186
A promising new, on-line method for determination of oD in water uses a chromium reaction
furnace (Gehre et al. 1996). The reaction unit,
which is directly coupled to the mass spectrometer
inlet, is filled with chromium powder «0.3 mm)
and set to 900°C. Very small sample sizes (1111) are
injected through a layered septum into a preevacuated reaction furnace, and are subject to flash
evaporation and reduction to H2 upon contact with
the hot chromium. The H2 gas flows directly into
the bellows of the sample side in the dual inlet
mode, because of pressure gradients. One advantage of this approach is that no special sample preparation is needed since the hot chromium reduces
water while simultaneously binding other elements
in the sample (C, N, 0, sulfur [S], and halogens) to
thermally stable compounds (Gehre et al. 1996).
The overall precision of this approach is ± 1 %0.
8 18 0 Analysis of Water
For relatively large volumes of water (0.2 to 0.5
ml), the 18 0/ 16 0 composition is determined by
equilibration with CO 2 of known isotopic composition (Dugan et al. 1985; Compston and Epstein
1958). A volume of CO2 adequate for analysis on
the mass spectrometer is placed along with the water sample in a preevacuated tube. After equilibration in a constant temperature water bath for at least
24 hr, the CO2 is extracted cryogenically using liquid nitrogen along with dry ice/ethanol traps to prevent water vapor and other gases from entering the
collection tube. The CO 2 collected will have the
18 0/ 16 0 signature of the water since the oxygen in
the water should completely exchange with the
much smaller mole fraction of oxygen in the CO 2 .
This method is easily batched, but requires large
volumes of water and is time consuming. This
method has a precision better than ± 0.2%0.
Equilibration techniques have also been developed for much smaller sample sizes (40 to 50 Ill)
(Kishima and Sakai 1980; Socki et al. 1992). These
methods add the sample using a micropipet into the
bottom of 9-mm Pyrex tube that was previously
evacuated and backfilled with dry nitrogen or argon. The sample is frozen with liquid nitrogen,
evacuated, and then 250 III of CO2 is injected
through a septa port. The tube is sealed under
vacuum and equilibrated in a constant temperature
water bath for at least 24 hr. After equilibration, the
James R. Ehleringer, John Roden, and Todd E. Dawson
CO 2 is extracted as discussed above and ready for
the mass spectrometer. An external liquid nitrogen
trapping device (similar to a cold finger) has proven
useful for enhancing the mass spectrometer precision for the small volumes of CO2 utilized. A variation on this method can be found in Scrimgeour
(1995).
Even smaller volumes (5 to 10 Ill) of water can
be analyzed by reacting with guanidine hydrochloride to produce CO 2 (Dugan et al. 1985; Wong et
al. 1987). The water and guanadine hydrochloride
are heated in an evacuated sealed tube at 260°C for
16 hr. The two gases formed in this reaction are
ammonia (NH 3 ) and CO 2 . Upon cooling, the CO 2
combines with NH3 to form an ammonium carbamate. The CO 2 is released from the ammonium carbamate by reacting it with phosphoric acid and
trapped at liquid nitrogen temperatures. The
method is most useful when very small sample
sizes are required, otherwise the method is more
time consuming than the H20-C02 equilibration
methods. The precisions of both techniques are
similar, 0.2%0.
Methods for Leaf and Stem
Organic Matter Sampling,
Extraction, and Analysis
Total Tissue Versus Cellulose Analysis
Most isotopic studies on water relations utilize cellulose since it has very slow turnover rates (Lea and
Leegood 1993) and represents the water signature
at the time of fixation. The study of various components of organic matter can yield information regarding the dynamics of isotopic fractionation during metabolism. However, for considerations of
water relations, the differences in oD between various components (e.g., lipids and carbohydrates,
Sternberg et al. 1984) make whole tissue samples
more difficult to interpret. Although the carbonbound H is considered nonexchangeable, postphotosynthetic heterotrophic metabolism has been
shown to modify the oD of plant carbohydrates as
they are exposed to isotopically different water
fractions (Yakir 1992). Even though the 0 of organic matter tends to be less exchangeable than H,
the use of cellulose rather than total tissue is particularly important for 0 18 0 studies due to the com-
A promising new, on-line method for determination of oD in water uses a chromium reaction
furnace (Gehre et al. 1996). The reaction unit,
which is directly coupled to the mass spectrometer
inlet, is filled with chromium powder «0.3 mm)
and set to 900°C. Very small sample sizes (1111) are
injected through a layered septum into a preevacuated reaction furnace, and are subject to flash
evaporation and reduction to H2 upon contact with
the hot chromium. The H2 gas flows directly into
the bellows of the sample side in the dual inlet
mode, because of pressure gradients. One advantage of this approach is that no special sample preparation is needed since the hot chromium reduces
water while simultaneously binding other elements
in the sample (C, N, 0, sulfur [S], and halogens) to
thermally stable compounds (Gehre et al. 1996).
The overall precision of this approach is ± 1 %0.
8 18 0 Analysis of Water
For relatively large volumes of water (0.2 to 0.5
ml), the 18 0/ 16 0 composition is determined by
equilibration with CO 2 of known isotopic composition (Dugan et al. 1985; Compston and Epstein
1958). A volume of CO2 adequate for analysis on
the mass spectrometer is placed along with the water sample in a preevacuated tube. After equilibration in a constant temperature water bath for at least
24 hr, the CO2 is extracted cryogenically using liquid nitrogen along with dry ice/ethanol traps to prevent water vapor and other gases from entering the
collection tube. The CO 2 collected will have the
18 0/ 16 0 signature of the water since the oxygen in
the water should completely exchange with the
much smaller mole fraction of oxygen in the CO 2 .
This method is easily batched, but requires large
volumes of water and is time consuming. This
method has a precision better than ± 0.2%0.
Equilibration techniques have also been developed for much smaller sample sizes (40 to 50 Ill)
(Kishima and Sakai 1980; Socki et al. 1992). These
methods add the sample using a micropipet into the
bottom of 9-mm Pyrex tube that was previously
evacuated and backfilled with dry nitrogen or argon. The sample is frozen with liquid nitrogen,
evacuated, and then 250 III of CO2 is injected
through a septa port. The tube is sealed under
vacuum and equilibrated in a constant temperature
water bath for at least 24 hr. After equilibration, the
James R. Ehleringer, John Roden, and Todd E. Dawson
CO 2 is extracted as discussed above and ready for
the mass spectrometer. An external liquid nitrogen
trapping device (similar to a cold finger) has proven
useful for enhancing the mass spectrometer precision for the small volumes of CO2 utilized. A variation on this method can be found in Scrimgeour
(1995).
Even smaller volumes (5 to 10 Ill) of water can
be analyzed by reacting with guanidine hydrochloride to produce CO 2 (Dugan et al. 1985; Wong et
al. 1987). The water and guanadine hydrochloride
are heated in an evacuated sealed tube at 260°C for
16 hr. The two gases formed in this reaction are
ammonia (NH 3 ) and CO 2 . Upon cooling, the CO 2
combines with NH3 to form an ammonium carbamate. The CO 2 is released from the ammonium carbamate by reacting it with phosphoric acid and
trapped at liquid nitrogen temperatures. The
method is most useful when very small sample
sizes are required, otherwise the method is more
time consuming than the H20-C02 equilibration
methods. The precisions of both techniques are
similar, 0.2%0.
Methods for Leaf and Stem
Organic Matter Sampling,
Extraction, and Analysis
Total Tissue Versus Cellulose Analysis
Most isotopic studies on water relations utilize cellulose since it has very slow turnover rates (Lea and
Leegood 1993) and represents the water signature
at the time of fixation. The study of various components of organic matter can yield information regarding the dynamics of isotopic fractionation during metabolism. However, for considerations of
water relations, the differences in oD between various components (e.g., lipids and carbohydrates,
Sternberg et al. 1984) make whole tissue samples
more difficult to interpret. Although the carbonbound H is considered nonexchangeable, postphotosynthetic heterotrophic metabolism has been
shown to modify the oD of plant carbohydrates as
they are exposed to isotopically different water
fractions (Yakir 1992). Even though the 0 of organic matter tends to be less exchangeable than H,
the use of cellulose rather than total tissue is particularly important for 0 18 0 studies due to the com-