Oleogels as spreadable fat and butter alternatives: sensory description and consumer perception

Abstract

Table spreads are one of the fundamental ingredients of human diets. Generally, butter and margarine are the most preferred spread products among others. Moreover, these products have some disadvantages such as high saturated fatty acids contents and the presence of trans fatty acids. In this study, virgin olive oil and hazelnut oil oleogels were prepared with beeswax and sunflower wax, while the hazelnut oil oleogels were aromatized with diacetyl. For the reason stated above, the first purpose of this study was to test the sensory properties and consumer acceptance of virgin olive oil when prepared as a spreadable fat in oleogel form. The second purpose was to determine how the hazelnut oil oleogels would be received as a butter alternative by the sensory and consumer tests as well. The results revealed that both types of oleogels are structurally and thermally suitable as alternative products. For the first time in the literature, these oleogels were described by a panel with thirteen sensory definition terms (hardness, spreadability, liquefaction, grassy, milky, rancid, fatty, sweet, salty, waxy, grittiness, cooling and mouth coating). Hedonic attributes (appearance, odor, flavor, spreadability) tested by the consumers proved that these oleogel products have potential as spread and/or butter alternatives. In the consumer survey study, one of the striking results indicated that more than 50% of the consumers would buy or try once, then decide to buy the oleogel products. In conclusion, these oleogels can be used as spreads or butter alternatives.

---

Introduction

Edible fat/oil products are served to consumers mostly either with flowing liquid or plastic consistencies. The physical state of the oil depends on the source fatty acid composition, applied processing technologies and degree of cooling. Meantime, this physical state of the oil is also the most crucial factor for determining its direct consumption intent and areas of food applications. Butter, margarine and spreads have been used for direct kitchen consumption as well as for many industrial applications like in baking, confectionary, candy manufacture, ready-to-eat foods, food frying, baby foods, etc.^1,2 Most of the time, some problems have been associated with the utilization of the traditional plastic fats, like shortages in availability of butter, higher saturated fat content and presence of trans fatty acids for butter and margarines and higher cooling energy expenses during transportation and storage.^1–5

Hence, alternative plastic fat sources are still in demand. Organogelation or for edible oils as it is called oleogelation, has emerged as a new method of oil plasticization technology. Briefly, edible liquid oils are gelled within a network created by organogelator molecules, and these gels truly behave like a plastic fat. There is an excellent book written about this topic covering most aspects of the oleogels.⁶ After many publications not to list here in the topic about how organogels form, what are the microstructures, what determines their stabilities, some studies headed towards their applications in food areas.^4,7–9 Furthermore, researches on new types of organogelators, health effects of the oleogels, and non-food applications will continue with curiosity.

Depending on the aims of this study, it is imperative to summarize briefly studies concerning oleogels for margarine/spread applications and butter alternatives. Applications of food grade organogelators like mono- and di-glycerols, fatty acids, fatty alcohols, waxes, wax esters, sorbitan esters, phytosterols and water soluble and modified polymers for the production of margarine/spread hard fat stocks have been reviewed.^4,7 Specifically, we studied the applications of beeswax and saturated monoglyceride as the organogelator in hazelnut oil to prepare hard fat stock as oleogels. We found that these preparates can successfully be utilized for spread/margarine type applications due to their suitable thermal, textural and stability properties.¹⁰ Similarly, olive oil organogels were made with beeswax and sunflower wax, compared directly with breakfast margarine, and were found highly similar in most properties.¹¹ One step further from our studies, Hwang et al.⁸ used organogels of sunflower wax, rice bran wax and candelilla wax as the solid fat stock in margarine formulations. They used 80% organogel as the solid fat phase together with other ingredients as liquid phase, created margarines, and compared them with commercial margarines. It was found that sunflower wax (SW) organogels were very suitable as hard fat stock for margarine production. In a continuation study, the same group¹² utilized 12 vegetable oils with 3, 5 and 7% SW organogels to produce margarines. It was found that 3% SW organogels containing margarines had greater firmness than commercial spreads, whereas margarines made with 7% SW were softer than commercial stick margarine. Although in these studies complete physical, thermal and textural characterizations were achieved, sensory definitions of the products and consumer perceptions were not investigated. In another study,¹³ edible applications of shellac oleogels were studied. Among them, the application of the oleogel as spread was characterized in terms of microstructure and rheology. The results showed that oleogel-based emulsions can be used as easily spreading fat materials. In this study as well, there was neither sensory definition nor consumer study for the spreads developed. We have reached only one study directly comparing the rheological properties of sunflower oil organogels made with different ratios of β-sitosterol/γ-oryzanol organogelator with those of butter.¹⁴ It was indicated that organogel prepared with 4.5% gelator concentration exhibited mechanical properties most similar to those of butter. In their study, there was no sensory and consumer studies. To our best knowledge, there is no other study reporting any sensory analysis or consumer tests for organogels.

The aims of this study were twofold. In the first part, we tried to fully characterize physico-chemically and sensorially, and to find out the consumer perceptions for virgin olive oil oleogels prepared with 5% beeswax (BW) and sunflower wax (SW). There was no color, antioxidant or other additives in these products, and sensory and consumer tests were accomplished in comparison with a commercial breakfast margarine (CBM). Since virgin olive oil (VOO) is an aromatic oil, it would not be compatible to add extrinsic aromas to it. The aim was to test the sensory properties and consumer acceptance of VOO when prepared as a spreadable fat in oleogel form. Since VOO is universally accepted as healthy oil, it would be meaningful to estimate its sensory properties and consumer acceptances in its spreadable form created naturally without any additives (except the organogelators) or change in fatty acids profile. Because of the fact that VOO has a very special and preferred virgin oil aroma, we decided to prepare it as spreadable fat and compare it with commercial breakfast margarine. Its distinct aroma makes impossible to consider it as butter alternative. In the second part of the study, our objective was to compare hazelnut oil (HO) oleogels made with 5% BW and SW and 0.5% (w/w on total weight) supplemented with butter flavor (diacetyl) with those of fresh commercial butter (CBT) through full physico-chemical analysis, sensory definition analysis and consumer tests. The aim was to determine how these oleogels would be expected as butter alternative. HO is a high oleic, neutral flavored oil, hence very suitable to be prepared as butter alternative after organogelation and aromatization with diacetyl. Since butter is a highly valued and in high market demand, it would be important to create an alternative product with added flavor to observe the similarities structurally and sensorially, and to observe how consumers perceive the new alternative. In our previously published studies, both hazelnut oil and virgin olive oil oleogels with different organogelators were prepared and characterized physico-chemically,^9–11 but in this study, both oleogels were prepared with the purpose of being breakfast margarine and butter alternatives, characterized fully by the physico-chemical analyses, and for the first time in literature, their sensory descriptive analyses and consumer tests were performed. Hence, this study is novel, and provides important contribution to the oleogel literature and those who are interested in oleogel applications.

Materials and methods

Materials

Hazelnut oil (HO), virgin olive oil (VOO), commercial breakfast margarine (CBM) and commercial butter (CBT) was purchased from local stores. Beeswax 8108 and Sunflower wax 6607L were purchased from KahlWax (Kahl GmbH & Co., Trittau, Germany). Beeswax was a whitish solid pellet with faint odor and 62–65 °C melting range, 17–24 mg KOH per g acid value, 87–104 mg KOH per g saponification value, 70–80 mg KOH per g ester value, max 65 °C saponification cloud test, no Japan wax, arsenic, lead and mercury and classified as GRAS. Sunflower wax was a yellowish solid pellet with soft odor, 74–80 °C melting range, 2–8 mg KOH per g acid value, 80–96 mg KOH per g saponification value, 0 mval per kg peroxide value. It was classified as free from dangerous chemicals (Annex I, Directive 67/548/EEC). Butter aroma (diacetyl) was provided by Aromsa Co. (Kocaeli, Turkey). All other chemicals were of analytical grade (Merck, Darmstadt, Germany and Sigma-Aldrich, St. Louis, US).

Preparation of the oleogels

In this study two different oils (HO and VOO) were used with two different organogelators (BW and SW) to prepare 4 different types of oleogels. The addition levels of the organogelators were constant and 5% (w/w). Only in HO oleogels, there was 0.5% (w/overall w) diacetyl addition as butter flavor. Briefly, calculated amounts of each oil and wax were put into separate beherglasses and heated in a water bath set to 90 °C. After complete melting of the waxes, and isothermal settings of the oils, known amounts of melted waxes were mixed with oil for 5 min. From our previous experiences, it was necessary to set temperature of the oils and waxes isothermal in order to mix the waxes well, and form uniform oleogels. For HO oleogels at last stage, calculated amount of diacetyl was added into the mixture and mixed thoroughly. Finally, the oleogels were filled up into plastic cups, tubes, and eppendorf tubes, and left overnight at ambient temperature for full gelation. The next day analyses were started to be done, meantime the samples were kept at ambient temperatures (20 ± 2 °C, laboratory temperature) during analyses.

Physico-chemical analysis

The instrumental color of the oleogel samples, CBM and CBT was assessed with Minolta CR-400 colorimeter (Konica Minolta Sensing, Osaka, Japan) calibrated against white tile. The peroxide values (PV) were measured by method Cd 8-53 using acetic acid–chloroform which determines all substances in terms of milliequivalents of peroxide per kg sample, that oxidizes potassium iodite under the test conditions.¹⁵ Iodine values (IV) were measured by method Cd 1-25 by Wijs method which determines the unsaturation of fats and oils and is expressed in terms of the number of centigrams of iodine absorbed per g sample.¹⁵ Free fatty acidity (FFA) values were measured by method Ca 5a-40 by titrimetric technique with phenolphthalein indicator, and expressed as % oleic acid.¹⁵ Combustion calorie values of the samples were analyzed with bomb calorimeter (Leco AC-350 model, St. Joseph, USA) according to its manual.

Thermal analysis

The thermal properties of the samples were analyzed with Perkin-Elmer 4000 Series Differential Scanning Calorimetry (Groningen, The Netherlands). 5–7 mg oleogel samples were weighed into aluminum pans and sealed hermetically. The temperature programming was heating from room temperature to 140 °C by 10 °C min⁻¹ heating rate. Then, cooling down to −20 °C by 10 °C min⁻¹ rate and keeping 3 min at that temperature for full crystal formation. Finally, the samples were heated to 100 °C by 5 °C min⁻¹ heating rate. Calibration of the instrument was achieved with indium and zinc. From the thermograms, the thermal parameters of the samples were calculated using the Pyris 1 Manager software on the instrument.^10,11 The solid fat content (SFC) of the samples were measured with Minispec Bruker NMR Analyzer mq20 (BrukerOptics, Inc.) at 20 °C. The oleogel samples were first completely melted in water bath at 90 °C. Then, 3.5 ml of each samples were taken into NMR tubes and conditioned in water bath at 0 °C for 1 h, and then at 20 °C for another 1 h, before reading. The calibration of the instrument was with the standard solutions including 0, 31 and 73.5% solid fat.

Texture analysis

The hardness and stickiness values of the samples were determined with a Texture Analyzer TA-XT2i (Stable Microsystems, Surrey, UK) at ambient temperature according to our previous technique described.^10,11 The penetration test was conducted by pushing the 45° conic acrylic probe with 3.0 mm s⁻¹ penetration speed into 23 mm sample depth which is placed in a custom-built block, and then pulling it out from the sample at 10 mm s⁻¹ speed. Texture Exponent software (v.6.1.1.0, Stable Microsystems) was used to calculate the parameters.^10,11 Crystal sizes of the samples were calculated from the X-ray diffraction data measured with Rigaku D-Max Rint 2200 model X-Ray Diffractometer and MDI Jade 7 software (Rigaku Int. Corp, Tokyo, Japan and Materials Data Inc., Livermore, USA). Angular scans (2θ = 2.0–50° by 2° min⁻¹) were performed using a Cu source X-ray tube (λ = 1.54056 Å, 40 kV, 40 mA).^10,11

Sensory descriptive analysis

Quantitative Descriptive Analysis (QDA) was applied to the four organogel, CBM and CBT samples.¹⁶ There were 8 female and 2 male panelists aged between 25 and 45. A consent form indicating that the samples are edible and safe was signed by them with their volunteer intentions to contribute to the test. The panel was trained at least 10 hours for the determination, definition and scoring of the sensory terms. The panel developed 13 different sensory description terms to assay the samples. The sensory terms, their definitions and references used to calibrate the panelists are given in Table 1. A 10 cm line scale anchored from 0 at the left end to 10 at the right end was used to quantify the sensory attributes. In each evaluation session, only 3 randomly selected samples coded with 3-digit numbers were served to the panel. Duplicate samples were analyzed in different sessions held in different days for each of the replicates of oleogel production samples and control samples. Sensory analysis was carried out at room temperature under daylight with panelists provided with water, bread slices, an apple slice and an expectoration cup.

Table 1 The panel defined sensory descriptive terms and their references used

Descriptor	Definition	References
Hardness	Force required to push a knife into sample	Min: yoghurt, max: tallow
Spreadability	Ease at which sample spreaded onto a surface as thin layer	Min: chewing gum, max: cream cheese
Liquefaction	Melting amount when sample spreaded on bread	Min: tallow, max: olive oil
Grassy	Aroma associated with cut grass	Min: absent, max: fresh cut grass
Milky	Flavour of dairy fresh milk	Min: absent, max: hot milk
Rancid	Aroma of oxidized oils	Min: fresh olive oil, max: heavily heated and stored oil
Fatty	All fatty taste and aroma notes	Min: no fat, max: Sunflower oil
Sweet	Taste associated with sugar	Min: no sugar, max: 5% sucrose solution
Salty	Taste stimulated by table salt	Min: no salt, max: 0.5% salt solution
Waxy	Aroma associated with waxes	Min: absent, max: paraffin, sunflower wax
Grittiness	Rough, particulate feeling in mouth	Min: none, max: semolina flour
Cooling	Cold feeling inside mouth	Min: none, max: menthol candy
Mouth coating	Layers of fat deposit on palate	Min: liquid oil, max: cream cheese, tallow

---

Consumer test

The attributes of appearance, odor, flavor and spreadability of the oleogel samples and CBM and CBT samples were determined by 120 different volunteer consumers. The samples were served as two separate triple sets (2VOO oleogels + CBM, and 2HO oleogels + CBT) in different days together with knife and slices of bread. 5-Point hedonic scale (1 = dislike extremely to 5 = like extremely) was used. After analyzing the data from the consumer tests, the two oleogels were identified with the highest scores (hazelnut oil–beeswax oleogel and virgin olive oil–beeswax oleogel) were selected to determine the buying intentions of the consumers. These two samples were coded and the consumers were asked with 4 selections options for their buying decision (definitely buy, try once-then decide, definitely not buy, no opinion). Score sheets were collected and the data were analyzed.

Statistical analysis

The parametric data group comparison was applied by one-way ANOVA, and Tukey's test for mean separation using statistical software. The non-parametric sensory data was analyzed by Kruskal–Wallis test (Minitab v.16.1).¹⁷ Differences were considered statistically significant when P values were ≤0.001.

Results and discussion

Physico-chemical, thermal and textural properties of the oleogels

A picture of the developed oleogels, and CBM and CBT samples used in this study is shown in Fig. 1. This picture provides a good visual feeling of what the oleogels look like, how they resemble the commercial breakfast margarine and butter, and how their texture is.

Fig. 1 Hazelnut oil and virgin olive oil oleogels prepared with BW and SW, and CBM and CBT samples.

Table 2 represents some important physical and chemical values of the samples. In this study, the samples are grouped in two and the samples of each group were compared among themselves, as stated in the objectives of the study under the Introduction section. From here onward, HO oleogels are compared with the CBT, and VOO oleogels are compared with the CBM. There were some color differences between the samples. The level of luminosity (L* value) was higher in CBT and CBM samples compared to the oleogels.

Table 2 Some physico-chemical features of the hazelnut oil and virgin olive oil oleogels (mean ± SD)^a

Sample	Instrumental colour			PV (meq. O2 per kg)	IV (g/100 g)	FFA (%)	Energy (cal g^−1)
	L	a*	b*
a Lowercase letters compare the samples within each column and type of oleogel (p ≤ 0.001). HBW: hazelnut oil–beeswax oleogel; HSW: hazelnut oil–sunflower wax oleogel; CBT: commercial butter; OBW: olive oil–beeswax oleogel; OSW: olive oil–sunflower wax oleogel; CBM: commercial breakfast margarine.
HBW	42.40 ± 0.04c	−2.18 ± 0.01c	−0.69 ± 0.04c	0.76 ± 0.10a	88.43 ± 3.92a	0.63 ± 0.01b	9677.80 ± 22.50a
HSW	59.52 ± 0.47b	−2.34 ± 0.01b	4.11 ± 0.07b	0.70 ± 0.15a	92.02 ± 2.78a	0.57 ± 0.01c	9876.60 ± 2.50a
CBT	91.39 ± 0.28a	−3.90 ± 0.02a	32.53 ± 0.06a	0.01 ± 0.00b	29.70 ± 0.59b	0.92 ± 0.01a	7968.90 ± 35.40b
OBW	40.10 ± 0.90c	−6.01 ± 0.21b	19.87 ± 0.84b	2.01 ± 0.01a	83.91 ± 2.89a	1.47 ± 0.01a	9914.10 ± 7.20b
OSW	60.30 ± 1.62b	−7.95 ± 0.22a	36.32 ± 1.53a	1.36 ± 0.17b	82.71 ± 4.81a	1.19 ± 0.02b	9992.60 ± 1.60a
CBM	88.68 ± 2.00a	−2.44 ± 0.03c	14.61 ± 0.76c	0.73 ± 0.06c	57.85 ± 7.91b	0.73 ± 0.03c	5299.40 ± 0.60c

---

Other color components (a* and b* values) were also different among the samples. Clearly, CBT is much more yellow than HO oleogels. It would be important to state here that there was no color addition into the oleogels, but it is always possible to add permitted food colorant to modify the color of the oleogels. VOO oleogels were greener than CBM, respectively, due to the natural greenish color of the virgin olive oil. The peroxide values (PV) of the HO oleogels and CBT were quite low. Although PV of VOO–beeswax oleogel was a little bit higher than the others, no PV exceeded the permitted limit values for vegetable fat and oil products in Turkey. According to the Codex, PV of up to 15 meq. O₂ per kg oil for virgin and cold pressed oils, and 10 meq. O₂ per kg oil for refined oils is permitted.¹⁸ In one study,¹⁹ four different margarines were stored at room and fridge temperatures and the PV was monitored. It was shown that storage at refrigerator temperature controlled PV considerably, and the PVs of the samples stored for 12 weeks at room temperature increased steadily up to 10–25 meq. per kg. Hence, PVs of our samples are quite acceptable. The measured iodine values (IV) of HO oleogels were higher than that of the CBT sample. Since butter contains considerable amounts of solid fat content, this finding is expected. The IV of the stock liquid oils, HO and VOO, were measured as 87.3 g/100 g oil and 81.5 g/100 g oil, respectively. Similarly, IV of CBM was higher than that of VOO oleogels because commercial margarines also contain around 10–15% or more solid fat to get the desired consistency. In oleogels, the IV can only indicate the total unsaturation level of the product and hence its nutritional quality. Contrarily, hardness and other textural properties of oleogels are controlled by the kind and amounts of the added organogelators. There was a slight increase in the IVs of the oleogels compared to their base liquid oils, most possibly due to the added waxes. Overall, it can be observed that organogelation does not cause any significant change in the saturation level of oil; hence, the nutritionally beneficial unsaturated fatty acids were unchanged, as the previously stated major advantage of oleogels.⁶ Free fatty acidity (FFA) levels of HO oleogels and CBT were quite low. Since HO was refined oil, it is expected the FFA to be lower. Although we used virgin olive oil with measured FFA level below 1.0, the FFA values of VOO oleogels were a little higher than 1.0, indicating that same free acidity may come from the waxes. Neither oleogels, nor the commercial products had any unaccepted levels of FFA, according to the codex. In one study, FFA values in the stored margarines were mostly below 0.3%, and did not increase during storage.¹⁹ The combustion calorie values were also measured. CBT had lower energy value than that of the HO oleogels, just like CBM had much lower energy value than those of the VOO oleogels. This might be due to the water content of the commercial products; hence, margarines and fresh butter are emulsion products and contain some water within, whereas oleogels are pure fat products without any water. Hence, during consumption or usage for food products this situation should be considered.

Some important thermal properties of the samples are presented in Table 3. DSC determined onset (start of melting), peak melting temperatures and melting enthalpies provide data to understand and compare the thermal behaviour of plastic fat samples. HO–beeswax oleogel seemed quite similar to CBT, although HO–sunflower wax oleogel had significantly higher melting temperatures. Melting enthalpy of CBT was significantly higher than those of the HO oleogels, indicating that solid fat content (higher temperature melting fractions) of butter must be in much higher amounts. Truly, oleogels had around 3.5% solid fat content (SFC) at 20 °C as measured by NMR, but it was 17.4% for CBT. Although CBT and HBW oleogel had almost the same melting temperatures, their saturated (solid) fat content was quite different. This condition specifies one of the main advantages of the oleogels. They are plastic consistency products with the same fatty acid composition of the liquid oil from which are made. Very similar situation was present for the VOO oleogels and CBM. Generally, SW had higher melting point oleogels than BW at the same addition levels. This situation was also observed previously in our other studies.^9–11 Hence, depending on the final purpose of the oleogel, either organogelators might be selected. It seems possible that at lower SW concentrations, the same thermal behaviours of BW oleogels can be obtained. Nevertheless, not only thermal properties, but also textural properties and most importantly sensory properties should be considered altogether during the selection of the organogelator type and addition level for suitable oleogel production.

Table 3 Thermal properties, solid fat content and crystal size of the samples (mean ± SD)^a

Sample	Onsetm (°C)	Peak (Tm) (°C)	ΔH (J g^−1)	SFC (%, 20 °C)
a Lowercase letters compare the samples within each column and type of oleogel (p ≤ 0.001). HBW: hazelnut oil–beeswax oleogel; HSW: hazelnut oil–sunflower wax oleogel; CBT: commercial butter; OBW: olive oil–beeswax oleogel; OSW: olive oil–sunflower wax oleogel; CBM: commercial breakfast margarine.
HBW	36.32 ± 0.54b	49.39 ± 0.05b	7.82 ± 0.97b	3.50 ± 0.07b
HSW	51.64 ± 0.41a	61.02 ± 0.02a	9.76 ± 0.19b	3.55 ± 0.05b
CBT	36.78 ± 0.01b	49.61 ± 0.01b	167.17 ± 0.98a	17.40 ± 0.08a
OBW	36.18 ± 0.24b	47.76 ± 0.01b	5.57 ± 0.27b	3.64 ± 0.12b
OSW	47.30 ± 0.38a	62.26 ± 0.01a	10.99 ± 0.69b	3.52 ± 0.16b
CBM	36.39 ± 1.46b	43.74 ± 4.77b	85.50 ± 12.30a	7.70 ± 0.01a

---

The hardness, stickiness values and crystal sizes of the samples were measured and summarized in Table 4. CBT is a much harder and stickier fat than those of the HO–beeswax and sunflower wax oleogels (HBW and HSW) measured at ambient temperature. Although there is no direct method to measure spreadability of plastic fats, hardness (firmness) and stickiness might be considered together for the estimation of spreadability. By definition, hardness is the force required to create a deformation to a sample, while stickiness is the force required to pull back the probe from the sample. Consequently, spreadability can be estimated from both parameters, and it is defined as the easiness of a sample to be applied as a thin layer onto a surface by a knife. Moderate levels of hardness and stickiness indicate good spreadability, as stated by Moskowitz,²⁰ although the best estimate would be by sensory analysis. In one study, correlations between texture parameters for table fats were investigated. It was found that high correlations exist between hardness and spreadability, and cohesiveness and spreadability.²¹ It was also suggested that spreadability is the best estimate to evaluate butter consistency, and firmness measured with cone penetrometer was recommended along with other methods to estimate butter spreadability.³ It was also indicated that good butter should have fine and close texture, have a firm and waxy body and be sufficiently plastic to be spreadable at lower temperatures.³ It is obvious that full estimation of spreadability by instrumental means is not possible, and sensory evaluation should be considered in this respect. Regardless of this, it is clear that both oleogels had sufficient levels of hardness and stickiness to be spreadable like CBT. Comparison of VOO oleogels with CBM revealed different results (Table 4). VOO–SW oleogel (OSW) was harder than CBM and VOO–BW oleogel (OBW) samples. Similarly, stickiness of the CBM sample was lower than those of the OBW and OSW samples. Although directly not compared, it is easy to observe that both hardness and stickiness values of the CBM were lower than that of the CBT at the same temperature. Hence, it is possible to indicate that oleogels of both oils had hardness and stickiness values between the values measured for butter (CBT) and breakfast margarine (CBM). Hence, spreadability of the oleogels might be between them and acceptable. It was indicated that spreadability is the most highly regarded attribute for margarines second to flavour, and margarines with 10–20% solid fat index at serving temperatures were found as optimal in consumer perceived spreadability.¹ Again, for plastic fat type products it would be always better to estimate spreadability by sensory assessments in addition to texture measurements. Crystal sizes of the samples were calculated from the X-ray diffraction (XRD) data and also presented in Table 3. The crystal size of CBT (606–650 Å) was significantly higher than that of the hazelnut oil oleogels produced as butter alternatives. Crystal size can give clue about the polymorphic type of the crystals. In fact, from the wide- and small-angle region peaks of the XRD data, it is quite possible to estimate polymorph types. In our previous studies^9–11 as well as in this study, it was observed from the XRD data (not shown again in this study) that both SW and BW oleogels were in β′ type polymorph. This type polymorph is characterized by smooth, homogeneous and fine texture. The CBM and CBT samples used in this study had also similar XRD patterns, indicating that they are mostly composed of β′ type crystals.^1,22

Table 4 Textural properties of the hazelnut oil and virgin olive oil oleogels (mean ± SD)^a

Sample	Hardness (g force)	Stickiness (g force)	Crystal Size (Å)
a Lowercase letters compare the samples within each column and type of oleogel (p ≤ 0.001). HBW: hazelnut oil–beeswax oleogel; HSW: hazelnut oil–sunflower wax oleogel; CBT: commercial butter; OBW: olive oil–beeswax oleogel; OSW: olive oil–sunflower wax oleogel; CBM: commercial breakfast margarine.
HBW	172.81 ± 2.35b	−133.82 ± 1.69b	43–64b
HSW	238.05 ± 37.41b	−87.38 ± 15.25b	47–86b
CBT	809.10 ± 59.02a	−270.61 ± 2.96a	606–650a
OBW	160.46 ± 30.46b	−114.46 ± 11.67a	26–51c
OSW	306.01 ± 14.33a	−85.86 ± 5.32ab	42–60b
CBM	189.45 ± 16.38b	−70.51 ± 5.02b	143–171a

---

Descriptive sensory analysis of the samples

Sensory Quantitative Descriptive Analysis (QDA) was completed for all samples, and the results are summarized in Tables 5 and 6. There were thirteen descriptive terms (Table 1) developed by the panel to describe the oleogel samples, CBT and CBM. The order of perception was considered during the sensory test through the order of definition terms on the evaluation ballot. First, the sensory terms perceived when the sample was cut and spread onto a slice of bread by a knife (hardness, spreadability and liquefaction) were measured. Then, aroma (grassy, milky, rancid) and taste attributes (fatty, sweet, salty, waxy) were measured when the sample was taken directly into mouth. Finally, mouth feeling properties (grittiness, cooling, mouth coating) were assessed as the sample was left for some time in mouth and swallowed. Sensory ‘hardness’ defined as the force required for pushing a knife inside the sample and acquainted with yoghurt as minimum and tallow as maximum intensity references (Table 1). The hardness of CBT was significantly higher than that of HO–oleogels. This sensory estimation was confirmed with the instrumental findings (Table 4). Contrarily, sensory ‘spreadability’ perceived by the panel for CBT was significantly lower than those of the oleogels. When instrumental texture parameters (Table 4) and spreadability predicted from those parameters are considered altogether, it would be possible to observe the difference. But the true determination of spreadability can only be through the sensory assessment. The panel found both oleogels very spreadable and very highly spreadable than CBT. ‘Liquefaction’ was defined as the amount of melted fat when it was spread onto a surface. Liquefaction of HO–oleogels was higher than that of the CBT. In fact, it is a well acknowledged phenomenon that either thermal or kinetic energy input cause oleogels to melt, but when the energy is cut, they immediately re-gel. This kinetic reversibility is a well-known phenomenon for oleogels,^4,6 and also observed during the sensory tests in this study. The aroma was associated with cut grass, named as ‘grassy’ and it was found that CBT had higher scores. Very interestingly, sensory ‘milky’ perception was lower in CBT, and in the oleogels. It would be expected to be lower in the oleogels, but it was also perceived in small scores for CBT. Similarly, ‘rancid’ scores were lower for all samples and not different among them. It would be better for such a negative image odor descriptor to be lower. The ‘fatty’ term as defined all fatty taste and aroma sensations together was higher in the CBT than the oleogels. This finding is also interesting, since oleogels are high fat (95% by weight) products. On the other hand, CBT has typical fatty aromas and color, most probably causing the panel to sense it as more fatty. ‘Sweet’ and ‘salty’ scores were not different among the samples, and CBT had a little higher value for salty, most possibly due to the added salt in the finished product. The ‘waxy’ sensations were significantly higher in the oleogel samples compared to CBT, expectedly. The organogelators, BW and SW, are true waxes and provide the typical waxy aroma associated with them. On a whole, the waxy scores were not much and below 4.0, on the 0–10 scale. Lastly, three mouth feeling properties were evaluated. ‘Grittiness’ defined as rough, particulate feeling of the sample inside mouth, and was not significantly different among the samples. ‘Cooling’ is caused by the melting of fat in mouth, and measured around 1.20–1.60 among the samples. There was no difference for cooling values among the samples. ‘Mouth coating’ was defined as the fat deposit on palate, and was also not significantly different among the HO–oleogels and CBT samples. Unfortunately, there is no study in literature about the sensory properties of oleogels produced as butter alternative, but sensory studies with different butters exist. In one of them,²³ enriched butter samples were evaluated with 11 sensory descriptive terms during storage. Although the goal of the researchers was to observe the oxidative changes during storage, they used some similar sensory terms with our study (cooked milk, rancid, spreadability, sweet).²³ In another study,²⁴ sensory attributes of whey, cultured, and regular sweet cream unsalted butters were evaluated by sensory tests, and differences between the 3 types of butters were obtained on yellow, shiny, acidic odor, melt rate, porous, hard, spreadable, cheese odor, mouth-coating, nutty, cardboard odors, acidic, nutty, diacetyl, and grassy flavors. In another similar study,²⁵ pure butterfat, butterfat–canola oil blend and butterfat–structured lipid blend were analyzed. The texture of the samples was defined by cold-spreadability, oiliness, hardness, rate of melting, adhesiveness and cohesiveness terms. Likewise, butter, margarine, caprylic acid, rancid and greasy terms was used as the flavor definition terms. These studies and our study concur on most of the sensory terms defined and used by the panels for these types of products.

Table 5 QDA results of the hazelnut oil oleogels prepared with BW and SW (mean ± SE; Me)^a

Descriptor	HBW	HSW	CBT
a Lowercase letters compare the samples within each row (p ≤ 0.001). HBW: hazelnut oil–beeswax oleogel; HSW: hazelnut oil–sunflower wax oleogel; CBT: commercial butter.
Hardness	2.44 ± 0.29	2.50 ± 0.21	7.81 ± 0.50
	2.00b	2.75b	8.50a
Spreadability	7.88 ± 0.35	7.75 ± 0.57	2.94 ± 0.76
	8.00a	8.00a	2.75b
Liquefaction	5.50 ± 0.88	5.38 ± 0.87	1.69 ± 0.95
	5.75a	5.00a	0.75b
Grassy	4.44 ± 0.79	4.81 ± 1.00	8.81 ± 0.31
	4.50b	5.75b	9.00a
Milky	2.44 ± 0.80	1.00 ± 0.28	0.50 ± 0.25
	2.00a	1.00ab	0.25b
Rancid	1.19 ± 0.30	1.25 ± 0.44	0.75 ± 0.35
	1.25a	0.75a	0.50a
Fatty	6.75 ± 1.04	8.44 ± 0.27	9.00 ± 0.19
	7.50a	8.75a	9.00a
Sweet	2.44 ± 0.56	2.44 ± 0.68	2.44 ± 0.48
	1.75a	2.25a	3.00a
Salty	0.88 ± 0.26	0.69 ± 0.19	1.75 ± 0.38
	0.50b	0.75ab	1.75a
Waxy	4.06 ± 0.98	3.56 ± 0.79	1.19 ± 0.43
	3.50a	3.25ab	0.75b
Grittiness	1.75 ± 0.37	1.44 ± 0.35	0.88 ± 0.28
	2.00a	1.00a	0.50a
Cooling	1.31 ± 0.56	1.19 ± 0.44	1.56 ± 0.38
	1.00a	1.00a	1.50a
Mouth coating	3.06 ± 0.55	2.69 ± 0.57	2.94 ± 0.68
	3.25a	2.50a	2.50a

---

Table 6 QDA results of the virgin olive oil oleogels prepared with BW and SW (mean ± SE; Me)^a

Descriptor	OBW	OSW	CBM
a Lowercase letters compare the samples within each row (p ≤ 0.001). OBW: olive oil–beeswax oleogel; OSW: olive oil–sunflower wax oleogel; CBM: commercial breakfast margarine.
Hardness	3.75 ± 0.46	6.50 ± 0.69	5.00 ± 0.66
	4.00b	7.25a	5.50ab
Spreadability	8.00 ± 0.27	5.88 ± 0.69	7.75 ± 0.37
	8.00a	6.75b	8.00a
Liquefaction	5.44 ± 0.73	3.00 ± 0.61	2.56 ± 0.75
	5.75a	3.50ab	1.50b
Grassy	3.63 ± 0.66	2.75 ± 0.82	0.50 ± 0.16
	4.00a	2.00a	0.50b
Milky	2.13 ± 0.76	2.00 ± 0.71	4.69 ± 0.86
	1.25b	1.25b	4.75a
Rancid	1.56 ± 0.53	1.88 ± 0.80	0.88 ± 0.36
	1.25a	0.75a	0.50a
Fatty	7.38 ± 0.65	6.94 ± 0.45	7.13 ± 0.63
	7.75a	7.00a	7.50a
Sweet	2.88 ± 0.77	2.31 ± 0.74	3.00 ± 0.78
	2.25a	1.75a	2.00a
Salty	0.63 ± 0.16	1.06 ± 0.22	1.38 ± 0.52
	0.75a	1.00a	0.75a
Waxy	3.88 ± 0.94	5.19 ± 0.67	2.00 ± 0.80
	2.75ab	5.00a	1.00b
Grittiness	1.25 ± 0.38	2.69 ± 0.43	0.56 ± 0.15
	1.25ab	2.50a	0.50b
Cooling	1.25 ± 0.60	1.00 ± 0.53	1.38 ± 0.34
	0.50a	0.25a	1.25a
Mouth coating	3.00 ± 0.71	4.19 ± 0.91	2.44 ± 0.60
	3.00a	4.00a	2.00a

---

The same panel evaluated the CBM and VOO-oleogels with the same descriptive terms (Table 6). Among these samples, some differences exist not only between the oleogels and the CBM, but also between the two oleogels. Maximum ‘hardness’ was in the OSW sample, followed by CBM and OBW. ‘Spreadabilities’ of OBW and CBM were very similar and higher than that of the OSW. In this group, ‘grassy’ was significantly lower in the CBM than those of the oleogels. Since the base oil was virgin olive oil (VOO) in the oleogels, there might be some grassy aromatics coming from the VOO. Contrarily, ‘milky’ score of CBM was higher, possibly due to milk based ingredients used to prepare breakfast margarine, as stated in the materials section. ‘Rancid’ was lower and not different among samples, whereas ‘fatty’ scores were around 7.0 and not different among the samples. Although there were some small differences among the samples for the ‘sweet’ and ‘salty’ tastes, they all have lower values. The ‘waxy’ value of OSW sample was higher than the others. This situation was possibly caused by the natural waxy aromas of SW. Obviously, SW yielded a little ‘grittier’ oleogel. There was no significant difference among the samples for ‘cooling’ and ‘mouth coating’ attributes. When Tables 5 and 6 are considered together, it can be observed that there were no great differences among all samples for the sensory definition terms. In literature, there were some sensory studies with different types of margarine and spread type products. Although margarines prepared from organogels^8,12,13 were evaluated for micro structural and rheological properties, their sensory analysis has not been published. In one study,²⁶ sensory properties of experimental trans-free margarine spread and two commercial margarine spreads were compared with sensory attributes of appearance, spreadability and texture-mouth with ranking (difference) test. Clearly, these sensory attributes were selected as the discrimination terms because of their importance in these type products, and similar terms and more were used in our study. Butter, margarine and two designer spreads were evaluated sensorially in another study.²⁷ The researchers used yellow, white, water, crumbly, shiny, dull and flaky as appearance attributes; storage, fruity, butter, greasy, sweet, rancid, margarine, and smoked as odour attributes; salty, greasy, margarine, butter, plastic aftertaste, sweet, rancid and minty as flavour attributes; slippery, rate of melting, smooth, soft and firm as texture attributes.²⁷ Depending upon the purpose of sensory evaluation, and variety of samples, different panels have developed numerous sensory terms to define the samples. It would be not practical to list other similar studies made with butter, margarine or spreads again. Sensory studies with oleogels and/or oleogel containing products are on demand.

Consumer tests of the samples

In order to acquire the consumer perceptions of the oleogel samples in comparison with CBT and CBM, hedonic tests were accomplished for the appearance, odor, flavor and spreadability attributes (Fig. 2). The appearance, odor and flavor scores of the CBT were higher than those of the two oleogels, HBW and HSW, whereas spreadability was higher in HBW and HSW samples (Fig. 2A). Most sensory attributes were close to 3.0 values, indicating a little acceptance over the neutral point of 2.5 on the hedonic scale, by the consumers. Spreadability scores of the oleogels were over 4.0, indicating high acceptance. Similar situation was also evident for the VOO-oleogel samples (Fig. 2B). Only spreadabilities of OBW and CBM samples were over 4.0 score. For the other attributes (appearance, odor and flavor) score of the OBW and OSW samples were lower than that of the CBM.

Fig. 2 Hedonic scores of the samples (A) hazelnut oil oleogels and commercial butter; (B) virgin olive oil oleogels and commercial breakfast margarine (HBW: hazelnut oil–beeswax oleogel; HSW: hazelnut oil–sunflower wax oleogel; CBT: commercial butter; OBW: olive oil–beeswax oleogel; OSW: olive oil–sunflower wax oleogel; CBM: commercial breakfast margarine; n = 120).

Overall, these hedonic results indicate that consumers have a little positive perception for the developed oleogels for the attributes of appearance, odor and flavor, except spreadability. Hence, more studies are needed to improve appearance (mostly color), aroma and flavor properties of the wax oleogels to make them highly acceptable as spread/margarine and butter alternatives.

In the last part of the consumer test, we have selected HBW and OBW samples due to their relatively higher hedonic scores, and asked to consumers about their buying decision with 4 answer selections (Fig. 3). Around 57% and 43% of the consumers indicated that they ‘definitely buy’ HBW and OBW, respectively. The ratio of ‘try once, and then decide’ consumers were 24% for HBW and 29% for OBW. The percent of the consumers who ‘definitely not buy’ these samples were 12% and 25% for HBW and OBW, respectively. There were also small percentages (4% and 7%) of consumers who had ‘no opinion’ about buying these samples. This buying decision data indicate that, consumers usually well accept and would buy especially HBW but also OBW samples. In this study, the consumers were not provided with any knowledge about the samples. Hence, some composition and health effect knowledge can also change their buying decision. In one study,²⁷ consumer acceptance of butter, margarine and two designer spreads were studied. Mean liking scores for consumption at breakfast, lunch and dinner's meals indicated that butter was the most liked spread. Usually, butter consumers accounted for the liking among samples, whereas margarine consumers liked all products equally. The two designer spread were found equally well accepted by the consumers. Further studies with different and detailed consumer preference and buying decision tests are determined as the immediate research need in this area to optimize oleogel products for successful market applications.

Fig. 3 Consumer buying decision scores of the hazelnut oil and virgin olive oil oleogels prepared with BW (HBW: hazelnut oil–beeswax oleogel; OBW: olive oil–beeswax oleogel).

Conclusions

In this study, virgin olive oil oleogels of BW and SW were produced as spreadable products and compared with commercial breakfast margarine. Similarly, wax oleogels of hazelnut oil with added diacetyl flavor were produced as butter alternative and compared with commercial butter. The most distinct part of this study is that sensory description of the prepared oleogels and commercial products were assessed by panel and compared. Furthermore, consumer perception of sensory attributes and consumer buying decisions were tested. Regardless of the accomplished physical and chemical analysis, these sensory studies are the novel parts of this study. It was observed that both types of the oleogels are structurally and thermally suitable as alternative spread products. Sensory descriptive analysis yielded full description as well as the differences among the samples. In most properties, the oleogel samples usually resemble commercial butter and margarine samples. Hedonic attributes tested by the consumers mostly took scores a little higher than neutrality point; hence, these oleogel products have potentials as spread and/or butter alternatives. Interestingly, buying decisions of the consumers revealed that more than 50% of the consumers would buy or try once, then decide to buy the oleogel products. Hence, some potential exist for these oleogels as spreads or butter alternatives. But, more studies to improve their color, aroma and taste are required. We suggest suitable color and aroma additives to improve sensory acceptability. Most importantly, organogelators with neutral taste and aroma or well gelling organogelator at much small addition levels might help to improve the sensory quality and consumer acceptability of the oleogels.

Acknowledgements

This research is funded by the Scientific and Technical Council (TÜBİTAK) of Turkey as the COST 112O038 project within COST FA 1001 Action. The authors thank gracefully for the support.

Notes and references

1. M. M. Chrysam, Margarines and Spreads, in Bailey's Industrial Oil & Fat Products, ed. Y. H. Hui, Wiley-Intersience Pub, New York, 1996, pp. 65–114.
2. R. D. O'Brien, Fats and Oils: Formulating and Processing for Applications, CRC Press, New York, USA, 2004.
3. D. Hettinga, Butter, in Bailey's Industrial Oil & Fat Products, ed. Y. H. Hui, Wiley-Intersience Pub, New York, USA, 1996, pp. 1–64.
4. A. Marangoni, J. Am. Oil Chem. Soc., 2012, 89, 749–780.
5. T. A. Stortz, A. K. Zetzl, S. Barbut, A. Cattaruzza and A. G. Marangoni, Lipid Technol., 2012, 24, 151–154.
6. A. Marangoni and N. Garti, Inform, 2011, 22(5), 317–320.
7. N. E. Hughes, A. G. Marangoni, A. J. Wright, M. A. Rogers and J. W. E. Rush, Trends Food Sci. Technol., 2009, 20, 470–480.
8. H.-S. Hwang, M. Singh, E. Bakota, J. Winkler-Moser, S. Kim and S. Liu, J. Am. Oil Chem. Soc., 2013, 90, 1705–1712.
9. M. Öğütcü, N. Arifoğlu and E. Yılmaz, Int. J. Food Sci. Technol., 2015, 50, 404–412.
10. E. Yılmaz and M. Öğütcü, J. Food Sci., 2014, 79, E1732–E1738.
11. E. Yılmaz and M. Öğütcü, J. Am. Oil Chem. Soc., 2014, 91, 1007–1017.
12. H.-S. Hwang, M. Singh, J. K. Winkler-Moser, E. L. Bakota and S. X. Liu, J. Food Sci., 2014, 79, C1926–C1932.
13. A. R. Patel, P. S. Rajarethinem, A. Gredowska, O. Turhan, A. Lesaffer, W. H. De Vos, D. Van de Walle and K. Dewettinck, Food Funct., 2014, 5, 645–652.
14. C. Alamprese, I. Ferrari, M. Rossi, M. Lucisano and M. Mariotti, Inside Food Symposium, 2013, pp. 1–4.
15. AOCS, Official Methods and Recommended Practices of the American Oil Chemists Society, AOCS Press, Champaign, 5th edn, 1997.
16. M. Meilgaard, G. V. Civille and B. T. Carr, Sensory EvaluationTechniques, CRC Press, Boca Raton, USA, 1991.
17. Minitab, Minitab 16.1.1 Statistical Software, Minitab Inc, State College, Pennsylvania, USA, 2010.
18. Turkish Food Codex, Codex for vegetable oils (number: 2012/29), Turkey, Ankara, 2012.
19. H. Zhang, C. Jacobsen, L. S. Pedersen, M. W. Christensen and J. Adler-Nissen, Eur. J. Lipid Sci. Technol., 2006, 108, 227–238.
20. H. R. Moskowitz, Food Texture-instrumental and sensory measurements, Marcel Dekker, New York, USA, 1987.
21. P. Glibowski, P. Zarzycki and M. Krzepkowska, Int. J. Food Prop., 2008, 11, 678–686.
22. A. Szydłowska-Czerniak, G. Karlovits, M. Lach and E. Szłyk, Food Chem., 2005, 92, 133–141.
23. S. Mallia, P. Piccinali, B. Rehberger, R. Badertscher, F. Escher and H. Schlichtherle-Cerny, Int. Dairy J., 2008, 18, 983–993.
24. S. Jinjarak, A. Olabi, R. Jiménez-Flores and J. H. Walker, J. Dairy Sci., 2006, 89, 2428–2440.
25. B. H. Kim, R. L. Shewfelt, H. Lee and C. C. Akoh, J. Food Sci., 2005, 70, s406–s412.
26. S. Lumor, A. Pina-Rodriguez, R. Shewfelt and C. Akoh, J. Am. Oil Chem. Soc., 2010, 87, 69–74.
27. M. Michicich, Z. Vickers, M. C. Martini and J. B. Labat, Food Quality and Preference, 1999, 10, 147–154.

---