Anal. Chem. 2006, 78, 7284-7291
Measurement of Ammonia in Human Breath with a Liquid-Film Conductivity Sensor Kei Toda,† Jianzhong Li, and Purnendu K. Dasgupta*
Department of Chemistry and Biochemistry, Texas Tech University, Lubbock, Texas 79409-1061
Measurement of breath NH3 is of interest in clinical applications as it can be used as a measure of kidney/ liver functions as well as halitosis. We have developed a liquid-film conductivity sensor to measure NH3 in human breath. A film of dilute H2SO4 is formed on the top of two metal capillary tubes placed in a concentric annular arrangement. The tube exterior has been specially treated to render it hydrophilic. As breath passes over the sensor tip, the film collects NH3 and the solution conductivity (measured by the concentric capillaries functioning as electrodes) decreases accordingly. This initial rate of conductivity decrease was determined to be the best metric (most rapid and least dependent on breath pCO2) for ammonia, relative to time to attain complete neutralization (conductivity minimum) or the final rate of conductivity increase as more ammonia dissolves after neutralization. The absorbing solution composition was optimized so that CO2 does not interfere. Both dynamic measurement using mask sampling and offline balloon sampling were performed. Ammonia readily absorbs on surfaces when significant concentrations of water vapor are present. As such, memory effects are common when analyzing human breath for ammonia. This problem was successfully eliminated. The results from this sensor agreed well with data obtained by a solution-phase fluorometric technique using a porous membrane diffusion scrubber and o-phthalaldehyde derivatization chemistry. For breath CO2 measurement, the applicability of a similar sensor that relies on a NaOH film was also demonstrated.
aspartate-N, originating either from amino acid-N or NH3-N. The relative importance of these two routes depends on the nutritional state.1 Patients who have impaired liver or kidney functions have elevated levels of NH3 in exhaled breath.2,3 Other disease conditions can elevate NH3 levels in breath. Helicobacter pylori infection can significantly elevate breath NH3 levels by breaking down urea.4 Ammonia can also be generated in the oral cavity from infections, and exhaled breath NH3 measurement can thus be used to assess halitosis.5 Selected ion flow tube mass spectrometry is particularly popular in measuring a variety of breath gases.6-10 Such systems can provide simultaneous data for various breath gases, including NH3, and thus constitute good research tools. The cost and complexity, however, deter routine use for dedicated breath NH3 measurement. Membrane extraction with a sorbent interference coupled with gas chromatography is another recently introduced technique for general use in measuring breath gases.11 There are a large number of chronic dialysis patients who must undergo dialysis several hours a day, several days a week. The reduction in blood urea nitrogen (BUN) and creatinine levels during the course of dialysis is a good indicator of how well waste products are removed from the bloodstream. However, this cannot be practiced in a continuous manner because of the invasive nature of such blood-based measurements. Patel et al. hypothesized that BUN levels should be related to breath ammonia levels and introduced a noninvasive CO2 laser-based approach to measure breath ammonia by photoacoustic spectrometry.12,13 They clearly established that breath NH3 levels also decrease during hemodialysis and can be used as surrogate measures for BUN and
Ammonia is the major basic gas in a variety of important sample matrixes, for example, the ambient atmosphere, indoor air, and human breath. Comparatively, breath contains high levels of NH3 (from tens of parts per billion by volume (ppbv) in the breath of a healthy individual to 1 ppmv or higher in individuals with renal failure). There are many different gases in exhaled breath. The last portion of deeply exhaled breath, representing alveolar air, can essentially be considered the headspace gas of blood circulating in the body. In humans, ammonia is converted to urea in the liver, and urea passes into urine via the kidney. Hepatic NH3 detoxification by ureagenesis requires an input of
(1) Weijs, P. J.; Calder, A. G.; Milne, E.; Lobley, G. E. 1996, Br. J. Nutr. 76, 491-499. (2) Shimamoto, C.; Hirata, I.; Katsu, K. Hepato-Gastoenterology 2000, 47, 443445. (3) DuBois, S.; Eng, S.; Bhattacharya, R..; Rulyak, S.; Hubbard, T.; Putnam, D.; Kearney, D. J. Dig. Dis. Sci. 2005, 50, 1780-1784. (4) Kearney, D. J.; Hubbard, T.; Putnam, D. Dig. Dis. Sci. 2002, 47, 25232530. (5) Amano, A.; Yoshida, Y.; Oho, T.; Koga, T. Oral Surg. Oral Med. Oral Pathol. 2002, 94, 692-696. (6) Smith, D.; Sˇ paneˇl, P. Rapid Commun. Mass Spectrom. 1996, 10, 11831198. (7) Davies, S.; Sˇ paneˇl, P.; Smith, D. Kidney Int. 1997, 52, 223-228. (8) Sˇ paneˇl, P.; Davies, S.; Smith, D. Rapid Commun. Mass Spectrom. 1998, 12, 763-766. (9) Smith, D.; Sˇ paneˇl, P.; Davies, S. J. Appl. Physiol. 1999, 87, 1584-1588. (10) Smith D.; Wang, T.; Sˇ paneˇl, P. Physiol. Meas. 2002, 23, 477-489. (11) Lord, H.; Yu, Y.; Segal, A.; Pawliszyn, J. Anal. Chem. 2002, 74, 5650-5657. (12) Patel, C. K. N. Gases Technol. 2002, (May/June), 24-30. (13) Pushkarsky, M. B.; Webber, M. E.; Baghdassarian, O.; Narasimhan, L. R.; Patel, C. K. N. Appl. Phys. B 2002, 75, 391-396.
* Corresponding author. Present address: Department of Chemistry and Biochemistry, University of Texas at Arlington, Arlington, TX 76019-0065. E-mail:
[email protected]. † Permanent address: Department of Science, Kumamoto University, Kurokami, Kumamoto 860-8555, Japan.
7284 Analytical Chemistry, Vol. 78, No. 20, October 15, 2006
10.1021/ac060965m CCC: $33.50
© 2006 American Chemical Society Published on Web 09/15/2006
creatinine. For the first time, this provided a noninvasive indicator of dialysis progress and thus a decision-making tool as to when a patient can be taken off the dialyzer.14 The laser photoacoustic spectrometry technique is very sensitive15 and can easily be used for the measurement of ambient levels of NH3 (rarely greater than 10 ppbv, often sub-ppbv). The basic technique has been commercialized for the measurement of ammonia in both exhaled breath and ambient air (www.pranalytica.com). We believe that breath NH3 measurement could be practiced more widely if a more affordable and compact sensing system was available. We present here a compact conductivity-based system for such measurement. It relies on titrating a thin reproducible H2SO4 film with gaseous ammonia present in breath and following the progress of the titration by conductometry. We show the applicability of a similar sensing principle for measuring CO2 in breath; CO2 data are often used for normalizing other breath gas results.16,17 Normally, the applicability of thin liquidfilm sensors would be limited for sensing gases because variabilities in the sample relative humidity (RH) and thus variable extent of film evaporation will create problems with a concentration-sensitive detection scheme. The presently proposed approach is uniquely applicable to situations where sample moisture content is essentially constant, as in breath. It is also applicable to a breath sample that has been passed through a dehumidifier that can produce a constant exit RH stream without removing the analyte. PRINCIPLES Ammonia is the only significant basic gas in breath. The sensor film containing a strong acid is titrated by ammonia in the breath sample resulting in the replacement of much higher mobility H+ (λH+ 349.8 S cm-2 equiv-1) with NH4+ (λNH4+ 73.5 S cm-2 equiv-1). The conductivity will thus decrease linearly with time and be a measure of NH3 concentration. When the free acidity is fully neutralized, a minimum in conductivity is reached; the time to reach this minimum, tnutr, is reciprocally related to the NH3 concentration. Further incorporation of NH3 leads to the formation of NH4+ and OH- and an increase in conductivity; this rate of increase is also related to the concentration. Variabilities in the extent of evaporation of the film can complicate data interpretation; as such, samples and calibrants must be at the same relative humidity. The following quantitative considerations apply. Consider an annular electrode sensor with a spherical capshaped liquid film (radius r mm, height h mm, volume υ µL) located centrally within a larger conduit of radius R mm. Geometrically
υ ) 1/3πh2(3r - h)
(1)
Consider that breath gases flow at a rate of n mol/s through the larger conduit and directly contact the sensor head on. As a first approximation, we assume that NH3 uptake is quantitative for the intercepted portion of the stream; i.e., if CNH3 ppbv ammonia is (14) Narasimhan, L. R.; Goodman, W.; Patel, C. K. N. Proc. Natl. Acad. Sci. U.S.A. 2001, 10, 4617-4621. (15) Pushkarsky, M. B.; Webber, M. E.; Patel, C. K. N. Appl. Phys. B 2003, 77, 381-385. (16) Schubert, J. K.; Spittler, K. H.; Braun, G.; Geiger, K.; Guttmann, J. J. Appl. Physiol. 2001, 90, 486-492. (17) Roller, C.; Namjou, K.; Jeffers, J. Optics Lett. 2002, 27, 107-109.
present in the gas, the rate of NH3 input to the film, I, will be given by
I (nmol/s) ) nCNH3r 2/R2
(2)
In practice, as ammonia is a small MW gas of high diffusivity, the actual rate of collection will be higher than the directly intercepted flux, as some will also be collected via diffusive transport. However, this basically means that the effective value of R (Reff) is lower than its physically measured value. The rate of NH4+ incorporation in the film is thus
d(NH4+)/dt (mM/s) ) I/υ
(3)
The liquid film contains very dilute (x mM, x typically 100× higher than ambient air. Compared to the ambient atmosphere, significant concentrations of strong acid gases do not exist in breath. Weak acid gases, e.g., H2S, mercaptans, CH3COOH, and NOx, cannot affect the conductivity of a 200 µM H2SO4 solution. Of course, the present technique cannot distinguish between volatile amines and ammonia. Although detectable levels of amines in breath have been identified in cases of renal failure, these levels tend to be much smaller than the NH3 levels.31 Direct Continuous Measurement. Exhaled breath was sampled at a constant rate from the mask outlet as the subject (31) Lin, Y.-S.; Wu, T.-Z.; Fang, T.-C. Chem. Sens. 2001, 7 (Suppl. B), 200-202.
7290
Analytical Chemistry, Vol. 78, No. 20, October 15, 2006
was sitting, reading, or working on a computer. In some cases, the exhaled breath was concurrently measured with the DSF method for validation. One set of 6 h of comparative data obtained is shown in Figure 8. Note that the relative phases of the two signals are slightly different: The DSF method has a greater time lag, an intrinsic 5+ min integration, as well as a memory effect that it displays under high moisture conditions. Nevertheless, the data obtained by these two very disparate methods, relying on altogether different principles, show good agreement. The chemistry used in the DSF system is essentially specific for NH3. The liquid film conductometric sensor uses no liquid pump and is intrinsically simple and robust. The excellent agreement shows that the latter is indeed applicable for the measurement of breath NH3. Figure 8 (circles/gray traces), shows exhaled breath NH3 levels for a subject increasing from 300 to 400 ppbv to ∼900 ppbv ∼2 h after lunch containing a significant amount of protein. It is interesting to note the lack of such increase in breath ammonia after a nonprotein lunch the next day for the same subject (dashed lines). Offline Balloon Sampling. While continuous monitoring is appropriate in situations such as with a patient under dialysis, a grab sample of alveolar air is sufficient in other cases. A Peltier-cooled condenser was first investigated where the condensed moisture present in breath was used for trapping the ammonia present. The use of a large sampling loop was also investigated. Overall, we found balloon sampling to be the most convenient; a large sample volume allowed checks of withinsample instrumental variations and smoothed out breath-to-breath variability. However, the memory effect of the balloon itself under high moisture conditions, even with maintaining it at 50+ °C, became an issue. Most drying agents are acidic or otherwise absorb ammonia under high moisture conditions. After investigation of many dehumidifying agents, we found that NaOH was very effective in removing significant amounts of moisture
without removing NH3. At room temperature, breath gases are supersaturated with moisture. With the NaOH-based dehumidifier, the effluent RH was a constant ∼50% and the size of the cartridge used permitted as many samples as can be processed during one 8-h working day before replacement was needed. The passage through the NaOH cartridge removed about half of the breath CO2. Illustrative data for a number of balloon-collected actual breath samples with 2-min sample cycles are shown in Figure S9 in SI. A comparison of balloon versus mask sampling and the effect of mouth rinsing are also provided in SI. In summary, we have demonstrated a simple affordable sensor for the measurement of breath NH3. This can be used in a number of modalities.
ACKNOWLEDGMENT The original concept of conductometric sensing of ammonia for atmospheric measurements was developed under the aegis of U.S. Environmental Protection Agency STAR Grant RD 831074010. However, this article has not been reviewed by the Agency, and no endorsement should be inferred. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review May 25, 2006. Accepted August 18, 2006. AC060965M
Analytical Chemistry, Vol. 78, No. 20, October 15, 2006
7291
