Vapour–phase activities of essential oils against antibiotic sensitive and resistant bacteria including MRSA


Valerie Edwards-Jones, Research, Enterprise and Development, Manchester Metropolitan University, Manchester, UK. E-mail:


Aims:  To determine whether essential oil (EO) vapours could reduce surface and airborne levels of bacteria including methicillin-resistant Staphylococcus aureus (MRSA).

Methods and Results:  The antibacterial activity of geranium and lemongrass EO individually and blended were evaluated over a range of concentrations by direct contact and vapour diffusion. The EO were tested in vitro against a selection of antibiotic-sensitive and -resistant bacteria, including MRSA, vancomycin-resistant Enterococci (VRE), Acinetobacter baumanii and Clostridium difficile. An EO blend containing lemongrass and geranium was used to formulate BioScentTM that was dispersed into the environment using the ST ProTM machine. The effects were variable depending on the methods used. In a sealed box environment, MRSA growth on seeded plates was reduced by 38% after 20 h exposure to BioScentTM vapour. In an office environment, the ST ProTM machine dispersing BioScentTM effected an 89% reduction of airborne bacteria in 15 h, when operated at a constant output of 100%.

Conclusions:  EO vapours inhibited growth of antibiotic-sensitive and -resistant bacteria in vitro and reduced surface and airborne levels of bacteria.

Significance and Impact of the Study:  Results suggest that EO vapours, particularly BioscentTM, could be used as a method of air disinfection.


Essential oils (EO) are naturally occurring organic compounds derived from plants usually by steam distillation. They can be extracted from herbs, flowers, trees, vegetables or spices and from the leaves, flowers or bark (Tisserand and Balacs 1996). EO and their components are widely used in medicines, in the food industry (Fisher et al. 2007) and in cosmetics and fragrances (Schelz et al. 2006). The ratios of the major and minor components characterize the EO and further chemotypes can be identified based on the levels of the major characterizing components.

In vitro studies have shown that EO and their vapours can inhibit a range of bacteria including MRSA and other antibiotic-resistant healthcare-associated bacteria. Staphylococcus aureus is susceptible to the vapours of tea tree (Carson et al. 2002) patchouli and geranium oil, but not to lavender and CitricidalTM (Edwards-Jones et al. 2004). The MRSA strains showed no susceptibility to the vapours of individual oils except one strain, where tea tree oil showed limited effect. However, antimicrobial activity increased when certain EO or extracts were blended together. Increased zones of inhibition were observed with the two MRSA strains when combinations of patchouli and tea tree, CitricidalTM and geranium, and CitricidalTM and tea tree were used (Edwards-Jones et al. 2004). The vapour activity of oregano, perilla, tea tree, lavender and clove oils was assessed against Trichophyton mentagrophytes in a sealed box and shown to be fungicidal, and occurred in a time-dependent manner (Inouye et al. 2006).

As opposed to most antimicrobial agents currently used for air disinfection, EO are low in toxicity (Inouye et al. 2003) and could be used in a healthcare environment while patients and staff are present. They possess a unique property of high volatility that is not seen in other antimicrobial agents. Unlike hydrogen peroxide vapour (Otter et al. 2007) that can sterilize the rooms and surfaces in an unoccupied room, EO can reduce airborne contamination while occupied.

This study was designed to identify a blend of EO that could be dispersed safely by the ST ProTM machine and where they had identifiable antimicrobial effects against a range of healthcare-associated infectious agents but most notable MRSA.

Materials and methods

Essential oils

The EO used in this study were Geranium (Pelargonium graveolens) and lemongrass (Cymbopogon flexousus). EO were supplied by Essentially Oils Limited (Oxfordshire, UK) in 100 ml amber glass vials that were stored in the dark at room temperature. These were analysed by thermal desorption gas chromatography mass spectrometry (TD-GC-MS) to determine their authenticity and purity.

ST ProTM machine

The ST ProTM machine (Scent Technologies Ltd, Wigan, UK) is routinely used in the healthcare environment as a fragrance generator that disperses aromas into the atmosphere without heat or vibration. It releases vapours into the air by means of negative and Venturi airflow whereby the centrifugal fan creates a vortex within the chambers that is directed into the essence vessel, forcing the vapour out into the environment.

The output level of the ST ProTM machine can be selected (0–100%) and by altering the volume, the fan speed alters accordingly. Maximum (100% output) will treat 575 cu m of air per hour, which in turn can increase up to 800 cu m when in the environment, dependant upon the external conditions (i.e. people traffic, air conditioning, etc.)

It is currently used in hospitals in the United Kingdom to enhance the hospital environment by neutralizing malodours. The machine was used in this study to disperse a natural antibacterial product containing EO, BioScentTM, which contains lemongrass and geranium EO.

Determination of minimum inhibitory concentration (MIC) of EO

The agar dilution assay was used to determine the MIC of geranium, lemongrass and BioScentTM against a selection of antibiotic-resistant and -sensitive bacterial strains. Gram-positive bacteria (n = 13) included Enterococcus sp. (n = 2), methicillin-resistant Staphylococcus aureus (MRSA; n = 4), methicillin-sensitive Staphylococcus aureus (n = 2), Staphylococcus saprophyticus, vancomycin-resistant enterococci (VRE; = 3) and Clostridium difficile. Gram-negative bacteria (n = 26) included Escherichia coli sp. (n = 8), Citrobacter sp. (n = 2), Klebsiella sp. (n = 5), Acinetobacter baumanii (n = 5), Pseudomonas sp. (= 3), Proteus rettgeri and Serratia sp. (n = 2).

A series of twofold dilutions of the EO and BioScentTM ranging from 4% (v/v) to 0·03% (v/v) was prepared in Mueller Hinton agar (MHA) with 0·5% (v/v) Tween-20 to allow dispersion of the oils in the aqueous culture medium. Plates were inoculated with 1 μl spots containing a 1/100 dilution of an overnight broth culture (c. 106 CFU ml−1) using a multipoint inoculator (Mast Diagnostics, UK). MHA agar containing Tween without EO was inoculated before and after the experiments to assess the growth of organisms and eliminate false-positive results owing to oil carry over on the inoculating pins. The MIC was defined as the lowest concentration that completely suppressed visible growth. The experiment was carried out in triplicate.

Disc diffusion assay and antimicrobial vapour evaluation in petri dish

The antimicrobial activity of lemongrass, geranium, BioScentTM EO and vapours were assessed against the same selection of antibiotic-resistant and -sensitive bacterial strains described in the MIC experiments. Lemongrass, geranium and BioScentTM (20 μl) were placed onto individual 6-mm filter paper discs. The surface of a MHA plate was inoculated with a standard inoculum of each micro-organism containing 106 CFU ml−1, to produce a semi-confluent growth. The EO-impregnated discs were placed into inoculated surface for the disc diffusion assay. The EO-impregnated discs were placed on the lid of the petri dish for the vapour assay and hence only EO vapours were capable of inhibiting the organism and were incubated for 24 h at 37°C. The antimicrobial effect of the EO and their vapours was assessed by measuring any zone of clearing by the micro-organisms under test. The experiment was carried out in triplicate.

Determination of the antimicrobial effect of EO vapours as a surface decontaminant in a sealed box

BioscentTM was most inhibitory to staphylococci and thus further experiments were carried out to assess vapours in the larger environment of a 64 l (L605 mm × W370 mm × D280 mm) sealed box against four strains of Staphylococcus sp.

MHA plates were inoculated with c. 200 CFU of Staphylococcus haemolyticus (NCTC 11042), Staphylococcus epidermidis (NCTC 7944), Staph. saprophyticus (NCIMB 7811) and EMRSA 16.

Exactly 2 ml of BioscentTM was pipetted onto a glass petri dish and placed in the centre of the plastic sealed container (64 l capacity). The plates were arranged around the dish within the container (on lid, base and side), and exposed to oil vapour for 20 h at room temperature. The plates were then removed from the box, incubated for at least 48 h and counted. The mean count from the three plates was calculated as the position in the chamber did not affect the results. Control plates were prepared and processed in exactly the same manner, but in the absence of EO in the container. The experiment was carried out thrice.

Antimicrobial vapour evaluation distributed by ST ProTM machine dispersal in an office environment

Initial experiments were undertaken to determine the time taken to reduce airborne counts using the ST ProTM machine dispersing BioscentTM at a maximum of 100%; and a subsequent output level to maintain the reduction (data not shown). Following this, a second experiment was undertaken in an office environment that was unoccupied for the first 15 h, then occupied by two people during the subsequent sampling period. Windows remained closed and access to the room by other people was not restricted. Air samples and settle plates were taken over six distinct sessions.

Air samples and settle plate samples taken at all sites at time zero and then the ST ProTM machine was set at 100% to run overnight (15 h). The samples were taken at 15 h and then the ST ProTM machine output was reduced to 30%. Samples were then taken at 19, 63 and at 87 h with the dispersal output still set at 30%. At 88 h, the machine was removed from the office and the last sample was taken at 111 h.

The room (25 m2) used for this part of the study was an office, with a window, radiator and one doorway. Settle plates were arranged on either side of the door and the window (A, B, C), and the air sample was taken in the middle of the room (D).

Sampling the air using settle plates

At time 0 h, Columbia blood agar (CBA) culture plates were opened for a period of 1 h. The plates were then closed, labelled and incubated at 37°C for 48 h. The number and identity of bacteria isolated were noted. Settle plates were placed in the same position at every sampling time.

Sampling using the air sampler

An air sample machine (Germ Sampler, GS 100; Desaga, GmbH) was used to sample the air as per manufacturer’s instructions for a period of 5 min, which was equivalent to 500 l of air (100 l min−1). The air sample was taken from the centre of the room onto CBA plates. These were then incubated at 37°C for 48 h. The experiment was carried out in duplicate. The results were calculated as CFU m−3.

TD-GC-MS analysis of EO and machine vapour

Lemongrass, geranium and BioScentTM were analysed using TD-GC-MS. In addition, headspace analysis was also carried out. Finally, the BioScentTM vapour was collected from the ST Pro machine output in a 20 l plastic bag. Using a syringe, 520 ml of vapour sample was taken from the bag via a septum, exhausted through a Tenax TA adsorbent cartridge and further analysed by TD-GC-MS. The experiment was carried out thrice.


Determination of MIC and zone of inhibition (ZOI) of EO

The general trends observed (Table 1) were that lemongrass EO individually by direct contact were most inhibitory to Gram-positive bacteria and geranium the least inhibitory. The single EO were effective at MIC between 0·16% and 2·4% and after effective blending was further reduced with a mean of 0·79%.

Table 1.   Mean minimum inhibitory concentrations (MIC) and zone of inhibition (ZOI) for lemongrass and geranium and a blend against Gram-positive (= 13) and Gram-negative (n = 26) bacteria
Essential oilsMIC (%± SE)Contact ZOI (mm ± SE)Vapour ZOI (mm ± SE)
G +ve (n = 13) G –ve (n = 26)G +ve (n = 13) G –ve (n = 26)G +ve (n = 13)G –ve (n = 26)
Geranium1·15 (0·004)2·4 (0·004)12·9 (3·9)9·25 (2·9)011 (3·4)
Lemongrass0·16 (0·006)0·34 (0·0006)24·3 (7·6)18·3 (5·7)35·6 (12·2)24 (7·5)
BioscentTM0·54 (0·002)0·79 (0·001)25 (4·7)18·9 (5·9)16·8 (5·3)7 (2·2)

BioscentTM vapours inhibited all Gram-positive (n = 13) micro-organisms demonstrated by MIC and ZOI (Table 1). The vapours (8 mm distance) inhibited antibiotic-resistant bacteria including MRSA with a ZOI of 29 mm, VRE 13 mm (±1 mm) and A. baumanii 11 mm (±1·5 mm).

Antimicrobial vapour evaluation in sealed box

In the box environment (64 l capacity), growth of the four Staphylococci sp., including MRSA were reduced by 38% after 20 h exposure to BioScentTM vapour (Fig. 1). Staphylococcus epidermidis was most susceptible and S. haemolyticus was the least susceptible.

Figure 1.

 Percentage reduction of bacterial growth (seeded plates) after 20 h exposure to blend vapours in a sealed box.

Antimicrobial vapour evaluation distributed by machine dispersal in an office environment

The airborne counts were much higher than those from settle but the predominance of staphylococci and micrococci were observed on both. A reduction in airborne counts on CBA (Fig. 2) was observed 15 h after the ST ProTM machine was set at a 100% constant output. The lowest count was observed at 63 h with the ST ProTM machine set at 30%. At 63 and 87 h with the ST ProTM running at 30% output, the airborne counts on CBA were 50% lower than samples taken at time O before the ST ProTM machine was switched on.

Figure 2.

 Air sample counts on Columbia blood agar over six distinct sessions.

TD-GC-MS analysis of EO and machine vapour

The TD-GC-MS analysis demonstrated that the profiles of the main constituents differ in lemongrass, geranium, BioscentTM and their resulting vapours. The major components of the individual oils being geranial and β-citronellol where also identified as the major components in BioscentTM but when headspace or collected vapours from the ST ProTM machine were analysed, this demonstrated major peaks to be 2-carene, α-pinene and limonene (Table 2).

Table 2.   Essential oil and headspace major constituents in descending order by their relative concentration
GeraniumLemongrass BioscentTM BioscentTM vapour headspace analysisHeadspace analysis of BioscentTM vapour from ST ProTM machine
β-citronellol, 36%Geranial, 38%Geranial, 22·26%Limonene, 33·2%2-carene, 18·43%
Citronellyl formate, 12%Neral, 28%β-citronellol, 18·38%α-pinene, 19·3%α-pinene, 17·53%
Geraniol, 4·5%Limonene, 4·7%Neral, 18·05%β-myrcene, 10·7%Limonene, 14·12%
Isomenthone, 4%Geranyl acetate, 2·8%Citronellyl formate, 6·88%Camphene, 8·6%Camphene, 10·43%
Rose oxide isomers, 2%2-carene, 2·6%Geraniol, 6·13%2-carene, 4·4%β-myrcene, 4·52%


This study has demonstrated that lemongrass and geranium EO are inhibitory to antibiotic-resistant bacteria at very low concentrations as little as 0·06% MIC. Gram-positive bacteria were more susceptible than Gram negative particularly when in contact with lemongrass or BioscentTM. The differences in susceptibility have been observed and previously reported (Smith-Palmer et al. 1998; Inoyue et al. 2001; Delaquis et al. 2002) and have been attributed to the presence of cell wall lipopolysaccharides that reduces the rate of diffusion. The greatest ZOI was demonstrated by lemongrass and BioscentTM that may be attributable to the high content of geranial as demonstrated by TD-GC-MS analysis. In accordance with the other and previous studies in our laboratories (Edwards-Jones et al. 2004), blending or combining EO increases the antimicrobial potency by direct contact and their vapours. Further studies in our laboratories are assessing the nature of this inhibition, i.e. whether this inhibition is bacteriostatic or bacteriolytic.

Experiments conducted within a sealed box (64 l) with the blend of oils created in our laboratory demonstrated a reduction in growth of four Staphylococcal sp. including EMRSA 16. These experiments were carried out over 20 h in a 64 l box but other studies have shown reduction in less time in a 1 l box (Inoyue et al. 2006). Inouye demonstrated that the vapours of geranium were more effective than tea tree against T. mentagrophytes in a box vapour assay over 24 h but the order was reversed over a short exposure of 4 h. By blending oils (more bioactive components) in this study in both the box and utilizing the ST Pro machineTM, it is hoped that different components of the vapours dispel at varying rates thereby inhibiting a range of bacteria overtime. The TD-GC-MS analysis demonstrated this by the presence of 2-carene, α-pinene and limonene in particular. Interestingly, the quantity of limonene analysed in the BioscentTM vapour was much higher than that observed in the EO solution and could be accountable for the greater ZOI observed compared with that recorded in direct contact.

The ST ProTM machine was effective in reducing airborne bacteria as shown by the reduction in microbial counts using the air sampler. Results from settle plates did not show any significant reduction after the ST ProTM machine was switched on, although the counts were generally lower.

The results in this study demonstrate that in an office environment ST Pro machine can reduce the airborne bacteria by 89% after 15 h saturation at a constant output of 100%. Note that the saturation was carried out overnight when the office was not inhabited. The study has also shown that samples taken at 111 h (24 h after the ST ProTM was switched off) were still lower than samples taken at time O before the ST ProTM machine was switched on. Further studies are in process to determine the percentage of saturation required and longevity in relation to room size. The long time course of this experiment would certainly allow for an accumulation of volatile vapours into the air. Further studies are being conducted to establish whether the instrument could be beneficial in a healthcare environment.


The authors are grateful to Scent Technologies for funding this research and also to Helen Duxberry, Anne Leahy-Gilmartin and Alan Surrey for their technical assistance.