ELM Scientific |
   
| Article: HPLC Variability and Temperature FluctuationsPublished in "What's New in Scientific and Laboratory Technology Magazine"Date: August / September Edition 2000Temperature changes outside often mean fluctuating climates in the laboratory. Even minor variations in room temperature can cause real problems with column temperatures and therefore analytical results. The variability in analytical results includes retention time drift, and therefore in peak areas and retention times. Retention Time Drift One of the major causes of Retention Time (Rt) drift in standard HPLC runs is Column temperature. This effect has been noted by official bodies such as the FDA1 and TGA2. These have approached the problem from a regulatory perspective. The FDA has stated that “ Although many HPLC separations can be carried out at ambient temperature, column operation in a thermostatted column oven is necessary for reproducible, quantitative results, because distribution coefficients and solubilities are temperature dependent”. If column temperatures can be stabilized, retention time drift can be greatly reduced. Peak Response Drift The TGA has formerly specified the allowable Rt (1% rsd) and peak response drift (3% rsd) without attributing cause. They are now concerned with drift effects over a total run i.e. K prime, Tf, Rt etc. for every injection. Various researchers have postulated not only external causes for this drift, but also internal ones. One reported possibility is that under the high pressures used in HPLC, local temperature fluctuations can be caused by solvent friction effects and these fluctuations directly affect the adsorption- desorption process. Others have noted that (empirically) a Mobile Phase (MP) containing components which suffer a high degree of volume change with temperature (e.g. acetic acid) are more likely to suffer from retention time and response drift. Regulatory Impact It should be noted that the bland statement in both the BP / Ph Eur that “The temperature of the chromatographic column is kept constant” has the force of law for all Australian and European pharmaceutical testing laboratories. The Mutual Recognition agreement between Australia and the European Community ensures that all such Australian companies are audited to the appropriate European standards. But what is “constant”? Unless otherwise specified in a specific monograph the BP and Ph. Eur. perform all sample analyses at laboratory temperatures of 15°C and 25 °C (ref 5). With a similar proviso the USP states that “columns are used at ambient temperature” (ref 6). If it is accepted that laboratories are a controlled environment, then the USP gives further guidance by defining “controlled room temperature” as “a temperature maintained thermostatically that encompasses the usual and customary environment of 20°C and 25 °C “. Running HPLC analyses at temperatures outside those specified requires the full re-validation of the method (including specificity, sensitivity, detection limit, ruggedness and robustness) as temperature changes affect distribution co-efficients. Published papers on these temperature effects detail the benefits of specific temperatures for specific analyses. One paper (ref) compared the results at 18, 22 and 30 degrees Celsius and concluded that 18 oC offered specific “advantages in resolution between critical compound pairs” and “additional interfering compounds…………can be separated “. Clearly altering methods to use temperatures other than those they were validated for is a problematic operation requiring a clear understanding of the practical as well as regulatory requirements. Practical Effects The effects qualitatively noted in the MLS Applications Laboratory and laboratories associated with MLS have been somewhat more mundane, but have wide-ranging implications. Diurnal Laboratory Temperature Changes The attached traces clearly show the effect of a laboratory’s temperature control being “tuned down” after hours. A simple Triglycerides (BP) analysis, using water mobile phase, was running when the laboratory temperature was on it’s daytime normal of 23oC (diag. 1). The laboratory’s heating was turned down overnight and the temperature decreased to 19°C-20 °C. Virtually every peak in this analysis (diag. 2) shifted and was mislabelled. System Suitability Summary Results
Air Con OFF < °C = 26 Heating ON < °C = 26 Heating OFF < °C = 18 Diagram 1 
Ordinary air conditioning is not just inadequate for the task of column temperature control, it also contributes to the problems experienced during an analytical run. Column heaters and ovens can heat but not cool. Without proper cooling, column temperature can fluctuate with ambient temperature fluctuations above that set on the oven, thereby creating instability in Rt and Response. The BP/Ph Eur, USP and numerous other bodies and official organizations have published most methods as being run under ‘ambient’ temperature. The “common ground” of their defined laboratory ‘ ambient’ temperatures is a temperature of between 20°C and 25°C. In instances where the Column temperature is outside this range eg. - In direct Sunlight in a well controlled laboratory - When air conditioning is turned down or off after hours - Where the column is in the path of the air conditioner air stream. Column heating would have to be set above the ambient range to control the column temperature. This would require method re-validation, could potentially cause damage to heat labile compounds and could reduce resolution between compounds in a multi component system. THE SOLUTION The solution is to control the column temperature to a preset ambient temperature. This can be achieved by the use of heater / cooler optioned devices, designed to maintain rigid control of column temperature. Given that Rt changes by between 1-3% for a 1°C change in temperature (ref John W Dolan) and keeping in mind the drift specifications for HPLC response and Rt set by regulatory bodies, it must be recommended that the device used have a tight temperature control. Precision and accuracy together should be +1°C or better. Control of Column temperature is critical to those laboratories requiring consistent, reproducible Rt and response. Heater-Cooler modules should be used for all HPLC methods. If these units use Peltier cooling with air convection heat exchange, they are more convenient than water-cooled systems as no extra hardware is required. Column temperature control at ‘ambient’ temperatures as well as at those specified in specific methods/monographs is now possible. Current microprocessor controlled heater/cooler units are capable of both elevated temperature and ‘ambient’ temperature control. |
|
| Affiliates:  
 |
| Copyright © ELM 2008 |
