2029
Anal. Chem. 1986, 58, 2029-2032 (13) Onuska, F. 1.; Kominar, R . J.; Terry, K. J . Chromatogr. Sci. 1983, 21, 512-518. (14) Watanabe, C.; Tomlta, H. Sato, K.; Masada, Y.; Hashlmoto, K. HRC CC, J . High Resolut. Chromatogr. Chromatogr. Commun. 1982, 5 , 10363- 10364. (15) American Chemlcal Society Committee on Environmental Improvement Anal. Chem. 1980, 5 2 , 2242-2248. (16) American Chemical Society Commktee on Environmental Improvement Anal. Chem. 1989, 5 5 , 2210-2218. (17) Method 608, Fed. Regist. 1984, 49(209), 90. (18) Wells, D. E.: Cowan, A. A. J . Chromatogr. 1983, 279, 209-218. (19) Glaser, J. A.; Foerst, D. L.: McKee, G. D.; Quave, S. A,; Budde, W. L. Envlron. Sci. Techno/. 1981, 75, 1426-1435.
(20) Alford-Stevens, A. L.; Bellar, T. A,; Eichelberger, J. W.; Budde, W. L. Anal. &am.. preceding paper In this Issue. (21) Elchelberger, J. W.; Kerns, E. H.; Olynyk, P.; Budde, W. L. Anal. Chem. 1983, 55. 1471-1479.
RECEIVED for review November 21,1985.
Accepted April 14, 1986. This article has not been subjected to review by the U.S.Environmental Protection Agency. Therefore, it does not necessarily reflect the view of the Agency, and no official endorsement should be inferred.
Chemistry of Hydrogen Gas Preparation by Pyrolysis for the Measurement of Isotope Ratios in Hydrocarbons Zvi Sofer Cities Service Oil and Gas Corporation, P.O.Box 3908, Tulsa, Oklahoma 74102
The conventlonal oxldatlon/reduction preparatlon method for the determlnatlon of hydrogen lsotope ratios In hydrocarbons suffers from being very tlme-consumlng. A fast method In whlch the organic hydrogen Is dlrectly converted to hydrogen gas is highly deslrable. This report discusses the chemlstry Involved In adaptlng high-temperature pyrolyds reactlons for the dlrect conversion of organk hydrogen to hydrogen gas for lsotope analysts. The report shows that hlgh-temperature pyrolysis can be adapted to hydrogen isotope determlnatlons provlded that the organic matter contains only carbon and hydrogen atoms and that the pyrolysis technique Is highly standardized.
In most laboratories that measure hydrogen isotope ratios in organic matter, the preparation of hydrogen gas involves the combustion (oxidation) of the organic matter to carbon dioxide and water. Water is subsequently separated cryogenically from the COz in a vacuum line and passed over hot uranium or zinc shavings, resulting in the quantitative conversion (reduction) of the water to hydrogen gas. Over the years two oxidation techniques have been developed: the dynamic method, which was described by Craig (I), and the static method, which was described by Frazer (2)and Sofer (3).In the dynamic method, the organic matter is combusted in a vacuum line in an excess oxygen atmosphere. The resulting water is then passed over hot uranium to obtain hydrogen gas according to the method described by Friedman (4). The main disadvantage of this method is its extreme slowness (about 30 min per sample). In the static methods, the organic matter is oxidized by copper oxide in evacuated and sealed quartz (2)or borosilicate (3)tubes. After combustion, tubes are cracked open in a vacuum line and COz is cryogenically separated from water; the latter is then reduced to hydrogen gas in a manner similar to that of the dynamic method. The static combustion method has the advantage of being fast (about 5 min per sample); however, the cryogenic separation and the conversion of water to hydrogen gas are slow and no substantial time is saved. With respect to deuterium determination both the static and dynamic methods also suffer from inherent problems that are associated with the difficulty of quantitatively moving water vapor in a vacuum system. These problems often result in an incomplete 0003-2700/86/0358-2029$01.50/0
conversion of water, isotopically fractionated hydrogen gas, and memory effects. In order to eliminate these problems, a fast method, which involves the direct conversion of the organic hydrogen to hydrogen gas (thereby eliminating the need to handle water vapor in a vacuum system), is highly desirable.
DISCUSSION Hydrogen gas can be directly generated from organic matter by pyrolysis a t elevated temperatures (>go0 "C) according to the following two reactions:
CH,
n -CH4
+ ( 1 - :)C CHd + C + 2Hz
+
4
(2) The equilibrium constant K = PH;/PCH, (where P H 2 and PcH, are the respective partial pressures of hydrogen and methane) for reaction 2 as a function of temperature is shown in Figure 1 (after Gulbransen and Andrew, ref 5). As can be seen, at elevated temperatures (>lo00 "C) K > 100; i.e., the formation of hydrogen is highly favored over methane. The isotopic composition of the generated hydrogen can be directly measured on this gas mixture. The presence of some residual methane would not affect the measurement of HD/H2 ratios in the mass spectrometer because of the large difference in the mass of hydrogen vs. methane. Because eq 2 is a second-order reaction, at equilibrium the hydrogen mole fraction X,i.e., the fraction of the total hydrogen atoms converted into hydrogen gas, is dependent on the partial pressure of hydrogen (or sample size) and the equilibrium constant (i.e., temperature) 'Hz
X= pHz
+ 2pCH,
--
1 1 + (2pHz/K)
(3)
Consequently, if the isotopic fractionation as defined below 6DHz 1 a = (4) 6DCHl + 1
+
between the hydrogen gas and the methane is different than unity, the isotopic composition of the hydrogen gas will also depend on the temperature and the hydrogen partial pressure. Assuming that the organic matter contains only CH, groups (i.e., no heteroatoms that might give rise to HzO, H2S, NH3, 0 1986 American Chemical Society
2030
ANALYTICAL CHEMISTRY, VOL. 58, NO. 9, AUGUST 1986
Table I. Calculated 8D Values at 1160 OC (Equilibrium Constant K = 260 atm) weight, mg
total pressure, atm
H2pressure, atm
8D,% (8Do 0.0%)
CH,
CHI
1150 "C
22 OC
115OOC
22 OC
a = 1.25
0.1 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0
0.06 0.57 1.14 1.71 2.29 2.86 3.43 4.00 4.57 5.14 5.71 6.29 6.86 7.43
0.16 1.51 3.00 4.47 5.93 7.38 8.82 12.23 11.65 13.04 14.43 15.82 17.18 18.54
0.03 0.31 0.62 0.93 1.23 1.53 1.83 2.12 2.42 2.70 2.99 3.28 3.56 3.84
0.15 1.50 1.96 4.40 5.80 7.18 8.53 9.86 11.17 12.45 13.71 14.95 16.18 17.38
0.03 0.31 0.61 0.91 1.20 1.49 1.77 2.04 2.32 2.58 2.84 3.10 3.36 3.60
0.23 2.28 4.48 6.59 8.62 10.58 12.47 14.30 16.07 17.79 19.45 21.07 22.63 24.15
cy=
1.20
0.19 1.90 3.73 5.48 7.17 8.80 10.37 11.89 13.36 14.78 16.16 17.49 18.79 20.05 -
a = 1.15
cy=
1.10
a = 1.05
0.11 1.04 2.09 2.98 3.90 4.78 5.63 6.45 7.24 8.01 8.75 9.47 10.16 10.84
0.15 1.49 2.92 4.29 5.61 6.88 8.10 9.27 10.42 11.53 12.60 13.64 14.64 15.62
0.06 0.54 1.06 1.56 2.04 2.50 2.94 3.37 3.78 4.18 4.56 4.94 5.30 5.65
Table 11. Calculated 8D Values at 1000 "C (Equilibrium Constant K = 100 atm) weight, mg
total pressure, atm
Hzpressure, atm
CH2
CH,
1000 OC
22 "C
1000°C
0.1 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0
0.06 0.57 1.14 1.71 2.29 2.86 3.43 4.00 4.57 5.14 5.71 6.29 6.86 7.43
0.15 1.34 2.65 3.92 5.18 6.41 7.63 8.83 10.00 11.17 12.33 13.47 14.59 15.71
0.03 0.31 0.61 0.91 1.20 1.49 1.77 2.05 2.31 2.59 2.86 3.12 3.38 3.64
0.14 1.32 2.58 3.78 4.94 6.05 7.12 8.16 9.17 10.14 11.10 12.02 12.93 13.81
22
O C
0.03 0.31 0.60 0.88 1.15 1.40 1.65 1.89 2.13 2.35 2.57 2.78 3.00 3.20
8D,% (&Do= 0.0%) a = 1.25
0.54 5.17 9.91 14.26 18.30 22.06 25.57 28.87 31.98 34.91 37.67 40.36 42.85 45.25
cy
= 1.20 0.45 4.31 8.24 11.86 15.20 18.31 21.22 23.94 26.50 28.92 31.21 33.39 35.45 37.42
a = 1.15
a = 1.10
0.35 3.37 6.44 9.26 11.86 14.28 16.53 18.64 20.62 22.49 24.26 25.94 27.53 29.05
0.25 2.35 4.48 6.43 8.42 9.91 11.46 12.92 14.29 15.59 16.79 17.94 19.03 20.07
cy
= 1.05 0.13 1.23 2.34 3.36 4.30 5.16 5.97 6.73 7.43 8.10 8.72 9.32 9.88 10.41
etc., are present), the true isotopic composition of the organic matter (6Do)and the isotopic composition of the hydrogen generated during pyrolysis (6DH,)are related according to the following mass balance equation:
(5) where Y is the mole fraction of hydrogen tied to CHI after pyrolysis. Since X + Y = 1, X = 1/[1+ (wH,/K)] and bDCH, = ( ~ D H+, l ) / a ]- 1,it can be shown by arranging eq 4 that
or
a(6Do
+ 1)
-1
(7)
where P, the total pressure, is defined as P = PCH,+ PH,. Based on eq 6 and 7, Tables I and Il give the results of some theoretical calculations for 6Dg at two different temperatures and for five different fractionation factors (a)as a function of total pressure and hydrogen partial pressure. These calculations assume a 6Do value of 0.0% for the true isotopic composition of the organic matter. Calculated data for 6DH2 vs. total pressure or hydrogen partial pressure as would be measured after cooling to 22 "C are shown in Figures 2 and 3. It was assumed here that cooling of samples from pyrolysis temperature to room temperature is fast and no noticeable isotopic or chemical reequilibration occurs during that period of time. As can be seen, at 1150 O C there is nearly a linear relation between 6DH2and total pressure or hydrogen partial pressure over the range of 0.0-3.8 atm (or 0.0-18.5 atm if
150
950
I IS0
1350
1550
O K
Flgwe 1. EqulHbrium constant (K)vs. temperature (K) for the reaction CH,
== C + 2H,.
ANALYTICAL CHEMISTRY, VOL. 58, NO. 9, AUGUST 1986
0.0
3.0
2.0
1.0
Y.0
0.0
1.0
2.0
3.0
4.0
3.0
v.0
2031
PIAtrn.1
P1Atm.l
Calculated 6D vs. total pressure at 22
OC.
0
I
"
"
I
"
"
1
"
"
'
-
1 15OoC 3 0 -
. 0
-
*mg CH,
-
CY-
D -
N
0.0
1.0
2.0
3.0
P, (Atm I
Y.0
0.0
1.0
2.0
PH7[Atm.l
Calculated 6D vs. H2 partial pressure at 22 O C . pressure was measured at 1150 "C). The above calculations were, in effect, based on a hypothetical hydrocarbon with a chemical composition of CH,; Le., it was assumed that 2CHz CHI + C (or, for example, 14 mg of hydrocarbons would generate 8 mg of methane) and that hypotheltical amounts of 0.1-13 mg of hydrocarbons were reacted. In order to calculate total pressure or hydrogen partial pressure, the volume of a hypothetical reaction chamber was set at 5.5 cm3 (this is the volume in which the pyrolysis is performed in the laboratory). Hydrocarbon sample weights (in mg) are shown in Figures 2 and 3 as numbers with asterisks and the corresponding total pressure or hydrogen
-
partial pressure is shown on the X axis. In general, however, because real hydrocarbons might have H/C atom ratios different than 2, the 6DH, vs. P or PH*plot is more universal than the 6DH, vs. mg of hydrocarbon plot. Accordingly, plots of 6D vs. mg of hydrocarbons were not generated. As was mentioned previously, the chemistry of reaction 2 and the isotopic mass balance (eq 5) dictate that in order to correlate 6DH, with totalpressure or hydrogen partial pressure, the only two compounds into which hydrogen atoms are allowed are methane and hydrogen gas (i.e., no heteroatoms in the organic matter). This limits the applicability of this method to organic compounds like methane, ethane, benzene,
2032
ANALYTICAL CHEMISTRY, VOL. 58, NO. 9, AUGUST 1986
and the heavier aliphdtic and aromatic hydrocarbon fractions of crude oils.
CONCLUSIONS On the basis of the above discussion it can be concluded that the pyrolysis method can be adopted for the determination of hydrogen isotope compositions of hydrocarbons that contain only carbon and hydrogen atoms, provided that a standardized procedure that corrects for variations in total pressure or hydrogen partial pressure is followed. Ideally, the pressure correction would give the true isotopic composition 6Do at P or PH2= 0. In practice, however, this is not necessary if some of the mass spectrometer readings (i.e., the difference between sample machine ratios and reference machine ratios) are first calibrated for the effect of P or PH2and then all the mass spectrometer readings are corrected to a standard (though arbitrary) pressure. [Machine ratio is defined as the ratio of the electrical current generated by the minor ion beam (i.e., positive ions with molecular mass 3) over that of the major ion beam (i.e., positive ions with molecular mass 2).] Once the correction for pressure has been applied to the mass spectrometer readings, the numbers can be converted into the 6Do values by using two hydrocarbon standards that have different but known &Dovalues and that have been prepared, measured, and corrected in the same manner (Le., the same pyrolysis temperature and corrected to the same pressure) as the unknown samples. The true 6Do values are calculated according to
6DOStl- 6D0, 6Dost,- m s t 2 (RR R ~ ) , t i- (RR - RD),, (RR - R ~ ) , t i- (RR - R d s t 2 (8) where RR - RD is the average difference between reference and sample machine ratios in mass spectrometer units after pressure corrections. The subscripts stl, st2, and u refer to the two standards and the unknown sample. Although the difference RR - RD is the machine units, the differences (RR - R D ) , -~ (RR- R D )and ~ ~(RR- R D ~- (RR I - RD),, are linearly related to real differences in isotopic ratios; Le., the denom-
inators represent real isotopic ratios and can be treated as such. Since RR is the same for all standards and unknown samples, eq 8 can be rewritten as
Equations 6 and 7 can be written in the form of isotope ratios rather than 6 values as
1+
(CY -
a 1)/(1
+ 4P/K)'I2RD, (10)
(or R D ( H=~A R d and substituted in eq 9; because all samples have been corrected to the same pressure, A is cancelled out and
6DOOtl- 6Do,, RDo(lti)
- RDwm
-
6DOStl- 6Dou RDo(ltl)
- RDo(n)
(11)
Equation 11contains the true isotopic ratios of the different hydrogen gases used in the measurement of unknown 6 values; however, because eq 11is equivalent to eq 8, the mass spectrometer readings (i.e., the pressure corrected RR - RD)can be used in combination with the true 6Dovalues of the standards to determine the true 6 0 value of an unknown sample. Registry No. Hz,1333-74-0;Dz,7782-39-0.
LITERATURE CITED (1) Craig, H. Geochim. Cosmochim. Acta 1853, 3 , 53-92. Frazer, J. M. Microch/m. Acte 1862, 993-999. Sofer. 2. Anal. Chem. 1880, 52. 1389-1391. Frledman. I . Geochim. Cosmochim. Acta 1953, 4 , 89. Gulbransen, E. A.; Andrew, K. F. I n d . Eng. Chem. 1952, 4 4 ,
(2) (3) (4) (5)
1034-1038.
RECEIVED for review January 31,1986. Accepted March 24, 1986.