INTRODUCTIONMayonnaise is an emulsion of oil in water. Therefore,dietary mayonnaise has a smaller dispersed stepand larger water content [1–3]. Mayonnaise consistsof 60–80% fat [4]. Conventionally, it contains eggyolk, oil, lemon juice or vinegar, and seasonings, e.g.salt, mustard, paprika, sweeteners, etc. Three maincomponents in mayonnaise perform as different phasesin the formulation: oil is the dispersed phase, water is thecontinuous phase, and egg yolk is the emulsifier [5, 6].Mayonnaise is fat-free if its fat level is at least 50%lower than that of standard mayonnaise; mayonnaiseis considered light if its fat level is 25% lower thanstandard [7].Eggs are a common mayonnaise emulsifierbecause their emulsifying properties are perfect formayonnaise production. However, the growing ratesof vegetarianism, egg allergy, heart diseases, andproduction costs make producers look for egg-freeformulation variants.Furthermore, plant-based diets have gainedpopularity not only due to the health benefits theypromise but as a way to reduce environmentalfootprint [8]. Therefore, new egg substitutes and eggfreeproducts are of great importance in vegetarianfood supplies [9]. In general, protein acts as a surfactantto reduce the surface tension between hydrophilicand lipophilic materials in food systems and stabilizeemulsions. Cashew nut protein isolates can serve asan egg alternative and a fat replacer agent due to theirexcellent emulsifying property [10]. However, cashewnuts are a much less popular plant protein, despite theirexcellent sensory and nutritional benefits [11].Copyright © 2022, Mohammed et al. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 InternationalLicense (http://creativecommons.org/licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix,transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.Foods and Raw Materials, 2022, vol. 10, no. 1E-ISSN 2310-9599ISSN 2308-405777Mohammed N. et al. Foods and Raw Materials, 2022, vol. 10, no. 1, pp. 76–85Several studies have evaluated plant-basedemulsifiers as potential substitutes for eggs. Chetanaet al. reported egg-free mayonnaise of rice bran oil andsesame oil produced by replacing egg with xanthangum [14]. Gaikwad et al. managed to replaceegg yolk with skim milk powder [15]. In anotherstudy, wheat germ protein isolate and xanthan gumsubstituted egg yolk to produce low-cholesterolmayonnaise with acceptable characteristics [16].Modified starch can also serve as an alternativeto fat and eggs in low-fat mayonnaise [17].Among vegetable oils, coconut oil obtained fromcoconut kernel (Cocos nucifera L.) was reported to haveantibacterial and antioxidant biological activities [18].Virgin coconut oil is widely used in other vegetable oilssince it has many health benefits. Virgin coconut oildecreases total cholesterol, triglycerides, phospholipidsand low-density lipoprotein (LDL) cholesterol, whileincreasing high-density lipoprotein (HDL) cholesterol inthe blood [19].Although many studies reported this or that kind ofegg-free mayonnaise produced from various oils andemulsifiers, none of them featured the combination ofcashew nut protein isolates, xanthan gum, and modifiedstarch. Consequently, the current study aims at selectingthe optimal concentration of emulsifiers of cashew nutprotein isolates, xanthan gum, and modified starch toproduce egg-free virgin coconut oil mayonnaise andcompare its properties with commercial mayonnaiseproducts.STUDY OBJECTS AND METHODSMaterials. Table 1 shows the ingredients of eggfreeLady’s Choice mayonnaise (Bangi, Selangor) usedas a reference sample. Xanthan gum, cashew nut proteinisolates, and modified starch (maize) were purchasedfrom Sigma-Aldrich Co. (St. Louis, MO, USA).Experimental design. The methods of responsesurface methodology and central composite design wereused with three independent variables of emulsifiers,namely cashew nut protein isolates (5–15%) (Xc),xanthan gum (0–1%) (Xx), and modified starch (0–0.5%)(Xm) (Table 2). Viscosity (Y1), stability (Y2), and firmness(Y3) served as response variables.Preparation of egg-free virgin coconut oilmayonnaise. The low-fat and egg-free mayonnaise-likeemulsion gel was prepared according to Mozafari et al.with some modifications [20]. Briefly, a fixed amountof distilled water, lemon juice, mustard, sugar, aceticacid, and salt (Table 1) were mixed in a blender 8010S(Waring Commercial Torrington, USA), at mediumspeed for 3 min to achieve a smooth and creamy coarsephaseemulsion. Virgin coconut oil was then graduallyadded to the coarse-phase emulsion, followed byemulsifiers, i.e. cashew nut protein isolates, modifiedstarch, and xanthan gum (Table 2). The mix (500 mL)was further homogenized at high speed for 2 min untilsmooth and creamy. All mayonnaise samples weretransferred into 500-mL sterilized glass jars, capped,tightly sealed, and kept at room temperature (25 ± 2°C)prior further analysis.Physicochemical properties. Physicochemicalproperties are given for the optimized formulation only.Viscosity. The viscosity measurement followedthe method developed by Makeri et al. [21]. It involveda rheometer HAAKE RheoStress RS600 (ThermoElectron Corporation, Karlsruhe, Germany) and aparallel stainless-steel plate with a 25-mm diameterat a 1-mm distance at 25°C. A sample of 10 mL wasloaded onto the plate with extreme carefulness toprevent emulsion softening. The excess sample wascarefully trimmed from the sensor edge with a thinblade [22]. The flow characteristics were determined ata temperature of 25°C and a shear rate of 1–100 s–1. Eachviscosity measurement was performed in triplicate, andmean ± SD values were plotted.Texture. The texture of the egg-free virgin coconutoil mayonnaise was determined using a texture analyser(XT2i, Surrey, UK) following the method describedin [23] with slight modifications. A total of 100 mg foreach sample was placed in round plastic containers ata depth of 30 mm. The texture was determined using aP/35-cylinder probe (Stable Micro System, Surrey, UK).The force was measured in compression mode at fixed75% strain at room temperature (25 ± 2°C). The testconditions included 10 mm penetration, 1 mm/s pre-testspeed, as well as 1 and 10 mm/s test speed. The testswere performed in triplicate, and the mean values weretabulated.Stability. The mayonnaise emulsion stability test wasbased on the amount of oil removed from the emulsionafter centrifugation [24]. Briefly, 1.5 g of the samplewas placed in a 25-mL centrifuge tube (Refrigeratedcentrifuge SIGMA 3-18K, Goettingen, Germany) andweighed (initial weight, F0). The sample was heatedfor 30 min at 80°C in a shaking water bath at 120 rpmto form emulsion. After heating, the emulsions werecentrifuged in a Thermo Sorvall Legend Micro 17micro-centrifuge (Thermo Science, Waltham, MA) forTable 1 Formulation for egg-free virgin coconut oilmayonnaiseIngredients AmountDistilled water, mL 32.20Virgin coconut oil, mL 32.20Lemon juice, mL 16.10Mustard, g 3.35Sugar, g 2.68Acetic acid, mL 2.68Salt, g 0.13 gCashew nut protein isolates, Xc, %* 5–15Xanthan gum, Xx, %* 0–0.1Modified starch, Xm, %* 0–0.5* % varies according to formulations generated using responsesurface methodology experimental design78Mohammed N. et al. Foods and Raw Materials, 2022, vol. 10, no. 1, pp. 76–855 min at 5000 rpm, and the top oil layer was extractedwith a long-needle syringe. The precipitated fraction(F1) was weighed, and the stability of the emulsions wasestimated using the equation below:Percentage of emulsion stability (%) = 𝐹𝐹1𝐹𝐹0 × 100 Wet − weight percentage (%) = (𝐴𝐴−𝐵𝐵)𝐴𝐴 × 100 Protein content ( g100g) = Fat content (g100g) =𝑊𝑊1 − 𝑊𝑊2𝑊𝑊1× 100 Ash percentage (%) =(a + b) − bc× 100(1)where F0 is the Initial weight; F1 is the weight of theprecipitated fraction.Water activity. The water activity test followedthe calibration procedure. The sample cup was filledhalfway with 3 g of mayonnaise sample using anAquaLab water activity meter (Model 3TE, DecagonDevices, USA). The sample chamber lid was sealed toreach vapor equilibrium. The dew point/temperaturewas later translated into water activity (Aw) reading.pH measurement. The pH values were assessed byusing a pH meter (S210 Seven compact, Mettler-ToledoInstrument Co., Ltd., Shanghai, China) at 25°C. The pHmeter was adjusted at pH 7.01, 4.01, and 10.01 buffersolutions. The pH values were presented as a mean ofthree readings for one sample.Proximate analysis. Moisture content. Themoisture content was determined using the methoddeveloped by the Association of Official AnalyticalChemists (AOAC) [25]. A sample of 5 g was put into acovered crucible and placed into a Memmert 800 oven(Schwabach, Germany). There it stayed for at least 7 hat 105°C; the temperature of the oven was constant.The crucible and its cover were set on the balance andweighed quickly and accurately. The weighing processwas repeated to obtain constant weight. The moisturecontent was calculated based on the percentage of wetweight:Percentage of emulsion stability (%) = 𝐹𝐹1𝐹𝐹0 × 100 Wet − weight percentage (%) = (𝐴𝐴−𝐵𝐵)𝐴𝐴 × 100 Protein content Fat content (g100g) =𝑊𝑊1 − 𝑊𝑊2𝑊𝑊1× 100 Ash percentage (%) =(a + b) − c(2)where A is the weight of sample before oven drying, g; Bis the weight of dried sample after oven drying, g.Protein content. This crude protein analysis methodwas designed by AOAC: it is based on the nitrogen (N)determination according to the Kjeldahl method in aKjeltec 2100 Distillation Unit (Foss Tecator, Hoganas,Sweden) [25]. The protein content was calculated usingthe following formula:stability (%) = 𝐹𝐹1𝐹𝐹0 × 100 Wet − weight percentage (%) = (𝐴𝐴−𝐵𝐵)𝐴𝐴 × 100 Protein content ( g100g) = Nitrogen content × 𝐹𝐹Fat content (g100g) =𝑊𝑊1 − 𝑊𝑊2𝑊𝑊1× 100 Ash percentage (%) =(a + b) − bc× 100(3)where F is the protein factor (6.25, depends on thesample).Fat content. The fat content was measured accordingto another AOAC method by petroleum ether extractionusing a Soxtec System (2055 Soxtec Avanti; FossTecator, Höganäs, Sweden) [25]. The fat content wascalculated by using the following formula:Percentage of emulsion stability (%) = 𝐹𝐹1𝐹𝐹0 × 100 Wet − weight percentage (%) = (𝐴𝐴−𝐵𝐵)𝐴𝐴 × 100 Fat content (g100g) =𝑊𝑊1 − 𝑊𝑊2𝑊𝑊1× 100 Ash percentage Percentage of emulsion stability (%) = 𝐹𝐹1𝐹𝐹0 × 100 Wet − weight percentage (%) = (𝐴𝐴−𝐵𝐵)𝐴𝐴 × 100 Protein content Fat content (g100g𝑊𝑊1 − 𝑊𝑊2𝑊𝑊1× 100 Ash percentage (%) =(a + b) cPercentage of emulsion stability (%) = 𝐹𝐹1𝐹𝐹0 × 100 Wet − weight percentage (%) = (𝐴𝐴−𝐵𝐵)𝐴𝐴 × 100 Protein Fat content (g100g) =𝑊𝑊1 − 𝑊𝑊2𝑊𝑊1× 100 Ash percentage (%) 10              (4) where W1 is the sample weight, g; W2 is the plainaluminum weight, g; W3 is the aluminum with sampleweight, g.Ash content. The ash content was measuredaccording to AOAC method: 10 g of the sample wereplaced into the crucible [25]. After recording the weight,it was put into a muffle furnace at 550°C. The sampleburned for at least 2 h to obtain permanent weight,until no black particles. Next, the crucible and ashwere cooled in the desiccators. Finally, the crucible wasweighed together with the ash.Percentage of emulsion stability (%) = 𝐹𝐹1𝐹𝐹0 × 100 Wet − weight percentage (%) = (𝐴𝐴−𝐵𝐵)𝐴𝐴 × 100 Protein content ( g100g) = Nitrogen content × Fat content (g100g) =𝑊𝑊1 − 𝑊𝑊2𝑊𝑊1× 100 Ash percentage (%) =(a + b) − bc× 100 ( 5 )where a is the weight of ash; b is the weight of crucible;c is the weight of sample.Carbohydrates content. The carbohydrate contentwas determined by extracting the protein, fat, moisture,and ash amount from 100%.Statistical analysis. Minitab 17.0 (Minitab, Inc,State College Pennsylvania, USA) was used foroptimization. The software programmed a face-centeredcomposite design with three independent variables,namely cashew nut protein isolates (Xc), xanthan gum(Xx), and modified starch (Xm) at three coded levels (–1,0, +1). The experiment involved six replicates at thecenter stage, with a total design of 20 experimental runsper sample. As a result, the effect of the two independentvariables on the response surface was obtained as 3-Dgraphs of response. The polynomial regression modelequation was used to define the performance of theresponse surface. The generalized response surfacemodel looked as follows:Table 2 Response surface methodology experimental designof the three independent variables in egg-free mayonnaiseformulationsRunorderBlock Cashew nutprotein isolateXanthangumModifiedstarch(Xc) (Xx) (Xm)1(c) 3 10(0) 0.5(0) 0.25(0)2(c) 3 10(0) 0.5(0) 0.25(0)3 3 15(1) 0.5(0) 0.25(0)4 3 15(1) 1(1) 0.5(1)5 3 5(–1) 0(–1) 0(–1)6(c) 3 10(0) 0.5(0) 0.25(0)7 3 5(–1) 0(–1) 0.5(1)8 3 15(1) 1(1) 0(–1)9(c) 1 10(0) 0.5(0) 0.25(0)10 1 10(0) 0.5(0) 0.5(1)11 1 10(0) 0.5(0) 0(–1)12(c) 1 10(0) 0.5(0) 0.25(0)13 1 5(–1) 1(1) 0.5(1)14 1 15(1) 0(–1) 0.5(1)15 2 10(0) 0(–1) 0.25(0)16(c) 2 10(0) 0.5(0) 0.25(0)17 2 5(–1) 1(1) 0(–1)18 2 10(0 1(1) 0.25(0)19 2 5(–1) 0.5(0) 0.25(0)20 2 15 (1) 0(–1) 0(–1)c is center point79Mohammed N. et al. Foods and Raw Materials, 2022, vol. 10, no. 1, pp. 76–85y= ß0 + ß1x1 + ß2 x2 + ß11 x12 + ß22 x22 + ß12 x1 x2 (6)where y is the response calculated by the model; ß0 isthe constant regression; ßi, ßii, ß1j are the linear, squared,and interaction coefficients, respectively; x1, x2 are theindependent variables.The responses were evaluated by multipleregressions and the square least method. A t-test wasperformed to compare the properties of both mayonnaisesamples.To validate the model, the experimental data werecompared to the predicted values using the t-test atP-value = 1 and F-ratio = 0 for each response. Therefore,the model was declared suitable when no statisticallysignificant difference was observed between theexperimental and predicted values.RESULTS AND DISCUSSIONResponse surface methodology. The goal ofthe optimization was to obtain target values forresponses, viscosity, and firmness, as well as tomaximize the stability of the emulsion. The initialstep was to decide on the experimental ranges for theindependent variables. The levels of variation wereselected according to a preliminary study. A uniformprecision type central-composite design consisted ofthree variables, namely cashew nut protein isolates,xanthan gum, and modified starch. It had a threelevelpattern with 20 runs and was prepared usingthe response surface methodology. The experimentaldesign contained six cube center points, where six out oftwenty runs were replications of the central points of allthe factors. Twenty samples of egg-free virgin coconutoil mayonnaise were prepared based on the emulsifierquantity proposed in the experimental design. Otheringredients remained constant.All twenty samples were measured for viscosity,stability, and firmness. Table 3 displays the variables,levels, and results obtained for all the responses.The analysis of variance was used to determine thesignificance of the linear, quadratic, and interactioneffects, as well as the lack of fit value against theresponses in the variables. The models fit well for all theresponse variables because they had acceptable levels ofR2 of more than 80%.Table 4 illustrates the summary of R2, %, P-value,and multiple regression equation of response for reducedregression equation model in the decoded units. Thebest model was the one with the highest R2, lowestP-value (model), and the highest number of significantfactors. The emulsifiers were optimized by identifyingthe desired response. The anticipated responses weredesignated based on the viscosity, stability, and firmnessof commercial mayonnaise. These properties are knownto be accepted by consumers. The reference mayonnaiseunderwent an analysis to obtain the desired response.The lack-of-fit in all the models had a P-value ≥ 0.05,i.e. the models were acceptable. The next step involvedthe P-value of individual factors of the quadratic andinteraction effect against response. The factors withinsignificant effects were removed to obtain a fittedreduced model equation.In this study, Xc, Xx, and Xm were coded values forindependent variables in the experiment, i.e. cashewTable 3 Viscosity, stability, and firmness of egg-free virgin coconut oil mayonnaise produced with different percentages of cashewnut protein isolate, xanthan gum, and modified starchRun Order Cashew nut protein isolate Xanthan gum Modified starch Viscosity, mPa·s Stability, % Firmness, g(Xc) (Xx) (Xm) (Yv) (Ys) (Yf)1 10 0.5 0.3 104.2 ± 11.4 95.2 ± 2.2 24.6 ± 3.72 10 0.5 0.3 98.2 ± 4.9 89.3 ± 0.6 25.8 ± 1.63 15 0.5 0.3 101.6 ± 6.9 92.9 ± 1.0 22.3 ± 3.04 15 1 0.5 120.3 ± 22.8 100.0 ± 0.0 21.1 ± 2.05 5 0 0 47.8 ± 5.7 81.8 ± 0.7 9.2 ± 2.26 10 0.5 0.3 100.3 ± 15.6 93.9 ± 1.4 30.0 ± 2.07 5 0 0.5 88.2 ± 3.9 93.4 ± 1.8 17.3 ± 1.58 15 1 0 92.1 ± 3.6 96.4 ± 0.1 10.8 ± 1.49 10 0.5 0.3 106.8 ± 6.5 94.3 ± 0.9 22.3 ± 2.210 10 0.5 0.5 127.8 ± 19.4 95.8 ± 0.2 30.8 ± 3.111 10 0.5 0 84.0 ± 6.8 93.1 ± 1.6 17.5 ± 1.612 10 0.5 0.3 107.9 ± 9.0 93.9 ± 0.4 21.1 ± 1.813 5 1 0.5 122.8 ± 14.3 100.0 ± 0.0 16.6 ± 1.714 15 0 0.5 97.8 ± 4.8 94.2 ± 2.0 19.3 ± 5.115 10 0 0.3 72.8 ± 7.3 87.2 ± 1.3 19.4 ± 2.016 10 0.5 0.3 103.4 ± 14.4 94.9 ± 1.6 27.6 ± 2.717 5 1 0 93.4 ± 3.6 97.4 ± 1.2 6.7 ± 1.218 10 1 0.3 123.0 ± 6.5 95.2 ± 0.2 27.7 ± 1.819 5 0.5 0.3 98.1 ± 7.6 89.8 ± 3.1 18.0 ± 1.520 15 0 0 44.1 ± 3.1 79.2 ± 0.9 13.1 ± 2.180Mohammed N. et al. Foods and Raw Materials, 2022, vol. 10, no. 1, pp. 76–85nut protein isolates, xanthan gum and modified starch,respectively. Likewise, Yv, Ys, and Yf were coded valuesfor viscosity, stability, and firmness dependent variables.All the initial and reduced model multiple regressionequations used the code above.Effect of independent variables on viscosity (Yv).Viscosity measurement is essential to characterize thestructure and stability of the food emulsion products,such as mayonnaise. Figure 1a shows that both the linearand square effects were significant for viscosity, whilethe overall model P-value was < 0.05 for both. Theinteraction effect of cashew nut protein isolates withxanthan gum and modified starch was not significant,with a P-value of 0.472 and 0.372, respectively.The analysis of regression coefficient showed thatviscosity experienced significant impact (P < 0.05)from the linear effect of cashew nut protein isolates (Xc),xanthan gum (Xx), modified starch (Xm), quadratic effectcashew nut – cashew nut (Xc·Xc), xanthan gum – xanthangum (Xx·Xx), and interaction effect xanthan gum –modified starch (Xx·Xm). The increased amount of theseTable 4 Summary of reduced model equation for all responsesResponse R2, % P-value Reduced model equationViscosity, mPa·s 97.5 0.00 Yv = 22.76 + 5.64 Xc + 84.9 Xx + 96.48 Xm – 0.2760 Xc·Xc – 35.56 Xx·Xx – 36.5 Xg·XmStability, % 88.7 0.00 Ys = 81.89 – 0.033 Xc + 15.74 Xx + 24.42 Xm – 20.40 Xx·XmFirmness, g 80.9 0.00 Yf = –21.18 + 7.74 Xc + 0.97 Xx + 19.10 Xm – 0.3683 Xc·Xc805101001201.00.50.015ViscosityXGCN801005101200.20.0150.4ViscosityMSCN50751000.00.51250.40.20.01.0ViscosityMSXGSurface Plots of Viscosity805101001201.00.50.015ViscosityXGCN801005101200.20.0150.4ViscosityMSCN50751000.00.51250.40.20.01.0ViscosityMSXGSurface Plots of Viscosity801005101201.00.50.015ViscosityXGCN801005101200.20.0150.4ViscosityMSCN50751000.00.51250.40.20.01.0ViscosityMSXGSurface Plots of Viscosityа90955101000.50.0151.0StabilityXGCN909294510960.20.0150.4StabilityMSCN85900.00.5951000.40.20.01.0StabilityMSXGSurface Plots of Stability90955101000.50.0151.0StabilityXGCN909294510960.20.0150.4StabilityMSCN85900.00.5951000.40.20.01.0StabilityMSXGPlots of Stability90955101000.50.0151.0StabilityXGCN909294596Stability85900.00.5951000.40.20.01.0StabilityMSXGSurface Plots of Stabilityb15205 10251.00.50.015FirmnessXGCN10205 10300.40.20.015FirmnessMSCN20250.0 0.5300.20.01.00.4FirmnessMSXGSurface Plots of Firmness15205 10251.00.50.015FirmnessXGCN10205 10300.40.20.015FirmnessMSCN20250.0 0.5300.20.00.4FirmnessMSSurface Plots of Firmness15205 10251.00.50.015FirmnessXGCN10205 30Firmness20250.0 0.5300.20.01.00.4FirmnessMSXGSurface Plots of FirmnesscFigure 1 Surface plots of viscosity (a), solubility (b), and firmness (c) changes in low-fat egg-free mayonnaise by formulationparametersvariables resulted in increased viscosity. The reducedmodel equation for viscosity was predicted as below:Yv = 22.76 + 5.64 Xc + 84.9 Xx + 96.48 Xm –– 0.2760 Xc·Xc – 35.56 Xx·Xx – 36.5 Xx·Xm (7)The equation above was fitted using a seconddegreepolynomial model for independent variableeffects of cashew nut protein isolates, xanthan gum,and modified starch on apparent viscosity response. Themodified value (R2 = 97.5) proved that more than 97%of the experimental points were adequate independentvariables.The highest viscosity reading obtained was127.8 ± 19.4 mPa·s, and the lowest was 44.1 ± 3.1 mPa·s.Among all three factors, xanthan gum had the mostsignificant effect on viscosity. This finding was similarto the results obtained by Mozafari et al., who found thatxanthan affected the viscosity of low-fat low-cholesterolmayonnaise [20].Also, Kumar et al. illustrated that xanthan gumsignificantly impacted the viscosity of egg-free81Mohammed N. et al. Foods and Raw Materials, 2022, vol. 10, no. 1, pp. 76–85mayonnaise produced by ultrasonication [26]. Theexperimental outcomes and contour plots showed thata larger amount of xanthan gum followed by modifiedstarch improved the viscosity of mayonnaise samples.For standard oil, the viscosity and flow behavior in wateremulsion was captivated by the dispersed phase andcontrolled by the hydrophilic additives, such as sugar,salt, and polymeric thickener [27].Effect of independent variables on stability (Ys).Egg-free virgin coconut oil mayonnaise samples provedmoderate to high stability, depending on the emulsifierused in the formulation. The linear effect of cashew nutprotein isolates (Xc), xanthan gum (Xx), modified starch(Xm), and a combination of xanthan gum and modifiedstarch (Xx·Xm) had significant effects on the stability ofegg-free virgin coconut oil mayonnaise. The remainingfactors proved insignificant (P-value > 0.05) and wereremoved. A reduced model equation for stability waspredicted as below:YS = 81.89 – 0.033 Xc + 15.74 Xx ++ 24.42 Xm – 20.40 Xx·Xm (8)The stability of samples was 100% for formulations4 and 13, which had the maximal amount of emulsifier.This result was similar to the findings obtained byMozafari et al., who achieved a good stability of low-fatlow-cholesterol mayonnaise with the maximal amountof xanthan gum and Zodo gum as emulsifiers [20]. Inthis study, the formulations of egg-free virgin coconutoil mayonnaise with xanthan gum and modified starchhad higher emulsion stability than the control samples.Lee et al. reported similar findings in their study of lowfatmayonnaise with gelatinized rice starch and xanthangum [24].Two formulations demonstrated a much lowerstability, namely formulation 5 (cashew nut proteinisolates (Xc) – 5, Xanthan gum (Xx) – 0, Modifiedstarch (Xm) – 0) and formulation 20 (cashew nut proteinisolates (Xc) – 15, Xanthan gum (Xx) – 0, Modified starch(Xm) – 0). The stability was 81.8 ± 0.7 and 79.2 ± 0.9%,respectively.Therefore, cashew nut protein isolates had an almostnegligible effect as natural emulsifiers on the stabilityof the emulsion. In addition, the percentage of virgincoconut oil used in this formulation was approximatelyonly 30–31%. This indicates that emulsion stabilitywas affected by the biopolymers used in the system.According to Lee et al., a lower amount of oil resultedin a significant decline in mayonnaise stability [24].Therefore, such biopolymers as starches and gumshave to be combined with such fat-reduced formulationproducts as stabilizers.Effect of independent variables on firmness (Yf).According to Khor et al., firmness is the product’sability to resist deformation or breaking and increaseswith the force required for penetration [28]. Higherfirmness of emulsion makes it difficult for the mouth tobreak the sample and swallow. The interactions betweenproteins and oils in a network structure are known toincrease mayonnaise firmness [29].Based on the P-value, all linear (Xc, Xx, and Xm) andquadratic effects of cashew nut protein isolates (Xc·Xc)had a significant impact (P < 0.05) on the firmness. Thebest reduced model equation for predicting firmness wasas follows:Yf = –21.18 + 7.74 Xc + 0.97 Xx ++ 19.10 Xm – 0.3683 Xc·Xc (9)In this study, fat content in egg-free virgin coconutoil mayonnaise was 30%, which was lower than thatin whole fat mayonnaise (70%). Such reduction of fatcontent caused a lower droplet density, which affectedthe emulsion stability by weakening the interactionsbetween droplets. However, such lower oil contentincreased the aqueous phase and decreased the dispersedphase, which reduced the firmness and viscosity of theemulsion [30]. Singla et al. reported similar findings: ahigher amount of xanthan gum with maltodextrin asthickener increased firmness and stickiness values [31].Response optimization and model validation. Agraphical optimization (Fig. 2) was performed usingMinitab 16 package to optimize the percentage ofFigure 2 Response optimizationCurHigh0.89643 LowDOptimald = 0.98947Targ: 120.0Viscosity = 120.1579d = 0.72917MaximumStability = 98.6458d = 0.99842Targ: 25.0Firmnessy = 24.97780.89643DesirabilityComposite0.00.500.01.05.015.0CN XG MS[13.0266] [1.0] [0.3586]Table 5 Optimal values of emulsifiers (factors) derivedthrough response surface methodologyFactor Optimizedvalue, gPercentage informulation, %Cashew nut protein isolates 13.0 12.6Xanthan Gum 1.0 1.0Modified Starch 0.4 0.482Mohammed N. et al. Foods and Raw Materials, 2022, vol. 10, no. 1, pp. 76–85emulsifier. The optimal values of emulsifiers were 13.0 gfor cashew nut protein isolates, 1.0 g for xanthan gum,and 0.36 g for modified starch (Table 5). The desiredresponse required the highest amount of xanthan gum.Table 6 illustrates the predicted optimal andexperimental values of response, viscosity, stability, andfirmness. Based on the two-sample t-test, the P-valuefor all responses was > 0.05. Statistically, there was nosignificant difference between the experimental andpredicted values. Thus, the model and the reduced modelequations were validated and accepted.Proximate analysis and physicochemicalproperties. Table 7 shows the proximate analysis andphysicochemical properties of the optimal formulationof egg-free virgin coconut oil mayonnaise and referencesamples. They demonstrated a significant difference(P < 0.05) in fat content, protein content, water activity,and consistency. In the egg-free virgin coconut oilmayonnaise, fat content, water activity, and consistencywere significantly lower, whereas the protein contentwas higher compared to the reference product. However,the comparative analysis showed no significantdifference in terms of viscosity, stability, firmness,cohesiveness, pH, moisture content, ash content, andcarbohydrate content.Singla et al. compared the firmness of the standardand the egg-free mayonnaise samples, and the egg-freemayonnaise showed a higher firmness [31]. However, thehigh-fat content in the standard mayonnaise caused anincrement in textural firmness and stickiness by keepingthe neighboring oil droplets flocculated to form a thingel network.In this study, thickeners enhanced the firmnessand stickiness values in the egg-free virgin coconut oilmayonnaise compared with the egg-containing sample.Generally, the texture of mayonnaise depends on theingredient selection and the effect of thickening agentsused in the system.The pH of the egg-free virgin coconut oilmayonnaise was acidic, and pH 4 was similar to thatof the reference mayonnaise. The acidic emulsion isformed when adding lemon juice or vinegar. Acidicstate extends the shelf life of the product and ensures itsmicrobiological stability [28].Based on [32, 33], mayonnaise producers favorhigher acidity because it improves the microbialstability, emulsion stability, and viscoelasticityproperties. Moisture content is a significant factor as itaffects stability and shelf life. The moisture content inthe sample produced by applying the optimal conditionswas 34.7 ± 2.9%, while for the commercial sample it was35.8 ± 4.3%, which indicated no significant differences.This result could be due to the similar content ofsolid materials used to formulate the egg-free virgincoconut oil mayonnaise. The water activity of the eggfreevirgin coconut oil mayonnaise was significantlylower compared to reference sample. Even though thepercentage of water was higher in this formulation, ahigher amount of emulsifier was expected to bind allthe molecules to obtain properties similar to standardmayonnaise.Ash content was 3.3 ± 0.5% for the egg-freevirgin coconut oil mayonnaise and 3.6 ± 1.1% for thecommercial sample. The differences between theseresults might be due to the different ingredients appliedfor the production.The protein content of the egg-free virgin coconutoil mayonnaise was 2.6 ± 0.2 g, which was higher thanthe labeled value of commercial sample (1.4 g). Thisis primarily because of the protein-based emulsifiersused in the formulation. The carbohydrate contentof the egg-free virgin coconut oil mayonnaise was14.0 ± 3.7 g, which was higher than in the labeled valueof commercial sample (9.2 g). This result also could bedue to the differences in the formulations.The fat content of the egg-free virgin coconut oilmayonnaise was 27.5 ± 3.6 g/100 g, whereas for thereference mayonnaise it was 66.2 g/100 g. This resultwas expected because the experimental low-fat egglessmayonnaise contained 30% of fat, while the commercialsample was a whole-fat mayonnaise. Standardmayonnaise formulation includes 60–80% of fat,depending on the composition and type of oil [33, 34].Table 6 Predicted optimal value and experimental valuesof responseResponse ExperimentalvaluePredictedvalueP-valueViscosity, mPa·s 102.4 120.2 0.1Stability, % 99.5 98.7 0.1Firmness, g 21.8 25.0 0.2* P-values < 0.05 are significant differences using Tukey Method testbetween experimental value and predicted valueTable 7 Proximate analysis and physicochemical propertiesof optimal formulation of egg-free virgin coconut oilmayonnaise and reference sampleAnalysis Egg-free virgincoconut oilmayonnaiseReferencemayonnaiseP-valueViscosity, mPa·s 102.4 ± 4.1a 121.1 ± 16.0b 0.2Stability, % 99.5 ± 0.3a 99.7 ± 0.2a 0.2Firmness, g 21.8 ± 1.5a 25.3 ± 5.1a 0.3Water activity 1.0 ± 0.0a 1.0 ± 0.0a 0.0pH 4.0 ± 0.0a 4.0 ± 0.0b 0.2Moisture content, % 34.7 ± 2.9a 35.8 ± 4.3a 1.0Ash content, % 3.3 ± 0.5a 3.6 ± 1.1b 1.0Protein content,g/100 g2.6 ± 0.2 1.4* 0.0Carbohydratecontent, g/100 g14.0 ± 3.7 9.2* 0.1Fat content, g/100 g 27.5± 3.6 66.2* 0.0*Values obtained from product nutritional information83Mohammed N. et al. Foods and Raw Materials, 2022, vol. 10, no. 1, pp. 76–85Therefore, a lower amount of oil in the formulationresulted in a lower fat content.CONCLUSIONThe research objective was to improve theapplication of egg replacers in low-fat virgin coconut oilmayonnaise using response surface methodology. Theoptimal combination of three independent variables wasas follows: cashew nut protein isolates – 12.6%, xanthangum – 1.0%, and modified starch – 0.3%. We produceda high-quality egg-free virgin coconut oil mayonnaisewith optimal viscosity, stability, and firmness. Thepredicted response values under the defined optimallevels were generally in accordance with the model. Theproximate analysis and physicochemical properties ofthe egg-free virgin coconut oil mayonnaise had a lowerfat content, water activity, and consistency, as well as ahigher protein content compared to the reference sample.Therefore, a mix of cashew nut protein isolates,xanthan gum, and modified starch at optimal levelscould be used as a plant-based substitute to improvethe viscosity, texture characteristics, and stability ofmayonnaise. More investigations are required to assessthe sensory properties and storage stability of the eggfreevirgin coconut oil mayonnaise, which could be agood product for vegan consumers.CONTRIBUTIONNameer Khairullah Mohammed performed theexperiments, drafted the manuscript, and proofreadthe article. Hemala Ragavan developed the researchconcept, performed the formal analysis, worked with thesoftware, and drafted the article. Nurul Hawa Ahmadperformed the data validation, wrote the review, andedited the manuscript. Anis Shobirin Meor Hussinsupervised the project, developed the methodology, andacquired the funding. The manuscript was checked andapproved by all the authors. All authors have read andagreed to the published version of the manuscript.CONFLICT OF INTERESTThe authors declare no conflict of interests regardingthe publication of this article.