QUALITY CHARACTERISTICS OF A MODEL BISCUIT CONTAINING PROCESSED MANGO (MANGIFERA INDICA) KERNEL FLOUR

ABSTRACT

Mango (Mangifera indica) kernel was processed to flour and incorporated at 50% level of wheat flour substitution into biscuit (PMKWB) preparation. PMKWB was comparatively analysed with a commercial whole wheat biscuit (BRSCB) for several storage characteristics at 31±2°C and rate of peroxidation at 14, 31 and 58±2°C. Results showed that both biscuits had similar composition except in fat content. PMKWB was slightly more acidic than BRSCB due to the associated polyphenolic matter in the processed kernel flour used. BRSCB had more but negligible microbial load than PMKWB. On a 9-point hedonic scale, both biscuits were scored between 6.0 and 8.5 for attributes assessed though BRSCB scores were statistically higher (p<0.05) than those of PMKWB. Both biscuits exhibited sigmoid-shaped moisture adsorption isotherms with calculated monolayer moisture values of 4.0–4.5% (dry basis) using Brunauer–Emmett–Teller (BET) equation, and upper stability limits of 8.5–8.7% using the Henderson equation (rel. humidity, ca 52%). Both mathematical models fitted the biscuit isotherms well (rel. deviation, 3.0 to 4.5%). Peroxide values on the 90th day were 13, 21 and 56 meq/kg of PMKWB sample at 31, 14 and 58°C respectively. It was concluded that hygienically and commercially produced PMKWB would be shelf-stable for 3 months at temperatures not exceeding 31°C.

INTRODUCTION

Biscuit is a popular snack world-wide. Wheat flour serves as the main and conventional ingredient in its preparation. The increasing choice of biscuit among other snacks being served at meetings and social gatherings in developing countries such as Nigeria partly describes the need for its increased availability. Unfortunately, the Nigerian tropical climate does not favor the growth of wheat. An alternative to the production of biscuits as well as promote backward integration has been through total or part-substitution of wheat flour with other raw materials.[1] Biscuits prepared from such composite flour programmes involved the use of rice, maize, sorghum, millet, peanut, pigeon pea, cowpea, and mango kernel singly or in combination with wheat.1-6

Our laboratory embarked on waste conversion programme of the abundant but discarded Nigerian mango (Mangifera indica) kernel. Earlier studies bordered on composition, toxicology, and functionality of the kernel before and after processing into flour,[7] the browning activity of its polyphenol oxidase,[8] characterization of the tannic constituents[9] and employing the processed flour in biscuit-formulation.[1] Recent work[[10],[11]] investigated the effect of temperature on the moisture absorption behavior of the biscuit, and the diversification of the flour in preparing some traditional Nigerian dishes in similar manner to efforts reported by Dhingra & Kapoor[12] on indigenous Indian dishes.

To further the study, the processed mango kernel–based biscuit (PMKWB) was re-produced and comparatively assessed with a commercial wheat biscuit (BRSCB) for chemical composition, moisture adsorption behavior, changes in microbiological and sensory properties after 90 days of stora ge at the ambient temperature of 31±21°C. The rate of peroxi dation of PMKWB lipid during the storage period was also monitored. The study, therefore, intends to elucidate the commercial viability of the model biscuit.

MATERIALS AND METHODS

Sample Procurement

Ripe mango (Mangifera indica) fruits were plucked directly from trees in the Polytechnic environ at Okenya–Idah, Kogi State during the March/April fruiting season. Farm fresh unsalted butter (WAMCO Plc., Nigeria) was purchased from a supermarket at Jos, Plateau State. Wheat flour, sugar, salt, egg, vanilla essence, and “Baker's Rich Shortcake” biscuit (Yale Foods Ltd., Ibadan) (BRSCB) were purchased from local supermarkets at Idah.

Biscuit (PMKWB) Production

Processed flour from the mango kernel (PMKF) was obtained by following the methodology described by Arogba.[7] The recipe and procedure described further by Arogba[1] using the 50:50 (PMKF: wheat) flour ratio were employed for the biscuit (PMKWB) production. In the present study, the flours were of not more than 0.6 mm granulation when passed through a standard sieve. Thereafter, PMKWB and the commercial BRSCB were comparatively analyzed except for peroxide value (on day 0, 14, 30, 60, and 90). The biscuit samples were stored at the ambient temperature or 31±2°C. Proximate analysis and isotherm study commenced on the ‘zero’ day and were conducted only once but other anlayses were conducted also at the end of storage period. Packets of biscuit taken out from storage on specific days for analyses were not returned for later use.

Chemical Analyses

Moisture content was determined by drying in an air-oven at 105°C to constant mass. Crude protein (N×6.25), fat, fibre, and ash were determined by AOAC[13] methods 14.086, 14.089, 14.087 and 14.085 respectively. Total carbohydrate content was calculated by difference. Aqueous filtrates of 10% (w/v) biscuit suspensions were used for pH determination using a digital meter (Toptronic, Milano-Italy) while chloroform extracts were employed in peroxide value determination.[14] Results were expressed as milliequivalents of peroxide oxygen per kg of biscuit sample. All determinations were conducted in duplicate.

Microbial Count

Standard methods described by Buchanan and Gibbons[15] were employed for bacteria and mold counts. Each biscuit sample (1 g) was blended and serially diluted in 1% peptone water. One ml of the dilutions were each cultured on duplicate plates of nutrient agar (Merck) for bacterial count, and malt extract agar (Oxoid CM59) for mold count using the pour plate technique. The plates were incubated at 37 and 32°C respectively, for 24 h. Colony counter (ALS) was used and results were expressed as mean of the counts.

Sensory Evaluation

A 15-member panel was randomly selected from people in the campus. The completely randomized experimented design[16] was employed. Each sample was evaluated for colour, crispness and flavour. Scoring was based on a nine-point preference scale of 9 (like extremely), 7 (like), 5 (neither like nor dislike), 3 (dislike), and 1 (dislike extremely). The least Significance Difference test[17] conducted at p=0.05 was used to separate the mean sensory scores.

Moisture Adsorption Isotherm (MAI)

MAI was determined gravimetrically by exposing one gramme of powdered biscuit sample to molal concentrations of sulphuric acid of know humidities, ranging from 11% to 97% as previously described by Arogba.[18] Triplicate samples of each biscuit type were stored in desiccators at the ambient temperature of 31±2°C. Gain in mass was monitored at regular intervals of 24 h until equilibrium moisture content (EMC) was attained.

The EMC data were fitted using two traditional isotherm models namely, Brunauer–Emmett–Teller (BET), and Henderson's models also previously described by Arogba.[18] The parameter values were derived from the linear regression equations of the two models. The monolayer and upper critical moisture contents for shelf-stability were evaluated at water activity (a_w) of up to 0.45 (BET) and up to 0.97 (Henderson) respectively.

The goodness of fit was evaluated and expressed as mean relative deviation (%E), using the equation described by Palou et al.[19]

RESULTS AND DISCUSSION

Chemical Considerations

The proximate compositions of the test (PMKWB) and commercial (BRSCB) biscuits are given in Table 1. These biscuits possessed similar moisture and protein contents. The 5% moisture of PMKWB and its calculated monolayer value of 3.9% discussed under moisture adsorption study below, implied that PMKWB had high nutritional integrity prior to storage. For a stored food commodity, periodic assessment of the moisture level has been described as a useful quality index of shelf-stability.[20] Arogba[7] reported a 9.4% protein content of 100% processed mango kernel flour. Despite its 50% level of substitution in PMKWB, the protein concentration of PMKWB was also comparable to that of BRSCB prepared from whole wheat flour. The observation formed the basis of the optimum protein supplementation earlier identified by Arogba.[1] PMKWB had higher fat content than BRSCB. This was ascribed to the milling characteristics of the flours. for instance, PMKWB was prepared from fine flour particles with very low fiber and negligible ash contents Table 1. However, the higher acidity of PMKWB compared with BRSCB had been associated with the presence of reducing substances such as tannin in the processed mango kernel flour. [1] The rate of peroxidation of the packaged PMKWB was assessed at storage temperatures of 14, 31 and 58°C (Fig. 1 and Table 4). In the first 15 days, the peroxide value (PV) at these temperatures were very low (≤6.5 meq). The induction period for the oil of the biscuit appeared to last for a further 15 days. However, the sample stored at the ambient temperature of 31°C still exhibited a PV of about 8.5 meq on the 60th day. On the 60th day, the samples stored at 14 and 58°C exhibited similar magnitude of deterioration and higher than the sample stored at 31°C. The observation was suggestive of two modes of rancidity, possibly influenced by differences in the relative humidity of the storage environments. For instance, the refrigerated sample at 14°C and that in the wet-heated environment at 58°C were less crispy, signifying higher moisture gain (that is, higher relative humidity equivalent) which could promote hydrolytic rancidity. As opposed to water vapour transfer through the polythene film package, more oxygen transfer from the air to the sample stored at the ambient temperature of 31°C with assumed lower relative humidity, would promote oxidative rancidity.

Table 1. Acidity and Percentage Proximate Composition of Biscuits

Property	Commercial BRSCB	PMKWB^a
pH	5.63±0.07	4.10±0.04
Moisture	5.6±0.1	5.0±0.2
Protein (N×6.25)	4.8±0.4	4.0±0.3
Fat	18.8±0.09	29.5±0.5
Fiber	0.4±0.03	0.01
Ash	0.08±0.01	Trace^b
Total Carbohydrate^c	70.3	61.4

All results are means of triplicate determinations and are expressed on dry matter basis.

^aBiscuit (PMKWB) contained processed mango kernel and wheat flours in ratio (1:1) and of ≤0.6 mm granulation.

^bLess than 0.005%.

^cCalculated by difference.

Figure 1. Changes in the peroxide value of PMKWB at different storage temperatures.

Table 4. Peroxide Value During Storage

	PEROXIDE VALUE (meq Peroxide O2 per kg Biscuit)
STORAGE PERIOD (Days)	14°C	31°C	58°C
0	0.5	0.5	0.5
15	6.5	4.0	5.0
30	7.8	4.5	10.5
60	14.9	8.8	13.5
90	21.3	13.5	55.8

The influence of temperature on the rate of deterioration of PMKWB became very significant by the 90th day, with PV increasing in the order for samples stored at 31, 14, and 58°C respectively. Thus, it was inferred that high temperatures catalyse hydrolytic rancidity while relative humidity appeared to primarily determine the mode of chemical deterioration as hydrolytic or oxidative.

Biological Changes

The results of viable counts are shown in Table 2. There was no visible growth of microorganisms on PMKWB sample cultured on the ‘zero’ day, suggesting that the quality of the processed mango kernel flour and other ingredients used in the biscuit production was high. Furthermore, abuse is post-production handling prior to packaging appeared to be negligible.

Table 2. Mean Microbial Counts of Biscuits Stored for 90 Days at 31±2°C

		Microbial Count (cfu/g)
Storage Time (day)	Biscuit^a	Bacteria	Mold
0	PMKWB	Nil	Nil
	BRSCB	3.0×10^1	TsTc^b
90	PMKWB	1.37×10^1	TsTc
	BRSCB	4.5×10^1	TsTc

^aPMKWB=Processed mango kernel-wheat biscuit (1:1 flour basis) in polythene pack of 0.038 mm thickness; BRSCB=Commercial sample of Baker's Rich Shortcake biscuit in polypropylene pack of 0.018 mm thickness.

^bTsTc=Too Small to Count, that is, Not Significant

At the time of purchase, the actual date of production of the commercial BRSCB was unknown. However, since the sample was serially diluted, the bacterial and mold counts on the ‘zero’ day indicated lowplate counts (TsTc). Nevertheless, this result sharply contrasted that of PMKWB, possibly due to storage abuse[21] prior to our purchase.

Both the PMKWB and BRSCB exhibited similar microbial growth rate in the 90 days of storage, with BRSCB being three-fold higher than PMKWB. Nevertheless, the results were too low to raise any serious alarm of potential public health hazard. Where the need arises, the use of benzoic acid as recommended for flours[[22],[23]] is advocated. Benzoic acid is water-soluble, colorless and ordorless. The advantage of PMKWB over BRSCB is the natural presence of residual tannins in the processed mango kernel flour[1] for similar preservative purpose. It was evident in the observed pH 4.1 for PMKWB and 5.6 for BRSCB.

Organoleptic Assessment

The mean sensory scores of the biscuits on the ‘zero’ and 90th days are shown in Table 3. The commercial sample (BRSCB) was liked more than the test sample (PMKWB) in respect of color, crispness and flavor, period of storage notwithstanding. Both biscuits had significantly different (p<0.05) overall mean scores of 8.0 and 6.6.

Table 3. Mean^x Sensory Scores of Biscuits Stored for 90 Days at 31±2°C

		Attributes
Storage Time (day)	Biscuit^y	Crispness	Color	Flavor	Overall Acceptability
0	PMKWB	6.5a	6.1a	7.2a	6.6a
	BRSCB	8.3b	8.1b	8.0b	8.1b
90	PMKWB	6.6a	6.1a	7.1a	6.6a
	BRSCB	7.2a	8.5b	8.1b	7.9b
Standard Error		0.41	0.64	0.26	0.41
Least Significant Difference		0.63	0.78	0.50	0.62

Scores in a column followed by different letters differ significantly at p<0.05.

^xMean Scores of 15 panelists using a 9-point preference scale, where 9=Like extremely, 7=Like, 5=Neither like nor dislike, 3=Dislike, and 1=Dislike extremely.

^yPMKWB=Processed mango kernel-wheat biscuit (1:1 flour basis); color=dull brown.

BRSCB=Commercial Baker's Shortcake biscuit; color=golden brown.

However, while the panelists retained similar level of appreciation for the three attributes of PMKWB assessed even after 90 days of storage at the ambient temperature, such appreciation decreased for the crispness of BRSCB significantly (p<0.05). The scores for crispness were not sufficient to affect the overall acceptability of BRSCB. It implied, however, that BRSCB was more hygroscopic than PMKWB. It was not clear in this study whether differences in the thickness of the packaging films (PMKWB, 38 micron; BRSCB, 18 micron) and/or the composition of the biscuit accounted for the observation.

Further difference in the biscuits was observed in the color. While visually PMKWB was dull brown, BRSCB was golden brown and consequently affected the scores given by the panelists. The observed color of PMKWB was attributed to the presence of residual tannic substances[[1],[7]] in the processed flour. The potential preservative role has been indicated above.

The high scores of more than 7.0 for the flavor of both biscuits were possibly due to similarity in the recipe. Sugar, egg and butter concentrations were to principally influence flavor perception.

Moisture Adsorption Isotherm

Table 5 and Fig. 2 shows the isotherms of PMKWB and BRSCB at 31±2°C. Both biscuits exhibited sigmoid-shaped isotherms, conventionally described as Type II. The isotherms nearly overlapped signifying similarity in the moisture adsorption behaviour of the biscuits, and was ascribed to similar chemical composition in respect of moisture, protein and carbohydrate contents (Table 1).

Table 5. Moisture Adsorption Isotherm

	EMC (kg H2O/kg Dry Matter)
aw	PMKWB	BRSCB
0.11	0.060	0.060
0.19	0.072	0.060
0.33	0.075	0.065
0.43	0.080	0.070
0.59	0.10	0.090
0.74	0.145	0.135
0.85	0.242	0.225
0.96	0.39	0.42

Figure 2. Moisture adsorption isotherms of biscuits PMKWB and BRSCB at 31±2°C.

Furthermore, Table 6 describes similar monolayer moisture contents for the biscuits as 4.18±0.28% (dry basis). The value is within the range of 3.78–5.33% for cookies studied between 20–35°C.[[19],[24],[25]] However, the great difference in the calculated values of constant C which is related to heat of absorption for both biscuits using the BET equation, reflects the severity of flour milling and heat treatment given during the baking of the biscuits. The lower the C value, the more exposed were the active absorption sites of the biscuit polymers during processing.[26]

Table 6. Parameter Values of BET and Henderson Models to Describe the Isotherms of Biscuits at 31±2°C

	BET^a			HENDERSON^b
				Moisture Content of
Biscuit	C	%m0	%E^0	Break	% Dry Basis	k2	n2	k1	n1	%E^c
PMKWB	18.51	3.90	4.50	0.94	8.7	0.3162	1.00	1.3×10^−4	4.22	4.16
BRSCB	70.65	4.48	3.03	0.95	8.5	0.3162	1.00	8.9×10^−4	3.21	2.93

^aUp to a_w 0.45.

^bUp to a_w 0.97.

^c%E=Mean relative deviation.

C=Constant related to heat of absorption.

m₀=monolayer moisture content expressed as % on dry weight basis.

break=Value on x-axis extrapolated from the juncture of two linear regression curves (not shown).

% dry basis=Equilibrium moisture content of biscuit corresponding to the value of ‘break’.

n and k=constants related to the biscuit.

The mathematical Henderson equation described the two biscuits as similar in several parameters analysed. Notably, the upper critical moisture content of 8.6±0.1% of the biscuits defined the moisture limit necessary for shelf-stability. The value, from Fig. 2, corresponded to a_w 0.54±0.02. Both the BET and Henderson equations described the isotherms of the biscuits very well with mean relative deviations (%E) of about 3.0 to 4.5%. Good mathematical models describe isotherms with %E of less than 7.0%.[27]

CONCLUSION

The storage studies have shown that biscuits prepared from whole wheat flour (BRSCB) and that substituted by 50% with processed mango kernel flour (PMKWB) were organoleptically acceptable for 3 months. Their rates of deterioration were influenced to similar extent by physical factors such as a_w and dry matter content. The PMKWB had a more acidic character than BRSCB and appeared to be more chemically resistant, more so, when stored in environments with low relative humidities and temperatures not exceeding 52% and 31°C respectively. Potential marketers of mango kernel-based baked confectioneries could, thus, be guided.

ACKNOWLEDGEMENT

The author sincerely appreciates the word-processing skill of Mr. Joshua B. Agbogun of Kogi State University, Anyigba, Nigeria.