Correspondence to: Prof S. Roller, 50 Argyle Road, London W13 8AA, UK (e-mail: firstname.lastname@example.org).
Aim: To establish whether or not carvacrol and cinnamic acid delay microbial spoilage of fresh-cut fruit.
Methods and Results: Dipping of fresh-cut kiwifruit in carvacrol solutions at 5–15 mM reduced total viable counts from 6·6 to < 2 log cfu g −1 for 21 d at 4°C; however, undesirable colour and odour changes were also observed. Treatment with 1 mM of carvacrol or cinnamic acid reduced viable counts on kiwifruit by 4 and 1·5 log cfu g −1 for 5 d at 4°C and 8°C, respectively. Treatment of fresh-cut honeydew melon with 1 mM of carvacrol or cinnamic acid extended the lag phase of the microbial flora from less than 1 d in the untreated controls to 3 d at 8°C and 5 d at 4°C. Viable counts on the treated melon were 6 log cfu g −1 lower on Day 3 at 8°C and 4 log cfu g −1 lower on Day 5 at 4°C, compared with the untreated controls.
Impact of the Study: Treatment with 1 mM of carvacrol or cinnamic acid delays spoilage of fresh-cut kiwifruit and honeydew melon at chill temperatures without adverse sensory consequences.
Consumer demand for minimally processed, ready-to-eat fruits and vegetables has led to growth in the fresh-cut industry of 10% per year since 1995 (Barth 2000). During minimal processing, spoilage and pathogenic microorganisms can gain access to the nutrients inside fruits and vegetables and multiply. As prevention of contamination is not always possible, washing and treatment with chemical disinfectants are necessary to decontaminate the surface of fresh produce. These procedures are only partially effective in reducing the microbial load and their efficacy depends on the type of fruit or vegetable treated, the surface characteristics, the conditions used (e.g. temperature, pH), and on the type of microorganism being removed. Treatment with 50–200 p.p.m. chlorine and a contact time of 1–2 min can reduce the number of microorganisms by 10–100-fold in some instances but can equally be completely ineffective in others (Madden 1992; Beuchat 1998, 2000). Furthermore, many chemical agents in current use are losing favour with both industry and the public, e.g. chlorine is corrosive, ineffective in the presence of high organic loads, may form organochlorines and may have long-term toxicological implications. A 1998 WHO report has recommended that new research should be carried out to develop new treatments for decontaminating fruits and vegetables (Beuchat 1998).
Carvacrol (C10H14O) is a major component of the essential oils of oregano and thyme (Lagouri et al. 1993; Arrebola et al. 1994). Cinnamic acid (C9H8O2) is structurally similar and occurs in cinnamon, cloves, black pepper, coriander and turmeric. Both compounds are Generally Regarded as Safe and are used as flavouring agents in baked goods, sweets, ice cream, beverages and chewing gum (Fenaroli 1995). The in vitro antimicrobial properties of plant essential oils and their components are well documented against a wide range of foodborne fungi and bacteria (Aureli et al. 1992; Beuchat 1994; Sivropoulou et al. 1996; Davidson and Naidu 2000). Recent work has shown that carvacrol can be used to inactivate Saccharomyces cerevisiae, Salmonella enterica sv. Typhimurium and Listeria monocytogenes adhered to stainless steel (Knowles and Roller 2001). However, widespread application of these compounds as natural preservatives in foods has been limited by the strong flavours associated with them. Although effective microbial control using plant essential oils in foods such as minced meat, rice, whole tomatoes, strawberries and banana purée has been reported, the concentrations used in these studies would impart a strong and possibly undesirable flavour on those foods (Smid et al. 1996; Alzamora et al. 2000; Ultee et al. 2000; Skandamis and Nychas 2001).
Preliminary work in this laboratory has shown that dipping whole and sliced fruits in cinnamic acid solutions at concentrations of 3–5 mg ml−1 (20·2–33·7 mM) delayed the onset of visible spoilage at both ambient and chill temperatures (Roller et al. 1998). In melon slices, treatment with cinnamic acid delayed the first appearance of visible spoilage from 8 to 26 d at 25°C and from 15 to 78 d at 4°C. The onset of visible spoilage on kiwifruit slices was delayed from 5 to 31 d at 25°C. However, treatment with cinnamic acid at concentrations of 3 mg ml−1 (20·2 mM) and above also rendered the fresh-cut melon and kiwifruit translucent and/or brown.
The objective of this study was to establish whether or not concentrations of carvacrol and cinnamic acid below 20 mM could be used to inhibit microbial growth on fresh-cut fruit destined for storage under chill conditions.
Materials and methods
Trans-cinnamic acid, 99 + % (C6H5CH=CHCO2H, FW 148·16) and carvacrol, 98% (C10H14O, FW 150·22) were from Sigma Chemicals Ltd (Poole, Dorset, UK). Dipping solutions (1, 5, 10 and 15 mM) were prepared in distilled water. Plate Count Agar (PCA) and Maximum Recovery Diluent (MRD) were from Oxoid (Basingstoke, UK). Honeydew melons and kiwifruit were purchased from a local retailer.
All fruit and dipping solutions were chilled overnight before processing. The peeling and cutting of the fruit was carried out at room temperature. Disposable gloves were worn during preparation of the fruit. Chopping boards and knives were disinfected with industrialized methylated spirits and air-dried immediately before use. Fruits were peeled, de-seeded and cut into wedges (kiwifruit) or slices (melon) of approximately 80 g each. Fruit pieces were dipped for 1 min in solutions containing 1, 5, 10 or 15 mM of carvacrol or 1 mM of cinnamic acid. The fruit was removed from the dipping solutions and allowed to drain on a metal rack for 30 min. The fruit pieces were packed individually in 250-ml Sterilin jars, sealed using screw-cap lids and incubated at 4° and 8°C for up to 21 d. Sampling for microbiological analysis was carried out periodically by removing one storage jar at any single time point from the incubator. The appearance (colour, texture) and odour of each sample was noted. Each fruit piece was sliced transversely into several sections. Sections were selected randomly from the middle and end to comprise 25 g. The 25-g sample was added to 225 ml of the MRD and homogenized with a stomacher for 2 min. The slurry was allowed to stand for 10 min to allow large particles to settle. Duplicate samples (1 ml) of the slurry were serially diluted in MRD, spread-plated (0·1 ml) in triplicate on PCA and incubated at 25°C for 3 d. Separate 25-g samples of fruit were used to prepare aqueous slurries in distilled water for pH determination.
As shown in Table 1, treatment with 10 and 15 mM of carvacrol followed by storage at 4°C reduced the microbial flora of the fresh-cut kiwifruit to below the detection limit of the viable counting method (2 log cfu g−1) for the duration of the trial (21 d). Treatment with 5 mM of carvacrol also reduced viable counts on the fruit to < 2 log cfu g−1 for 12 d but a count of 5·2 log cfu g−1 was recorded after 21 d of storage. By contrast, viable counts on the untreated fruit increased steadily from 3·7 log cfu g−1 on Day 1 through to 4·9, 5·6 and 6·6 log cfu g−1 on Days 6, 12 and 21, respectively. However, at the higher concentrations of carvacrol tested (15 mM) browning of the fruit wedges was observed. Furthermore, all the treatments at 5–15 mM of carvacrol imparted a pungent and unpleasant aroma to the fruit which did not abate over the course of the trial.
Table 1. Total viable count in kiwifruit treated with 5, 10 and 15 mM of carvacrol and stored at 4°C for up to 21 d. Results are means (± 0·5 log cfu g −1 ) of six replicate determinations
* The detection limit of the viable counting method was 2 log cfu g −1 .
Treatment of the kiwifruit wedges with low concentrations of carvacrol and cinnamic acid (1 mM) prevented the appearance of visible spoilage and inhibited growth of the microbial flora for 5 d at both 4° and 8°C (Table 2). The reduction in viable numbers was not as substantial as seen in the presence of carvacrol at concentrations of 5 mM and above (Table 1). Treatment with 1 mM of carvacrol and cinnamic acid reduced microbial counts to a greater extent at 4°C (3·9 log cfu g−1 and 4·1 log cfu g−1, respectively) than at 8°C (1·3 log cfu g−1 and 1·6 log cfu g−1, respectively), compared with the untreated controls (Table 2). The pH of all the kiwifruit samples varied within the range 3·2–3·6 both before and after storage and irrespective of the treatment received. The aroma of both carvacrol and cinnamic acid was detectable in the treated fruit after storage but was not considered unpleasant. No visible spoilage or colour change was detected on the treated fruit for the duration of the trial.
Table 2. Total viable count and appearance of kiwifruit treated with carvacrol (1 mM) and cinnamic acid (1 mM) and stored at 8° and 4°C for 5 d. Results are means (± 0·4 log cfu g −1 ) of six replicate determinations made on two separate occasions
Temperature of storage
Total count (log cfu g−1)
Total count (log cfu g−1)
Yes (shrivelled and mouldy)
Yes (shrivelled and mouldy)
Treatment of fresh-cut honeydew melon with 1 mM of carvacrol or cinnamic acid extended the lag phase of the microbial flora from less than 1 d in the untreated controls to 3 d at 8°C and 5 d at 4°C (Fig. 1). The viable counts on the treated melon were 6 log cfu g−1 lower on Day 3 at 8°C (Fig. 1a) and 4 log cfu g−1 lower on Day 5 at 4°C (Fig. 1b), compared with the untreated controls. The reduction in counts afforded by the treatments was less pronounced on Day 7. By Day 10, all counts were very similar at approximately 8·5 log cfu g−1 irrespective of the treatment received or the storage temperature (Fig. 1). The results show that treatment of honeydew melon extended the lag phase, but once initiated growth proceeded at a similar rate in both the treated and the untreated fruit. The pH of both the treated and the untreated melon at the start of storage was 5·4–5·5, but by Days 7 and 10 at 8°C and by Day 10 at 4°C it had dropped to pH 4·4–4·9, reflecting the increase in microbial counts shown in Fig. 1. The characteristic ‘spicy’ odour of carvacrol and cinnamic acid was detected less readily on the melon than on the kiwifruit after treatment.
Microbiological criteria (IFST 1999) for nonheat processed fruits intended for use in chilled desserts recommend that yeasts are present at less than 3 log cfu g−1 immediately following production under Good Manufacturing conditions. In this study, the yeasts were not enumerated but the total viable count in the honeydew melon on the day of preparation was at or below the sensitivity limit of the plate counting method (< 2 log cfu/g), indicating that the hygienic conditions used in the preparation of the fruit met the criteria recommended for industrial practice. According to the same criteria (IFST 1999), a maximum yeast count of 6 log cfu g−1 is considered acceptable at any point in the shelf life of a fruit product. In this study, a total viable count of 6 log cfu g−1 was exceeded in the untreated kiwifruit after 5 d at both 4° and 8°C but not in the fruit treated with 1 mM of carvacrol or cinnamic acid. In the untreated honeydew melon, a count of 6 log cfu g−1 was reached earlier (Day 2) at 8°C than at 4°C (Day 5), as expected (Fig. 1a); however, treatment with either carvacrol or cinnamic acid extended the time to reach this level of contamination to 6 d at 8°C (Fig. 1a) and 8 d at 4°C (Fig. 1b).
As the results reported in this study were obtained using fresh-cut fruits stored in sealed jars, further work is needed to establish the efficacy of the treatments on fruits stored under semipermeable plastic films, which may be more typical of commercial practice.
Refrigeration at or below 4°C in combination with other preservation factors (e.g. modified atmosphere packaging) is already used widely for extending the shelf life of many ready-to-eat vegetables and salads but relatively few fruits (Ahvenainen 2000). In practice, the temperature of many domestic refrigerators is approximately 8°C and abuse temperatures of up to 15°C are not uncommon (Alzamora et al. 2000). While it was evident in this study that carvacrol and cinnnamic acid inhibited microbial growth at both 4° and 8°C, there was little evidence of synergism between the antimicrobial compounds and the physical hurdle of chilling. It is possible that carvacrol was not disproportionately more effective at 4°C than at 8°C, as its mode of action depends on migration into bacterial membranes (Ultee et al. 1999), which are less fluid at chill temperatures.
Increased susceptibility to the biocidal action of carvacrol has been reported in Bacillus cereus and Aspergillus niger at acidic pH values (Thompson 1990; Ultee et al. 1998). It is debatable whether results obtained from in vitro studies in which the pH was adjusted using mineral acids can be extrapolated to kiwifruit and melon, in which the principal acidifying agents are citric and malic acids (Lund and Snowdon 2000). In this paper, the pH of the kiwifruit (3·2–3·6) was within the range (3·1–4·0) reported in the literature but the pH of the melon on the day of preparation (5·4–5·5) was lower than published figures (pH 6·2–6·7; Lund and Snowdon 2000). However, there was insufficient data to draw conclusions about the possible increased efficacy of carvacrol in more acidic fruits.
Smid et al. (1996 ) have shown that the biocidal efficacy of cinnemaldehyde as a treatment for whole tomatoes could be improved by increasing the dipping time from 10 to 30 min. Similarly, Ultee et al. (1999 ) have shown that a 2 log cfu ml −1 reduction in B. cereus in suspension could be doubled by extending the exposure time from 5 to 25 min. It is conceivable that further improvements in the biocidal efficacy of carvacrol and/or cinnamic acid on fresh-cut fruit could be made by extending the exposure time from the 1 min used in this study to several minutes; however, it is unlikely that dipping times approaching 0.5 h would be considered practicable in an industrial processing operation.
The use of plant essential oils and their components as food preservatives is limited by the strong aromatic and flavour characteristics with which they are associated. The minimum inhibitory concentrations (MICs) for carvacrol against a range of food-borne organisms have been reported previously to be approximately 0·5–3·0 mM when tested in Brain Heart Infusion Broth at 30°C (Ultee 2000). The results presented here show that treatment with cinnamic acid or carvacrol at or very near the reported MICs afforded additional protection of fresh-cut fruits from spoilage at chill temperatures.
In this study, no attempt was made to identify the surviving organisms in the fruits treated with carvacrol and cinnamic acid and it is possible that some of them were pathogenic. The growth of pathogenic bacteria in the acidic environment of fruit is uncommon but the survival of salmonellae, listeriae and verotoxigenic Escherichia coli has been documented (Beuchat 1998). Several outbreaks and many hundreds of cases of salmonellosis including a number of deaths have been attributed to precut watermelon, cantaloupe melons and mangoes (Anonymous 1991; Madden 1992). Studies on the fate of pathogenic organisms on fresh-cut fruit treated with phenolic substances such as carcacrol or cinnamic acid would be essential to ensure that the elimination of the spoilage flora does not inadvertently enhance the survival of pathogens.