extraction time (min)

4 Total amount of bixin extracted out from annatto seeds

The total amounts of bixin, which was freed out from annatto seeds into the extracts by

the three methods, are showed in Fig 6 After calculating the bixin yield and its loss, it was

observed that the factual percentage of bixin transferred from seed coat of annatto into acetone

extracts and soybean oil extracts really high Because of the very high bixin yield obtained from

the extraction by acetone, the total amount of bixin was free from seed by this method was the

highest among the methods throughout extraction time; it was 44% at the period of 20 min and

rose to 92.7% after 120 min This high extracted amount was the result of the strong solubility of

bixin in acetone and the continuously washed process of annatto seeds by pure solvent in the

Soxhlet extractor The total extracted amount of bixin by soybean oil was also significantly

higher than that of sodium hydroxide extraction (p<0.05) Unlike acetone extraction, total

extracted bixin still increased and reached to a high value at the end of extraction although the

extraction yied of soybean oil method was very low and decreased during process This can be

explained by the large amount of bixin lost in the process that reduced the obtained pigment

intensity of soybean oil extract Bixin amount extracted from annatto seed in this method was

40.2 and 67.4%, after 20 min and at the end of process, respectively

Figure 6 Total amount of bixin extracted out from annatto seeds by vegetable oil,

acetone, and sodium hydroxide solution

Extraction time (min)

Extraction by soybean oil Extraction by acetone Extraction by NaOH