Proceedings of the International Conference « Maladies tropicales, aspects humanitaires et économiques », Luxembourg, June 3-4 2008.

Bactericidal properties of Artemisia annua tea and dosimetry of artemisinin in water by fluorescence under UV light.

Pierre Lutgen, Bernard Michels
Luxembourg Senior Auditors and « Iwerliewen fir bedreete Volleker »
Contact address : lutgenp@gms.lu, and for fluorometry only : michelb@pt.lu


When Artemisia annua tea is exposed to UV light of 365 nm wavelength it becomes fluorescent to a high degree. This phenomenon can be developed into dosimetry for artimisinine in water.

The addition of Artemisia annua tea to contaminated water drastically reduces the bacterial load and levels lower than those obtained by boiling can easily be obtained.


Lorsqu'on expose la tisane d'artemisia annua à une lumière UV de 365 nm la solution est fluorescente. Ce phénomène peut aisément être utilisé pour développer une dosimétrie des substances actives de la tisane.

L'ajout de tisane d'artemisia annua à une eau polluée réduit fortement la charge bactérienne.

Il est même possible d'obtenir une contamination inférieure à celle obtenue par ébullition

Background and Introduction

Artemisia annua, a Chinese herb used since more than 2000 years, has been rediscovered some 30 years ago as a potent cure against different forms of malaria. The antimalarial effect essentially is due to the active substance artemisinine present in the dried plant. The concentration of this endoperoxide however varies greatly depending on climate, geology, harvest period etc. It is thus important to know the concentration of the active substance in plant from different origins before recommending their tea or extracts as a cure.

Many attempts to find a simple, rapid and cheap technique for dosimetry have been elusive (1), and HPLC basically remains the sole available and reliable technique. Such equipment however is too expensive and difficult to maintain in developing countries.

It is known that Artemisia tea not only cures malaria, but many other diseases of viral or microbial origin, or even cancer. Some preliminary results obtained at the LCDI laboratory at Marange-Silvange near Metz with the Microtox equipment measuring the light emitted by vibrio fischeri bacteria indicated that Artemisia tea indeed has a noticeable effect on the luminescence of these bacteria. This could be confirmed at the same laboratory by using the Pallcheck Luminometer (ATP phosphatase) technique. Artemisa annua tea has a strong bactericidal effect on contaminated water. A cup of room temperature tea added to a litre of water is sufficient to show this effect.

By running some tests at the same LCDI laboratory to study the effect of UV light on contaminated water it was noticed that artimisia annua tea becomes strongly fluorescent under a 365 nm UV lamp.


The lamp used was a Vilber-Lourmat CN 6 box equipped with a curtain and the 365 nm lamp (fig 3.). Photographs of the samples can either be taken through a window at the top or through the front gate by raising the curtain. This allows to check the fluorescence horizontally and vertically through the thin water layer in a Petri dish. We also run experiments with a lamp used for checking bank notes (fig 4.) The antiseptic properties of artemisia annua tea were measured in a Pallman Luminometer based on the ATPase reaction with specific enzymes. For chromatography we used the AGILENT GC-MS at the Laboratoire National de la Santé in Luxembourg.

The Artemisia annua tea (fig 1) was obtained from a sample obtained from the 2007 Colabor plantations at Walferdange, Luxembourg, on seeds supplied by ACP Paludisme-France. The herb had been dried and packed at the “Téi vum Séi” facilities at Winseler Luxembourg, on semi-industrial equipment purposely built to this end. Artemisia annua tea with higher artemininin content was obtained from Pedro Melillo, Campinas, Brazil. Commercial tea samples of various brands were obtained at the local shop (fig 2).

The Artemisia annua seeds were of 4 origins: One Garden-Wyoming, ACP-France, Ruehlemans – Germany and Keipes-Luxembourg. Tea was prepared by adding boiling water to tea bags or to 5 g of the dry herb in a cup and leaving to infuse for 10 minutes before filtering. The Readycult Coliforms 100 enrichment broth was obtained from the Merck Cy at Darmstadt.

Fig. 1 : Artemisia annua leaves from Luxembourg

Fig. 2: Commercial tea samples used in this work

Fig.3 : Viber-Lourmat UV lamp with box

Fig.4 : UV lamp used for checking bank notes


1. The fluorescence of water extracts of artemisia annua

Tea from Artemisia annua has a strong fluorescent light emission under irradiation with a 365 nm 15 W lamp. Artemisia vulgaris and artemisia absinthum also give a very weak fluorescent signal, a. absinthum slightly more than a. vulgaris, several orders of magnitude weaker as for Artemisia annua as could be checked by dilution. Tea prepared with green leaves from Artemisia annua seedlings also is fluorescent. But while the infusion from dry leaves is only fluorescent in the top layer, indicating some strong absorption of this UV wavelength in the skin, the infusion from fresh leaves is brightly fluorescent in the bulk. Infusions made from Artemisia annua seeds (very weak signal) or stems give a similar fluorescence in the bulk of the solution.

Also tea prepared after 4 extractions of the same dry herb has a weaker fluorescent signal. By heating previously prepared tea in a microwave oven, the solution darkens and fluorescence changes. The fluorescence of different batches of tea prepared under apparently the same conditions is not necessarily the same.

Fig. 5: On the left, Petri dish with solution of green tea (Lipton)
without fluorescent signal (not visible in this photograph

Similar results were obtained with the UV lamp of Fig. 4 by depositing droplets of the samples on filter paper in front of this lamp.

None of the other commercial tea samples (fig.5) gives any fluorescent signal. The fact that none of the other 8 teas gives a fluorescent signal, although they all contain numerous essential oils, some probably identical to those of Artemisia annua, suggest that the molecule of artemisinin might be responsible for the fluorescence. Some endoperoxides containing singlet oxygen are known for their fluorescence (4), (5).

None of 3 samples of ACT pills imported from Cameroon or Brazil however gave a fluorescent signal, neither in water, nor in whey (milk).

Adding ferrous sulphate or chloride to the tea darkens the solution and strongly diminishes fluorescence. This is indicative of a Fenton type reaction of iron with artemisinine Raising the pH to 9 by adding NaOH or lowering it to 4 by addition of chlorhydric or acetic acid, changes and lowers the signal but doesn't make it disappear.

Trying to extract the fluorescent substance in the tea with hexane or trichloroethylene fails. The organic layers stay clear of all fluorescence and the fluorescence of the tea doesn't change.

Strange however is the fact that tea obtained by infusing the herb with room temperature water also displays fluorescence. This is completely contradictory with the opinion that artemisinine is not soluble in water.

Possibly artemisinine forms a soluble complex with another molecule present in the tea and/or water. Scopoletin is also present in Artemisia annua tea and has fluorescent properties. But scopoletin's fluorescence is extinguished by hydrogen peroxide, which doesn't occur after the addition of hydrogen peroxide to the tea.

Fig.6: Artemisinine annua tea from 2 origins:
Luxembourg on the left and Brazil on the right

When extracting tea from two different origins we got a first indication that the fluorescence is proportional to the artemisinine content of the dry herb. Dried leaves from Brazil contain on the average 1.0 g/l of artemisinine and those from Luxembourg only 0.3 g. The Artemisia annua tea from Brazil has a stronger fluorescence as can be seen in Fig. 6.

Extracting dry leaves of Artemisia annua at room temperature with pure ethanol doesn't give a solution with a fluorescent signal. Only if 20% of water is added to the pure alcohol before extraction does the solution fluoresce. This further substantiates the hypothesis that artemisinin forms a soluble complex with water.

Our hypothesis that the fluorescence of the artemisinin complex in water could be used for dosimetry purposes was verified on a simple device built by Bernard Michels, IFBV, Luxembourg michelb@pt.lu. The artemisia annua tea has to be diluted several orders of magnitudes for the test. This per se is no problem, but rather an advantage as it allows to detect minute quantities of artimisinine.

Fig.7 : Dosimetry curve of Artemisinin obtained for dilute solutions of artemisia annua tea on fluorescence meter built by Bernard Michels, IFBV, Luxembourg. michelb@pt.lu

If the concentration of the solution is increased linearity is lost. The response looks polynomial of the third order. The behaviour is first linear and than gets into a plateau, for higher concentrations the RLU decreases. It is a combination of fluorescent emission and light absorbance.

This to a large extent explains the empirical observations of paragraph one of this section. The peak of the curve also shifts to the left after 24 hours ageing which is visible to the naked eye by darkening of the tea.

Fig. 7a: Concentration vs. fluorescence

Fig 7.b The possibility to use this equipment and dosimetry was tentatively demonstrated on samples of tea from Brazil and and Luxembourg which based on HPLC results lie in the range of 1.0% and 0.3% artemisinin respectively. RLU was recorded with same amplification for both samples, set to teach maximum scale with Brazil sample.
2. Bactericidal properties of artemisia annua tea

It is known that Artemisia tea not only cures malaria, but many other diseases of viral or microbial origin, or even cancer. Additional indications in the literature (7) on the anti-inflammatory and antimicrobial properties of artemisinin, on inhibitory effects on hepatitis virus (M.Romero Univ of Salamanca, 2005) , inhibitory effect on protozoaria (8) and the report of several people that they had been cured from influenza by artemisia annua tea led us to make some assays concerning these properties.

Some preliminary results obtained at the LCDI laboratory at Marange-Silvange near Metz with the Microtox equipment measuring the light emitted by vibrio fischeri bacteria indicate that Artemisia tea indeed has a noticeable effect on the luminescence of these bacteria.

Fig. 8 Mictrotox results

When trying to reproduce these bactericidal tests in standard Petri dish incubation experiments we were surprised to find that after 16 to 48 hours the colony forming units had multiplied much more with tea solutions than with the control ( enterococcus faecalis and pseudomonas aeruginosa at the Laboratoire des Services de la Gestion de l'Eau in Luxembourg and pseudomonas aeruginosa at the Laboratory of Microbiological Ecology at the University of Gent)

In a second series of experiments we used the ATPase luminescence technique on the Pallman Luminometer. The bactericidal effect of artemisia annua tea on three different samples of water were tested : water from the sewage plant at Stenay-F, river water from Niederanven-L and tap water. The experiments were run by Haythem Riahi from the Pall company.

Fig. 9 : Microbial concentration for different samples
as measured by ATPase Luminosity Units

Fig.10 : Results obtained on sewage water from Stenay-Metz

Fig.11 : Results obtained on river water from Niederanven

Fig.12 : Results obtained on tap water

In all cases adding 50% of fresh artemisia annua tea had a stronger bactericidal effect than boiling the contaminated water for 5 minutes or irradiating it for 10 minutes under a 365 nm UV lamp.

A simple method to check the bacterial concentration in water and the bactericidal effect of an additive, is the Readycult Coliform 100. After 24 hours at 35°C this enrichment broth for coliforms is developing a green colour if bacteria have multiplied. This happens even after 4 hours with the river water of Niederanven.

But if fresh artemisia annua tea is added to the mixture at the beginning no green colour develops as can be seen on the right side of Fig.13. The same happens when the mixture is exposed to solar irradiation (Fig. 14). This effect is well known and has been used to promote the Sodis water sterilization technique (9)

Fig. 13 and 14 : On the left hand side bacteria have multiplied leading to the green color of the Readycult coliform mixture. On the right side, no bacteria have developed under the influence of fresh artemisia annua tea or solar radiation
One strange effect had been noticed when generating the data of Fig.11 as shown by the second bar of that chart. Tea which had been prepared the day before led to a significant increase in bacteria count. We were able to reproduce this puzzling phenomenon with the Readycult enrichment broth. Only freshly made tea inhibited the bacterial growth. When running the experiment with tea made the day before the mixture became rapidly green, even more than in the river water alone (Fig.15). This goes in line with the CFU results described in the 3d§ of this section.

We also noticed that although fresh tea initially inhibited the bacterial growth, after 8 hours it seemed to loose its power and the solution developed a dark green colour. In all cases the mixture on the far right of Fig.14 which had been SODIS treated however stayed clear and yellow for at least 24 hours.

If this is confirmed it would mean that artemisinin has a short shelf live in tea and that it is mandatory to only use freshly prepared tea in the fight against malaria and bacterial diseases.

Fig. 15 After 8 hours the mixture on the right prepared with 24 hours old artemisia annua tea become very dark , indicating an enhanced bacterial growth

Chromatographical results

In order to better understand the differences between different samples they were submitted to chromatographical analysis. In Fig 16 a-d the qualitative differences in composition are shown for the following samples:

  • pure arteminsin extract (origin BEST)
  • tea prepared with dried Artemisia annua leaves from Brazil (1g/L)
  • tea prepared with dried leaves from Luxembourg
  • tea prepared with dried leaves from France
  • artemisia vulgaris leaves

Fig 16. Active constituents in different types of artemisia tea.

The composition of the teas from Luxembourg and France are quite similar but differ greatly from the sample from Brazil which is a hybrid with an enhanced artemisinin content. Artemisia vulgaris doesn't contain any of these which are all of bactericidal nature. Artemisinin extract doesn't contain any of the other bactericidal substances. Others have analyzed the constitution of the essential oils of Artemisia annua (10) and find immunostimulants like coumarin, antipyretics like menthol, antiseptics like kaempferol, parisiticides like artemitin and many others This substantiates our claim that Artemisia annua tea is far from being a monotherapy, whilst the use of pure artemisinin extract is.

Fig 17: By chromatography it could be demonstrated that ageing of the tea for 24 hours had a detrimental effect on the concentration of active constituents.

It is fully recognized that this ageing effect deserves further study taking into account factors like, impurities (iron, bacteria) in the water, temperature, illumination…

Other techniques like UV absorption might be useful in elucidating the composition of artemisia annua tea, which with peaks at 280 and 324 nm, is different from that of commercial black tea.

Fig 18. UV-vis absorption scan of Artemisia annua tea extract (obtained from David van der Ha, University of Gent)

Discussion and Conclusions

A qualitative test based on this fluorescence technique would be able to select teas with a higher artemisinin content, and to eliminate fake ACT pills which inundate the market in Africa and Asia (more than 50% of the pills sold).

Further experiments need to confirm that the fluorescent signal is due to artemisinin, or a complex of artemisinin with an essential oil contained in Artemisia annua tea. In some cases indeed a synergetic effect between peroxides and other substances has been found for their bactericidal properties (2). It also needs to be checked if artesunate or artemether are fluorescent or if the ACT pills tested were fakes or if these substances have a short shelf life in the drugs. The pills we tested came from Cameroon and at temperatures above 30°C the shelf life is limited to a few days (1), (3).

It is likely that the strong fluorescence in the bulk of the solution for stems, seeds and fresh leaves is simply due to a lower concentration of the fluorescent complex in the solution. By diluting an infusion obtained from dry leaves the fluorescence also moves from the surface or skin to the bulk.

The difference in fluorescence between tea from Brazil and tea from Luxembourg containing different amounts of at artemisinin as determined by HPLC, difference which is visible to the naked eye (Fig.6) is a first indication that the phenomenon can be used for dosimetry purposes.

This was confirmed by a dosimetry curve obtained on a rudimentary device (Fig.7)

A more quantitative approach is probably possible but needs to be developed on professional spectroscopic instruments. As far as we can judge by using simple diffraction techniques, the light transmitted through a thin layer of artemisia annua tea is monochromatic. This would simplify the dosimetry.

Similar approaches are used for the dosimetry of the siderophore pyoverdine, either measuring its natural fluorescence, or the fluorescence quenching by iron salts (6).

The bactericidal effect detected during our assays could have a major impact on diseases and the death toll of several million people related to contaminated water in tropical countries. It indicates that artemisia annua tea is a complex mixture which is far from monotherapy as some people claim. It contains for example scopoletin which is known to be bactericidal. Artemisia annua tea possibly might also be used as an antiseptic for minor scratches and wounds.

But it seems prudent at this stage to only recommend the use of freshly made tea.


(1) S. Houzé et al.,(1) J.Clin Microbiol 45(8). 2007
(2) WO/1989/007396 Disinfectant compositions containing hydrogen peroxide
(3) WHO, ACT Drug, Addis Abeba workshop, 25-26 May 2004, recommendations k&l
(4) Mingqian Tan et al., Free Radical Biology and Medicine, 40, 1644-1653, 2006.
(5) Bo Song et al., J Am Chem Soc., 128, 13442, 2006
(6) Rong Xiao et al, Appl and Environ Microbiology, Apr 1998, 1473-1476
(7) J Wang et al., Antimicrobial agents and chemotherapy 50-7, 2420, 2006.
(8) Y.Mishina et al.., AAC, May 2007, vol 51.
(9) N. Aguirre, P Lutgen et Paul Schosseler. Mise en œuvre de la méthode Sodis en Colombie, Revue Technique Luxembourgeoise, 3.2006.
(10) Handbook of Biological Active Phytochemicals, J.A. Duke CRC Press, ISBN 0-8493-3670-8

This article has been submitted for peer review to Yves-Jacques Schneider, University of Louvain- Belgium, Willy Verstraete, University of Ghent-Belgium, Luc Pussemier, CERVA-Tervuren-Belgium, Dr Carlo Steffes, Santé au Travail, Luxembourg, Pedro Melillo de Magalhaes, University of Campinas-Brazil, Ivan Velez, University of Antioquia, Colombia, Pr Josph Mabingui, Faculté des Sciences, Université de Bangui, Pr Vera de Miranda, Universitate Federal de Ouro Preto, Gilbert Hansen, Laboratoire National de la Santé, Luxembourg.

A patent request has been filed.

Back to Chemical & Pesticides Page         Back to English Version

See the weather in Argentina

See here where our readers come from

Locations of visitors to this page

Tell us your opinion!

(Please include the name of this article!)

¿Which countries visit us?
¿Who and how many our visitors are?

Make a long URL a short one: