The shelf life of tempeh stored at 20, 30, and 40 °C can be extended by treatment with CO₂ at sub-supercritical pressure.

Tempeh, in the form of a cylinder with a diameter of 35 mm and a length of 100 mm, fermented for 36 hours at 30 °C, was obtained from the Center of Home Industry Tempeh Making Palembang, Indonesia, placed in a cooler box and carried to the laboratory for direct processing (Kustyawati et al., 2018).

The high-pressure CO₂ installation used for experimental treatments, consisting of a CO₂ gas cylinder, a cylindrical pressure chamber, pressure gauges, and a water bath at a constant temperature, is shown in Figure 1 (Saputra, 2006). Fresh tempeh was placed in a pressure chamber and then closed tightly. When the designated temperature in the water bath reached a constant 25 °C, and all pipe connections were secured, commercially available CO₂ was injected through the gas inlet valve from the gas cylinder into the pressure chamber, within 1 minute, until it reached the desired pressure of 6.3 MPa (sub supercritical CO₂ condition), which was indicated by the pressure gauge. After 10 min of treatment with sub-supercritical CO₂, the pressure was lowered to atmospheric pressure within 2 minutes by slowly opening the gas outlet valve. The tempeh was then collected aseptically from the pressure chamber using sterilized tongs, placed in a sterilized container, and stored in a refrigerator for further study. The processed and unprocessed tempeh (control) were analysed further for colour and texture, before and during the storage period.

Tempeh processed with sub-supercritical CO₂ (6.3 MPa, 25 °C for 10 minutes) and unprocessed tempeh were employed as the treatment and control, respectively, in this experiment. All of the tempeh was stored for 5 days. Tempeh was stored as follows: tempeh samples were placed on a Styrofoam plate and covered with plastic film then stored at 20, 30 and 40 °C with the same relative humidity. Observations on quality parameter changes (C) were carried out by measuring the quality attributes represented by L* and texture. Observations were made daily. A storage time of 5 days was chosen considering that the shelf life of fresh tempeh is normally around 24 – 48 hours at room temperature (28 – 30 °C).

The Accelerated Self Life Test (ASLT) with the Arrhenius model was used to determine the shelf life of tempeh, in which, if the food product deteriorates faster, then the shelf life is determined based on extrapolation to storage temperature. Changes in the quality factor were used to determine the degree of decrease in quality. Data were transformed into a kinetic plot, and an appropriate kinetic parameter model was obtained. The quality decrease in food is given by equation (1).

(1) d Q d t = k . Q n

Where: Q is the quality factor, t is time, k is a rate constant that depends on temperature, n is a degree factor or reaction order and dQ dt^-1 is the change in the Q factor per unit of time.

Most of all, a decrease in food quality includes zero-order (order 0) and first-order (order 1) reactions. The Arrhenius correlation chart was generated by evaluating the rate constant (k) at three different temperatures. The rate constant (k) was predicted by extrapolating the correlation between ln k and 1/T at three temperatures. Shelf life is determined on the basis of the most influential factors on the product. One of the factors that can affect a product’s shelf life is temperature. The Arrhenius kinetic approach was used to determine the shelf life and temperature limit factor. The equation for the Arrhenius model is shown in equation (2).

(2) k T = k o e − E a R T

Where: k_T was the reaction rate constant of quality degradation, k_o was a constant (frequency factor, not dependent on temperature), E_a was activation energy, T was absolute temperature (K) and R was the gas constant (8.341 J.mol^-1.K^-1).

The zero-order and first-order quality reaction was measured by using equation (3) and equation (4), respectively (Labuza and Szybist, 2001).

(3) Zero   order:       t = ( C 1 − C o k

(4) First   order:       t = l n C 1 C o k

Where: C_o was the initial quality value of storage, C_t was the quality value at the storage time t, k was the reaction rate constant and t was the storage time (days). The determination of the order of the most suitable reaction was performed by selecting the equation with the highest R².

The surface colour analysis of tempeh was evaluated as the CIE L*a*b* value and LCH colour scale using a colour difference meter (TC-1500, Tokyo, Japan). Results were expressed as L* (Lightness), a* (redness) and b* (yellowness).

The total colour difference (ΔE*) between the control and the processed tempeh was obtained using the following equation (5):

(5) Δ E * = Δ L * 2 + Δ a * 2 + Δ b * 2

Where: ΔL*, Δa* and Δb* were the differences between L*, a* and b* after treatment and the L*, a* and b* values of the standard colour. The standard colour used in this experiment was the L*, a* and b* values of the unprocessed tempeh (control), which was L*=76.6, a*=3.1 and b*=7.5.

Statistical analysis was carried out using Microsoft Excel 2003 and Statistica 8.0 StatSoft software. Analysis of variance (ANOVA) was used to study differences between samples. A software program using Duncan’s multiple range test was used to compare treatment means. A value of p <0.05 was considered statistically significant. All experiments were performed with at least three replicates.

Processed and unprocessed (control) tempeh were used as the models in this experiment. Colour is very important to the sensory nature of tempeh because it is the first characteristic observed by the consumer. The colour of tempeh produced by the growth of mould was influenced by changes in the chemical composition of the tempeh and storage temperature. The lightness (L*) value of the control was high, approximately 75.7 on the initial day, and then the lightness darkened to 57 at the third day of storage. The initial L* value of the tempeh processed with subsupercritical CO₂ glimmered to 74.4 and decreased to dark (69.3) by the fifth day (Table 1 and Table 2).

The data obtained from the experiments were plotted on a graph of the relationship between the degradation in the quality of L* and texture, and the storage time at various temperatures. Based on the correlation coefficient (R²) of texture and L* (Table 3), the rate of change in the quality of tempeh followed the first-order reaction model. A higher correlation value indicates a faster decline in reaction product quality. This was in agreement with the findings of Ahmed, Shivhare and Raghavan (2001) that the degradation of betanin, a natural colour compound in beets induced by heat, followed a first-order reaction.

Figure 2 shows that the lightness value (L*) weakened during storage at various temperatures, where it developed from light to dark. The colour of fresh tempeh is brownish yellow, due to the compounds furosine, hydroxymethylfurfural (HMF) and acrylamide, which are the products of the Maillard reaction in beans (Zilic et al., 2014). Tempeh is made from cooked soybeans which have been heated to boiling temperature. In addition, during the depressurization process of the high-pressure CO₂ treatment, the mycelia of the mould are wiped from the surface of the tempeh, resulting in a brownish-yellow colour appearing on the beans. Moreover, a study performed by Handoyo and Morita (2006) found that over-fermentation of tempeh that occurred during storage could bring about protein depletion and produced a blackish-brown colour.

The hard texture of tempeh increased significantly during storage (p <0.05) (Table 1 and Table 2). The texture of tempeh is dense, compact and sliceable. It is formed by soybean cotyledons intertwined with the mycelia of moulds. As mould grows, it produces fluffy white mycelia which bind the beans, squeezing and penetrating the cell walls to create a cake texture and simultaneously producing enzymes that cause softening of the beans due to hydrolysis of various compounds during fermentation (Duniaji et al., 2019; Jones et al., 2017; Wati et al., 2020).

The influence of temperature on the reaction rate was described using the Arrhenius equation. The regression equation and value of the R² data for tempeh at 20, 30 and 40 ℃ are shown in Table 3. The lightness (L*) decreased faster than the texture, which was indicated by the slope values. Plotting ln k against 1/T produced a linear regression of the Arrhenius model in which the slope represents the E_a value (Table 4). The E_a values of L* and texture were 12.27 and 25.59 kcal.mol^-1, respectively, indicating that lightness was more sensitive to temperature. The sensitivity of quality parameters to changes in temperature can also be evaluated based on the value of the correlation coefficient R², where the greater the value of R², the greater the relationship between changes in the rate constant (k) and temperature. The dates on which characteristic limits for processed tempeh were attained with respect to lightness characteristic criteria were 2.43, 4.88 and 9.36 days at storage temperatures of 20, 30 and 40 °C, respectively, while those with respect to texture characteristic criteria were 8.6, 3.7 and 1.4 days at storage temperatures of 20, 30 and 40 °C, respectively. The shelf life was defined as the earliest date of all the dates on which characteristic limits were attained when each characteristic criterion reached its limit. Therefore, the shelf life of processed tempeh was estimated to be 2.43 days at 20 °C, 3.7 days at 30 °C and 1.4 days at 40 °C. The dates on which characteristic limits were attained for the control (unprocessed tempeh), with respect to lightness characteristic criteria, were 3.33 days, 2.90 days, and 2.56 days at 20, 30 and 40 °C, respectively, while those with respect to texture characteristic criteria, were 6.89 days, 4.47 days and 2.99 days at 20, 30 and 40 °C, respectively. Therefore, the shelf life of unprocessed tempeh was estimated to be 3.33 days at 20oC, 2.90 days at 30 °C and 2.56 days at 40 °C. Summarizing the results, the shelf life processed tempeh was longer than that of unprocessed tempeh at 30 °C. However, the shelf life estimated in this study cannot be applied to all tempeh, because many factors including consumer palatability and consumer perspective, also play vital roles.

All food expiration dates could be established as selfapplied safety factors by each producer. For tempeh, expiration dates may not be mandatory because tempeh categorized as a fresh food product has a very short shelf life, and the spoilage of tempeh is easily detectable by looking at the colour, texture and aroma. Therefore, an expiration date is not necessary. From the point of view of microbial safety, tempeh, fermented soybean, is a reliably safe food because bacteria, yeasts and moulds that grow in tempeh have their own specific role. R. oligosporus, an important fungus in tempeh, is known to produce antibiotics against bacteria (Kobayasi, Okazaki and Koseki, 1992; Wang et al., 1969). Bacillus subtilis, the most common bacteria in tempeh, contribute to the production of fatty acids and isoflavones (Barus et al., 2017; Kanghae, Eungwanichayapant and Chukeatirote, 2017). The role of yeast in tempeh is not clear (Nout and Kiers, 2005; Pleva et al., 2018); however, the authors’ previous results show that co-culturing Saccharomyces cerevisiae with R. oligosporus in soybean fermentation produced tempeh with a pleasant yeast/tapai aroma that was liked by panellists (Kustyawati, Nawansih and Nurdjanah, 2017). Further studies on the role of yeast in tempeh production are needed.