| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
BACKGROUND: Xenon has many characteristics that make it very attractive as an anesthetic and therapeutic drug. Unfortunately, the supply of xenon is fixed, and therefore reclamation and recovery from even the most efficient breathing circuits is desirable. We built and evaluated a cryogenic device to recover xenon from waste anesthetic gases. METHODS: Xenon was selectively frozen to –139.2°C from test gas mixtures at ambient pressure (STP). The machine ran on standard 240 V 13 A electrical current without refrigerants that required replenishing, e.g., liquid nitrogen. A wide range of xenon/oxygen mixtures were processed over a range of freezing chamber temperatures. Efflux gas and thawed reclaimed xenon were collected separately. Xenon purity and yield (fraction recovered) were measured and calculated on each occasion.
RESULTS: Gas was processed at 300 mL/min, and the operating temperature was –139.2 (0.096)°C [Mean (sd)]. Purity and yield were >90% and >70% for gas mixtures containing
CONCLUSIONS: Xenon of high purity (>90%) and yield (>70%) for such a machine was recovered from all gas mixtures containing
Xenon is a noble gas with many attractive anesthetic properties such as favorable hemodynamics, minimal side effects, and fast onset/emergence (1–8). It has been licensed as an anesthetic in Russia (2002) and Europe (2007). Xenon may also be a potent neuroprotectant against hypoxic/ischemic brain injury (9–15). Proposed therapeutic uses include perinatal asphyxia (15), cognitive deficit reduction after cardiac surgery (16,17), and stroke (18). If xenon does find a clinical role as an anesthetic or therapeutic agent, practical methods must be found to minimize xenon consumption per patient episode because of its scarcity (approximate cost US$10 per liter). Xenon is a by-product of industrial oxygen production from the liquefaction of air. Annual global production is therefore relatively fixed at approximately 9–12 million liters. If solely used in medicine, this would still only be approximately 800,000 anesthetics/treatments, even if xenon availability per patient was severely restricted to, for example, 15 L. The notion that the price of xenon will decrease as demand increases cannot be assumed, as liquefying additional air purely to extract xenon would be prohibitively expensive in energy terms. Given this fixed supply, recovery methods could help to conserve xenon. The blood solubility of xenon is much lower than that of other anesthetics (blood/gas solubility ratio 0.0115) and uptake from the lungs is also very low (19). Consequently, although conventional circle anesthesia systems are usually considered to be gas-efficient, when used with xenon most of the xenon is still lost as waste gas. Even with a fresh gas flow (FGF) as low as 500 mL/min, >80% of the supplied xenon is lost as waste and if higher initial flows are used to enhance wash-in; the waste fraction is even larger (20–22). However, the low solubility and uptake properties suggest that closed circuit breathing systems, where FGF matches patient uptake with no "spill," might prove unusually gas-efficient with xenon (19). This is borne out by the closed circuit in vivo xenon consumption data of Luttropp et al. (6.5 L for initial 15 min and 2.5 L/h thereafter) and Ferrari et al. (13 L for initial 30 min then 3.69 L/h thereafter) (23,24). Other groups have also successfully used closed circuit xenon delivery systems (24–29). Therefore, by optimizing breathing systems, this scarce resource can be used both responsibly and at acceptable cost. Clearly, because of limited supply, methods that could produce further efficiency gains are worthy of investigation, and xenon recovery is the next logical step. There are two general approaches to xenon recycling/recovery.
The advantages of cryogenic approaches over adsorption methods include no gas contamination by the process itself and a more easily sterilized gas pathway. Our first objective was to design and construct an automated ambient pressure cryogenic machine with a removable sterilizable freezing chamber that did not rely on cryogenic liquids, which evaporate and need replenishing as the cooling method. The second objective was to experimentally evaluate the design.
Description of Apparatus The final freezing chamber design consisted of an elongated tapered cylinder with proximal inlet and distal outlet ports (Fig. 1). This fitted into a corresponding port in the refrigeration system, permitting easy removal for cleaning/sterilization. This chamber contained copper rings to present a large surface area for xenon to freeze onto and to allow even heat transfer throughout its interior.
The refrigeration unit comprised three stages, one precooling the refrigerant circulating in the next stage until a final working temperature in the freezing chamber of –139.2°C was attained (NBS Cryoresearch Ltd., Tollesbury, UK) (Fig. 2). This was the lowest temperature the machine could achieve after several design modifications, as pilot studies had demonstrated that temperatures well below the freezing point of xenon were needed to maximize yield. The first stage dissipated the heat energy via an air-cooled radiator. Each stage was thermally insulated. The refrigerants were: Stage 1, R404A (150 g); Stage 2, R508B (110 g); and Stage 3, R14 (86 g) + R290 (8 g). The design included an ultrasonic (1 MHz) xenon analyzer ("Minison," Thomas Swan and Co. Ltd., Cambridge, UK) (29,31,32). All gas storage bags were metallic to reduce xenon losses by diffusion (Hans Rudolph, Inc., KS City, MO).
A control unit was built and programmed by one of the authors (JD) (Fig. 3).
Experimental Evaluation
Gas flow rate.
Main Experiments The machine functioned automatically under computer control. Once the chamber had cooled to –139.2°C, the inlet valve opened and a gas pump propelled the gas to be processed through the freezing chamber. The xenon froze, forming xenon "ice," whereas the unfrozen efflux waste gas from the chamber outlet was directed to a gas collection bag. After 5 L of gas had been processed, the pump stopped, the inlet valve closed and the chamber warmed to –100°C, allowing frozen xenon to thaw to gas. The chamber outlet valve changed, to direct the expanding xenon gas to a xenon collection bag. After each run, the volume of gas in the efflux waste and xenon storage bags (V1) and (V2) was measured by emptying each bag in aliquots with a calibrated 200-mL syringe. The purity of recovered xenon was measured using the xenon analyzer. This analyzer has been extensively evaluated and found to be very accurate relative to a "gold standard" laser refractometer method (mean difference of –0.74% and 2 standard deviation limits of agreement of +1.08% to –2.56%) (33). We applied a slight correction factor (multiply ultrasonic reading by 0.9835) derived from this previous work to improve accuracy. The above-mentioned procedure was repeated with a series of five different oxygen/xenon input gas mixtures. Three complete series of measurements were made.
Relationship between efflux xenon concentration and freezing chamber temperature.
Results are shown in Table 1. The mean freezing chamber temperature was –139.2 (±0.1), at an ambient temperature of 19.5 (±1.7)°C mean (sd). It was noted that xenon always started to thaw at a chamber temperature of between –107°C and –108°C, in-keeping with the quoted boiling point of –107°C ± 3°C (32). Purity and yield are shown graphically in (Fig. 4).
The volume of gas processed was measured by summation of the gas volume in the xenon and waste gas collection bags after each run (Table 1, columns 3 and 5). There was a small difference between the calculated total xenon content of the gas being processed and that of the same gas once processed. This could result, for example, from under-measurement of the xenon collection bag or over-measurement of the waste gas bag volumes. The opposite measurement error could cause a slight apparent xenon "gain." Yield calculations could therefore be performed in two different ways. Although there is very little practical difference between the results of either calculation, for the purpose of Figure 4, we have erred on the side of caution and presented the lowest, most conservative yield values. The xenon concentration in the efflux gas was rather independent of the concentration in the input gas (Fig. 5). The freezing chamber temperature influenced the efficiency of recovery, as evidenced by the percentage of xenon present in the efflux gas, which increased as temperature increased (Fig. 6).
This study demonstrated that the device developed for cryogenic recovery of xenon from waste anesthetic gases was able to extract xenon reliably, and that the purity of the extracted xenon exceeded 90%. The percentage unfrozen xenon in the efflux gas is mainly determined by the temperature of the freezing chamber and independent of the input gas xenon fraction. Since the experiments were performed at atmospheric pressure, the expected percentage of unfrozen xenon can be derived from the vapor pressure. At –139.2°C, the machine operated exactly as predicted from published physical chemistry data; however, at temperatures close to its freezing point, the efficiency of xenon recovery appeared to be much better than predicted (Fig. 6) (32). This requires some consideration. The published vapor data in the –139 to –112 temperature range are for xenon gas in equilibrium (by sublimation) with frozen solid xenon. However, before multiple layers of solid xenon can form, gas will first be adsorbed onto the surfaces of the copper rings in the freezing chamber. Xenon adsorbs much more readily to copper than to its own solidified molecules (i.e., xenon "ice"); therefore, the intense cooling we used may not be required if the copper surface area can be increased (34). Also, the xenon content decreases as the gas moves along the chamber. Further xenon can only reach the distal chamber by diffusion from the proximal section (diffusion limited transport). It is clear that we observed the mixed effects of several processes. Potential improvements to the design would include the use of a honeycomb structure within the freezing chamber to increase the copper surface area (for which xenon has a high affinity) and a further reduce operating temperature. The gas processing rate could also be improved. Currently, every 5 L of gas processed takes 53 min, mostly due to the pauses in processing during cooling and thawing. The heat transfer (removal) rate from the gas as it passes through the freezing chamber can be estimated, from the heat capacities of oxygen and xenon, to be approximately 52 J/min. Improving heat transfer properties where the chamber and its outer cooling "sleeve" are in contact would permit both faster gas flows and more rapid freeze/thaw cycling times. For a 70% xenon, 30% O2 inhaled gas mixture (approximately 1 MAC), collection of useful gas for recovery could include: gas flushed from a closed breathing system to counteract N2 build up, xenon-rich gas remaining in the circle and that eliminated from the body in the first few exhaled breaths at the end of an anesthetic. This would likely produce several liters of gas with >20% xenon content, from which our machine should be able to recover >70% of the xenon. For xenon, very low-flow circuits are not as efficient as one might assume and it is clear that xenon recovery in this situation would greatly reduce overall xenon use. For example, Burov et al. (20,35) described a very-low flow protocol in which anesthesia is maintained with FGFs of 300 mL/min xenon and 300 mL/min oxygen. If, during maintenance, we assume typical uptake values of 60 mL/min xenon and 250 mL/min oxygen (metabolic), this would still generate an overspill or wastage rate of 240 mL/min (14.4 L/h) xenon (i.e., 80% of the xenon FGF) and 3 L/h oxygen. A closed-circuit breathing system would be a better choice of breathing system as it would minimize the volume of gas to be processed (36). From these closed-circuit breathing systems, the waste gas collected would mainly be produced by denitrogenation circle "flushes" with fresh gas, residual xenon purged from the circle, and the initial exhaled gas collected during emergence from anesthesia. This gas would again be rich in xenon. By extrapolation of data from Luttropp et al. and Ferrari et al. (in which a denitrogenation "flush" was included) a 2-h closed circle anesthetic would consume 10.9 L ($109) and 18.6 L ($186) of xenon, respectively (23,24). Furthermore, by returning all "waste" gas collected in this way to the xenon manufacturer, Ferrari et al. (24) managed to recover 60% of the total xenon used. The cost of industrially "reprocessed" xenon has been estimated at $3 per liter, and thus if 60% could be recycled, this would reduce the above estimates for a 2-h anesthetic from $109–$186 to $63–$108. It is likely, therefore, that even with closed-circuit breathing systems, a recovery device could still usefully reduce overall xenon consumption per anesthetic, a 50% reduction perhaps being a realistic target. By combining xenon recovery with closed circuit use, it is conceivable that a 2-h administration for anesthesia or neuroprotection could be achieved using less than 10 L xenon overall. There are a number of issues that would have to be addressed before a cryogenic machine of this type could be used in a clinical setting. Waste gas from a circle system might be very humid, containing carbon dioxide and possibly volatile anesthetic vapors. These could be removed by removing gas for processing from a point distal to the circle soda lime canister to ensure carbon dioxide removal, by using a dehumidification chamber containing silica gel beads for example or by using a volatile vapor adsorbent, such as activated charcoal, as seen in the Physioflex closed circuit machine (37). Alternatively, a "precooling" chamber at approximately –50°C could selectively freeze water vapor, volatiles, and any residual carbon dioxide. From a legal perspective, the recovery process could be considered the manufacture of a medical gas and the operator to be a medical gas manufacturer, responsible for its medical purity (typically 99.9% or above) and freedom from microbial contamination. If recovered xenon were only to be reused during the same patient anesthetic then, so long as the gas pathways could be sterilized between patients, the sterility of the gas itself would be less of an issue than if recovered xenon were intended for other patients. Although the removable chamber can be sterilized more readily than, for example, membrane or zeolite-based devices, sterility concerns over the gas itself mean that, realistically, recycling would need to be restricted to the same patient episode only. It is more likely that such a device might form part of a theater complex central xenon recovery facility, from which the relatively pure xenon could be returned at intervals to a gas manufacturer for final purification, sterilization, and recertification as a medical gas. In conclusion, we have demonstrated that an automated xenon recovery machine running on 240 V AC, 13 A standard electrical current with autoclavable components is technically feasible. It functioned, as intended, without the use of refrigerants (e.g., liquid nitrogen) that require replenishment. Within-patient xenon recycling and reuse might prove feasible, but concerns over purity, gas sterility, and legal issues preclude interpatient xenon reuse. A machine such as this might be best suited for use in a central xenon recovery facility. Even with the most efficient breathing systems, this recovery method could still usefully reduce overall xenon consumption. Such unusual attention to detail may be necessary if xenon finds a clinical role as a neuroprotective drug, since demand could easily outstrip a fixed supply. Regardless of the purchase price, medical xenon will always be a scarce commodity. Ten liters of xenon per 2 h delivery period may be a realistic overall target using closed circuit breathing systems in combination with a xenon recovery device of this type.
The authors thank Dr. Alexei Moozyckine, Swansea, UK, for technical assistance; Mr. John Minister, NBS Cryoresearch, Tollesbury, UK, for advice and construction of part of the machine; Dr. Per Blom and Wolfgang Shmehl of Linde Gas therapeutics, Lidingo, Sweden for donation of xenon gas.
Accepted for publication June 4, 2007. Supported by Department of Health, London, UK; New and Emerging Applications of Technology (NEAT) funded project AO94. Reprints will not be available from the author.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|