Secondary Logo

Journal Logo

Comparing charcoal and zeolite reflection filters for volatile anaesthetics

A laboratory evaluation

Sturesson, Louise W.; Frennström, Jan O.; Ilardi, Marcella; Reinstrup, Peter

European Journal of Anaesthesiology (EJA): August 2015 - Volume 32 - Issue 8 - p 521–526
doi: 10.1097/EJA.0000000000000177

BACKGROUND A modified heat-moisture exchanger that incorporates a reflecting filter for use with partial rebreathing of exhaled volatile anaesthetics has been commercially available since the 1990s. The main advantages of the device are efficient delivery of inhaled sedation to intensive care patients and reduced anaesthetic consumption during anaesthesia. However, elevated arterial CO2 values have been observed with an anaesthetic conserving device compared with a conventional heat and moisture exchanger, despite compensation for larger apparatus dead space.

OBJECTIVE The objective of this study is to thoroughly explore the properties of two reflecting materials (charcoal and zeolites).

DESIGN A controlled, prospective, observational laboratory study.

SETTING Lund University Hospital, Sweden, from December 2011 to December 2012.


INTERVENTIONS Three filters, with identical volumes, were compared using different volatile anaesthetics at different conditions of temperature and moisture. The filtering materials were charcoal or zeolite. Glass spheres were used as an inert control.

MAIN OUTCOME MEASURES Consumption of volatile anaesthetics using different reflecting materials in filters at different conditions regarding temperature and moisture. CO2 reflection by the filtering materials: glass spheres, charcoal or zeolite.

RESULTS Isoflurane consumption in an open system was 60.8 g h−1. The isoflurane consumption in dry, warm air was 39.8 g h−1 with glass spheres. Changing to charcoal and zeolite had a profound effect on isoflurane consumption, 11.8 and 10.7 g h−1, respectively. Heating and humidifying the air as well as the addition of N2O created only minor changes in consumption. The percentage of isoflurane conserved by the charcoal filter was independent of the isoflurane concentration (0.5 to 4.5%). Reflection of sevoflurane, desflurane and halothane by the charcoal filter was similar to reflection of isoflurane. Both charcoal and zeolite filters had CO2 reflecting properties and end-tidal CO2 increased by 3 to 3.7% compared with glass spheres. This increase was attenuated to 1 to 1.4% when the air was heated and humidified, and isoflurane was added.

CONCLUSION Charcoal and zeolite possess gas-reflecting properties, which can be used to conserve volatile anaesthetics. They also reflect CO2. The degree of CO2 reflection was reduced by heating and humidifying the air.

From the Department of Anaesthesiology and Intensive Care, Skane University Hospital Lund, Lund, Sweden

Correspondence to Peter Reinstrup, MD, PhD, Department of Anaesthesiology and Intensive Care, Skane University Hospital Lund, SE-221 85 Lund, Sweden Tel: +46 46 174223; fax: +46 46 176050; e-mail:

Published online 9 November 2014

Back to Top | Article Outline


Inclusion of various filters in the anaesthetic circuit has been a common practice worldwide for years. Filters have been developed to reduce bacterial contamination and maintain a warm and humid environment for the airway. The rationale for using filters with antimicrobial properties as well as for providing heat and moisture homeostasis has been established by thorough investigation.1–4 Inclusion of a filter in a partial rebreathing anaesthetic circuit reduces consumption of volatile anaesthetics. The principle was originally described by Thomasson et al.5 in 1989 using zeolite as the filter material. In 1998, the first report on the use of charcoal was published.6 In the late 1990s, a filter with reflecting properties, an anaesthetic conserving device (ACD), was developed to reduce volatile anaesthetic agent consumption during anaesthesia.5–7 In the ACD system, volatile anaesthetic agents are delivered on the patient side of the filter. The working principle of the filters is the same as that of a heat and moisture exchanger (HME). During expiration, the volatile gases are adsorbed onto the filter material. During the subsequent inspiration, volatile anaesthetics are displaced from the filter back into the gas stream, reducing the need to replenish with additional vapour.

The efficacy of inhaled sedation using the ACD during anaesthesia8,9 and in the ICU10 has been established. Inclusion of a filter in the airway increases the dead space and thereby rebreathing of CO2. However, higher arterial CO2 values in patients have been observed during the use of the ACD despite compensation for larger internal dead space.11 What actually happens when different filters for reducing volatile anaesthetics consumption, reflectors, are used, and when conditions of use vary, has not been well studied.

The aim of this study was to explore the reflecting properties of filters with charcoal and zeolite filters looking firstly at their volatile anaesthetic conserving effect, and secondly at their CO2 reflection. The reflecting materials in the filters were evaluated following exposure to varying combinations of dry, humid, cold and heated gases, both with and without isoflurane. In addition, the reflecting properties of charcoal- and zeolite-containing filters were tested for commonly used volatile anaesthetics.

Back to Top | Article Outline

Materials and methods

Experimental set-up

A Servo Ventilator 900C (Siemens-Elema, Solna, Sweden) was used in all experiments. In the baseline experiments, the Y-piece was connected directly to the lung model (2-l ISO antistatic ventilation bag; RÜSH, Kernen, Germany) with a Fisher & Paykel MR 450 humidifier (Fisher & Paykel, Auckland, New Zealand) interposed in the inspiratory limb. The volatile anaesthetic gas inlet system was connected to the low-pressure gas inlet port of the ventilator. This open set-up was used to check the validity of the system when compared with calculated values as described by Perhag et al.7

In the main experiments, the following equipment (Fig. 1) was connected in a serial manner after the Y-piece (Y): molecular sieve-reflector (R), side-stream Hewlett-Packard Agent Monitor M1025B [McPeak HB 1993; Hewlett-Packard, Palo Alto, CA, USA (M)] measuring volatile anaesthetics as well as CO2 and N2O with photo-acoustic spectroscopy, humidifier (H) controlled by a temperature probe (T), T-piece for volatile anaesthetic gas inlet system (F) and a CO2 inlet system which, through a thin plastic tube, ended in the lowest part of the ventilation bag used as a test lung (L).

Fig. 1

Fig. 1

The volatile gas inlet system (G) consisted of a precision flow meter (F) (administering atmospheric air), and a TEC 1 vaporiser (Ohmeda, Little Chalfont, Bucks., UK) (V) modified to deliver near-saturated vapour of anaesthetic agent. The vaporiser was placed on an electronic scale (G 1200, Ohaus, New Jersey, USA). A magnetic valve controlled by the ventilator allowed the anaesthetic gases to be injected only during inspiration.

Dead space consisted of tubing (7 ml), reflector (60 ml) and humidifier (85 ml). Dead space of the humidifier was kept as low as possible.

The filters were custom built by emptying Siemens HME filters (Servo humidifier 152; Siemens-Elema) of their contents and refilling them with 60 ml of glass spheres (control), zeolite or charcoal.

Back to Top | Article Outline

Glass spheres filter

The filter was filled with glass spheres with a diameter of 1 mm.

Back to Top | Article Outline

Charcoal filter

The adsorption material was granulated charcoal derived from burned coconut shells.

Back to Top | Article Outline

Zeolite filter

The adsorption material consisted of ultrastable zeolite Y (US-Y) pellets with a diameter of 1 mm, a synthetically prepared material using Al2O3 as the binder. Ultrastable zeolites represent a group of crystalline microporous silicates built by corner-sharing [SiO4]4--tetrahedral molecules forming a framework structure characterised by the presence of cavities and channels. Zeolite represents a unique framework, the FAU (Faujasite) structure type, in three dimensions connecting into micro-void cages [O(II)-cages] within the structure.12

Back to Top | Article Outline

Experimental procedure

All experiments started with a 30-min equilibration period. Initial measurements were made with the glass spheres configuration. Ventilation was set at 9.0 l min−1, frequency to 15 min−1, inspiratory time at 33% and pause time at 0%. The CO2 flow to the lung model (L) was set to produce an end-tidal concentration of 4% and this flow of CO2 was kept constant.

When a volatile anaesthetic was used, the flow through the vaporiser was adjusted to produce and maintain the desired end-tidal concentration throughout the experimental period (1 h). The different tests reported below were performed by substituting the glass spheres filter with either a charcoal or a zeolite filter. Weight changes in the vaporiser were recorded every 5 min.

CO2 reflection was monitored by the changes in end-tidal CO2 in the majority of experiments. In a parallel series of experiments, end-tidal CO2 was kept constant at 4% by adjusting the tidal volume of the ventilator to maintain constant end-tidal CO2.

The experiments were carried out with different configurations to evaluate:

  1. CO2 reflection in warm and dry, cold and dry and warm and moist air in turn, without any volatile anaesthetic. Warm, cold and moist air was produced by having the reservoir of the humidifier (H) empty with the middle part of the reservoir removed, filled with ice or with warm water, respectively.
  2. CO2 and 1.5% isoflurane reflection in warm, dry and in warm, moist air.
  3. The CO2 and 1.5% isoflurane reflection in warm, dry and warm, moist 70% N2O in O2 mixture delivered by the ventilator. In the following experiments, only the glass spheres and charcoal filter were used.
  4. The reflection of increasing concentrations of isoflurane.
  5. The reflection of 1.5% isoflurane, halothane, sevoflurane and desflurane.
  6. The reflection of one MAC isoflurane, halothane, sevoflurane and desflurane.
  7. The effect of isoflurane delivered only during inspiration or continuously.
Back to Top | Article Outline

Calculations of measurement errors

The results of the consumed volatile anaesthetics, expressed in grams per hour, can differ from the exact values due to measurement errors in the balance, the gas analyser and/or inaccuracy in the delivered volume from the ventilator. According to manufacturer's specifications, the measurement errors of the balance, gas analyser and ventilator are ±2%, ±5% and ±5%, respectively, giving a combined measurement error of 7%.

Back to Top | Article Outline


Isoflurane consumption in an open system using 1.5% isoflurane concentration

In this study, before inserting the different filters, the validity of the laboratory assembly was tested in an open system, in which isoflurane consumption was measured as 60.8 g h−1 corresponding to a calculated value of 58.9 g h−1.

Back to Top | Article Outline

The impact of an inspiratory-guided versus continuous delivery of isoflurane on the effectiveness of the charcoal filter

In all experiments, volatile anaesthetics were delivered exclusively during inspiration. The isoflurane consumption in warm dry air with glass spheres and charcoal filter was 39.8 and 11.8 g h−1, respectively. Bypassing the valve, resulting in a continuous delivery of isoflurane, did not change the consumption (39.6 and 11.5 g h−1, respectively).

Back to Top | Article Outline

Consumption of isoflurane using glass spheres, charcoal and zeolite filter

Changing from glass spheres to the charcoal or zeolite filter had profound effects on the isoflurane consumption. Warming and humidifying the air as well as adding N2O had only minor effects on isoflurane consumption (Table 1).

Table 1

Table 1

Back to Top | Article Outline

Reflection properties of the charcoal filter at different isoflurane concentrations

The isoflurane reflection using the charcoal filter was independent of isoflurane concentration from 0.5 to 4.5% (Table 2).

Table 2

Table 2

Back to Top | Article Outline

Halothane, isoflurane, sevoflurane and desflurane consumption

Reflection properties of the charcoal filter with a concentration of 1.5% of halothane, isoflurane, sevoflurane and desflurane and at MAC 0.5 and 1.0, respectively

These different, commonly used, volatile anaesthetics were reflected similarly by the charcoal filter, shown by similar conservation (Tables 3 and 4).

Table 3

Table 3

Table 4

Table 4

Back to Top | Article Outline

CO2 reflection

Compared with the glass spheres filter, the charcoal and zeolite filters increased end-tidal CO2 by 3.0% and 3.7%, respectively, in cold (17.2oC) air. To keep end-tidal CO2 constant at 4%, the minute volume had to be increased from 9.0 to 13.0 l min−1 and from 9.0 to 13.2 l min−1, respectively. CO2 reflection was reduced, when the air was heated, humidified and when isoflurane was added. No further reduction was noted when 70% N2O in oxygen was used.

In the clinical simulation, with volatile anaesthetics in warm humid air, end-tidal CO2 increased by 1.0% for the charcoal and 1.4% for the zeolite filter. In order to keep CO2 constant, the minute volume had to be increased from 9.0 to 10.1 l min−1 and from 9.0 to 10.4 l min−1 for charcoal and zeolite filter, respectively.

The CO2 reflection data are presented in Table 5.

Table 5

Table 5

Back to Top | Article Outline


The present study demonstrates similar reflection properties of charcoal and zeolite filters on volatile anaesthetics, with approximately 70% conservation of volatile anaesthetics compared with the glass spheres filter. Charcoal and zeolite filters also displayed CO2 reflection properties that were reduced when the air was warmed and humidified to physiological values. The CO2 reflection was further attenuated when isoflurane was added, resulting in a final 1% increase in end-tidal CO2 compared with the glass spheres filter.

Back to Top | Article Outline

Isoflurane reflection

The open system, in which isoflurane was delivered through the ventilator guided by the anaesthetic agent monitor, was used to address the combined accuracy of the delivered volume by the ventilator, the measured concentration of volatile anaesthetic and the scales apparatus. The actual consumption of the open system corresponded to the calculated values, indicating accuracy of these instruments.

In the filtering system, the volatile anaesthetic was delivered on the patient side of the filter and solely during the inspiratory phase. This was made possible by a magnetic valve controlled by the ventilator. Inspiratory delivery of volatile anaesthetic reduces consumption compared with the open system.5,6 However, with a charcoal filter in use, bypassing the valve did not affect the consumption. Thus, a filter with no magnetic valve will be just as effective at volatile anaesthetic sparing as a system with a magnetic valve, but the present experiments were still performed using this device.

Both charcoal and zeolite had marked reflecting properties when combined with isoflurane in accordance with previous studies, reducing the need for adding isoflurane to the system by almost 75% at 1.5% end-tidal isoflurane, compared with a glass spheres filter.5–7 Increasing the temperature from 17oC to 33oC did not change this reflection. Contrary to this, when the air was humidified, the raised water content increased the loss of isoflurane through both filter types. This increase in apparent permeability in humidified air was also seen with CO2, indicating that trapped water in the filter material attenuates the adsorbing properties of the filtering material in general. The zeolite filter was, by volume, somewhat more effective in reflecting isoflurane, as seen by a lower hourly consumption in all situations. Considering charcoal's ease of availability and the possible carcinogenic effects of zeolite, the rest of the experiments were performed using the charcoal filter.13,14

The amount of reflected isoflurane caused both by zeolite and charcoal has previously been found to be smaller at higher tidal volumes.5,6 One would therefore expect that increasing concentrations of the volatile anaesthetic would increase the loss through the filter. This was, however, not the case as seen in Table 2. This might be due to sufficient amount of binding sites existing on the reflecting materials in this filter, coping with even larger concentrations or amounts of volatile anaesthetics. The binding process in the filter material appears to be time-dependent so that the higher flow during expiration of higher tidal volumes allows the volatile anaesthetics to pass through the filter.

Back to Top | Article Outline

Halothane, isoflurane, sevoflurane and desflurane reflection

In previous investigations, the properties of the different filters have been evaluated solely with isoflurane.5–7,12 Other volatile anaesthetics are similar in chemical composition, and similar effects were seen for reflection of halothane, sevoflurane and desflurane by charcoal (Tables 3 and 4).

Back to Top | Article Outline

CO2 reflection

Rebreathing of CO2 and CO2 retention have been described but have not been thoroughly investigated.6,15 A charcoal filter used in a 17oC system was found to have profound CO2 reflecting properties compared with the glass spheres filter, resulting in an increase in end-tidal CO2. Accordingly, in order to keep end-tidal CO2 constant, the tidal volume of 600 ml had to be increased by 266 ml. However, as soon as the filter was warmed to 33oC, which in clinical practice should take less than 10 min, the CO2 reflection was reduced.16,17 Furthermore, when the filter had been humidified by a patient's expired gas over 5 to 10 min, the CO2 increase on the patient side of the filter was further reduced to 30%.16,17 In such a clinically relevant situation, without volatile anaesthetics, the CO2 increase was 1.2% for the charcoal filter and 0.8% for the zeolite. The corresponding increases in tidal volume necessary to compensate for this reflection were 233 and 153 ml, respectively. Adding isoflurane to the ‘dry’ system caused a considerable reduction in end-tidal CO2. This increased CO2 permeability was first believed to be caused by a relative decrease in dead space due to the isoflurane inlet in front of the reflector. However, the same assembly and flow, but without isoflurane, only caused a 0.2% CO2 reduction, suggesting the mechanism to be a higher affinity to isoflurane and the binding sites in the adsorbing material than CO2. The CO2 reflection of the filters differed when the system was humidified. This was shown by a further reduction of end-tidal CO2 using glass spheres and charcoal filters, but not when using zeolite filters. The binding sites between the two materials must therefore differ in their reaction to H2O.

HME filters have also been suspected to reflect CO2, but this effect has been found to be due solely to the increased internal dead space of the filter.18 In our investigation, the internal dead space was the same for the different types of filters, the increased CO2 being unique to zeolite and charcoal properties as filter materials.

In conclusion, charcoal and zeolite filters with an internal volume of 60 ml had similar reflection properties on volatile anaesthetics, with approximately 70% conservation compared with the glass spheres filter. In addition, they displayed CO2 reflection properties, which were reduced when the air was warmed and humidified to physiological values. The CO2 reflection was further attenuated when isoflurane was added, for a total of a 1% in end-tidal CO2 compared with the glass spheres filter.

Back to Top | Article Outline

Acknowledgements relating to this article

Assistance with the study: we wish to thank the late Associate Professor Olof Werner, MD, PhD, at the Department of Anaesthesiology and Intensive Care, Skane University Hospital Lund for suggesting the reflector idea and for invaluable discussions throughout the work.

Financial support and sponsorship: none.

Conflict of interest: none.

Presentation: none.

Back to Top | Article Outline


1. Vanderbroucke-Grauls CM, Teeuw KB, Ballemans K, et al. Bacterial and viral removal efficiency, heat and moisture exchange properties of four filtration devices. J Hosp Infect 1995; 29:45–56.
2. Luttropp HH, Berntman L. Bacterial filters protect anaesthetic equipment in a low-flow system. Anaesthesia 1993; 48:520–523.
3. Lee MG, Ford JL, Hunt PB, et al. Bacterial retention properties of heat and moisture exchange filters. Br J Anaesth 1992; 69:522–525.
4. Mebius C. Heat and moisture exchangers with bacterial filters: a laboratory evaluation. Acta Anaesthesiol Scand 1992; 36:572–576.
5. Thomasson R, Luttropp HH, Werner O. A reflection filter for isoflurane and other anaesthetic vapours. Eur J Anaesthesiol 1989; 6:89–94.
6. Dahm SL, Steptoe P, Luttropp HH, et al. Charcoal as an airway isoflurane reflection filter. Eur J Anaesthesiol 1998; 15:230–233.
7. Perhag L, Reinstrup P, Thomasson R, et al. The reflector: a new method for saving anaesthetic vapours. Br J Anaesth 2000; 85:482–486.
8. Enlund M, Wiklund L, Lambert H. A new device to reduce the consumption of a halogenated anaesthetic agent. Anaesthesia 2001; 56:429–432.
9. Tempia A, Olivei MC, Calza E, et al. The anesthetic conserving device compared with conventional circle system used under different flow conditions for inhaled anesthesia. Anesth Analg 2003; 96:1056–1061.
10. Sackey PV, Martling CR, Granath F, et al. Prolonged isoflurane sedation of intensive care unit patients with the anesthetic conserving device. Crit Care Med 2004; 32:2241–2246.
11. Sturesson LW, Johansson A, Bodelsson M, et al. Wash-in kinetics for sevoflurane using a disposable delivery system (anaconda) in cardiac surgery patients. Br J Anaesth 2009; 102:470–476.
12. Zeolites Thomasson R. curvature and anesthesia [thesis]. Lund, Sweden: University of Lund; 1991. ISBN 91-628-0486-5.
13. Feigin DS. Misconceptions regarding the pathogenicity of silicas and silicates. J Thorac Imaging 1989; 4:68–80.
14. Simonato L, Baris R, Saracci R, et al. Relation of environmental exposure to erionite fibres to risk of respiratory cancer. IARC Sci Publ 1989; 398–405.
15. Sturesson LW, Malmkvist G, Bodelsson M, et al. Carbon dioxide rebreathing with the anaesthetic conserving device, anaconda. Br J Anaesth 2012; 109:279–283.
16. Bengtson JP, Sonander H, Stenqvist O. Preservation of humidity and heat of respiratory gases during anaesthesia - a laboratory investigation. Acta Anaesthesiol Scand 1987; 31:127–131.
17. Turtle MJ, Ilsley AH, Rutten AJ, et al. An evaluation of six disposable heat and moisture exchangers. Anaesth Intensive Care 1987; 15:317–322.
18. Romano E, Gullo A, Vacri A, Bonifacio R, Caristi D. Effects of a heat and moisture exchanger on carbon dioxide equilibrium during mechanical ventilation with the Bain circuit. Eur J Anaesthesiol 1987; 4:183–186.
© 2015 European Society of Anaesthesiology