5
02 W. Meredith et al.
results in the same highly selective conversion of unsatu-
rated cholesterol into cholestane, as previously found for
saturated cholestanol when using the molybdenum catalyst,
together with a far higher degree of conversion. Such
improvements are necessary if the potential of hydropyr-
olysis for the analysis of structurally complex compounds
such as steroids is to be realised.
capillary column (DB5, Agilent Technologies, Wokingham,
UK; 50 m ꢃ 0.25 mm i.d. ꢃ 0.25 mm film thickness), with
helium as the carrier gas and a temperature programme of
ꢂ1
1808C (hold for 2 min) to 3008C (hold for 5 min) at 88C min
.
RESULTS AND DISCUSSION
Hydropyrolysis of steroids
The selective defunctionalisation of steroids by hydropyr-
olysis represents a significant analytical challenge owing to
their structural complexity, both in terms of oxygen
functionality and the presence of carbon double bonds.
EXPERIMENTAL
Combustion/IRMS of starting materials
The carbon isotopic composition of an aliquot of three C27
steroids; 5a-cholestanol, 5a-cholestan-3-one and 5-cholesten-
Previous experiments with
a sulphided molybdenum
3
b-ol (cholesterol) (Sigma-Aldrich, Gillingham, UK), were
catalyst have shown a very high selectivity of conversion
of cholestanol into cholestane, but the double bond present in
cholesterol resulted in significant generation of cholestenes,
isomerisation of produced cholestanes and rearrangement of
measured in triplicate by combustion/IRMS using an
elemental analyser (EA1112, Thermo Scientific, Bremen,
Germany) coupled to an isotope ratio mass spectrometer
plus
18
(
DELTA
XP, Thermo Scientific). The combustion
the products to diasteranes. Such transformations are
temperature was controlled at 10208C.
undesirable prior to the GC/C/IRMS analysis of the
products and must be eliminated if the potential of
hydropyrolysis for the analysis of more structurally complex
steroids is to be realised. Therefore, to assess the efficacy of
the platinum catalyst we subjected 5a-cholestanol, 5a-
cholestan-3-one and 5-cholesten-3b-ol (cholesterol) to hydro-
pyrolysis.
Hydropyrolysis
Aliquots (ꢁ1 mg) of the steroids were adsorbed onto a bed
(
150 mg) of 5 wt % platinum on activated carbon, and
subjected to hydropyrolysis at 3008C and 15 MPa, as
1
7,18,21
previously described in detail.
Briefly, samples were
ꢂ1
heated resistively from 508C to 1008C at 3008C min , and
GC/MS and GC/C/IRMS analyses indicated that hydro-
pyrolysis of 5a-cholestanol over the platinum catalyst
produced very similar results to those seen with the
ꢂ1
then from 1008C to 3008C at 4C min , under a hydrogen
pressure of 15 MPa. For partial conversion experiments, the
1
8
heating programme was halted at 2008C and 2508C. A
hydrogen sweep gas at a flow rate of 5 L min , measured at
molybdenum catalyst, with highly selective reduction to
a single molecular product in 5a-cholestane. The silica
recovered from the dry ice cooled trap was found to contain
no detectable products indicating that the lower tempera-
tures employed with the platinum catalyst were insufficient
to volatilise either the starting samples or the defunctiona-
lised products. Therefore, the degree of conversion at each
experimental temperature (Table 1) could be accurately
assessed by the distribution of the compounds recovered
from the reactor at the end of each run, as shown for 5a-
cholestanol in Fig. 1, and cholesterol in Fig. 2.
ꢂ1
ambient temperature and pressure, ensured that any gaseous
products were quickly removed from the reactor vessel.
Unlike previous hydropyrolysis procedures where the
2
2
products were recovered from a dry ice cooled silica trap,
the lower temperatures employed ensured that the defunc-
tionalised products, together with any unreacted material
remained adsorbed to the activated carbon in the reactor.
These were then recovered from the carbon residue by means
of a short chromatographic column with successive elutions
of 5 mL dichloromethane (DCM) and 5 mL DCM/methanol
At 2008C it was clear that the degree of conversion of 5a-
cholestanol was very low (ꢁ5%), with large amounts of the
starting compound remaining (and minor amounts of
cholestanone generated). Raising the temperature to 2508C
significantly increased conversion to almost 50%, with the
product largely composed of 5a-cholestane (>95%). A very
small amount of 5b-cholestane was also formed. At 3008C,
however, while the degree of conversion was greater than
99%, it appeared that the higher temperature promoted some
isomerisation, with the proportion of 5b-cholestane pro-
duced increasing to ꢁ10%. This was confirmed by a further
run at 3508C (not shown), in which the proportion of 5a-
cholestane recovered was ꢁ20%, while that for 5b-cholestane
was ꢁ15%, with the remainder comprised of more highly
altered isomers of cholestane.
(
1:1 v/v), with the two fractions then combined before
analysis.
Gas chromatography/mass spectrometry (GC/
MS) of hydropyrolysis products
GC/MS analyses in full scan mode were performed on a CP-
3
800 gas chromatograph (Varian, Oxford, UK), interfaced to
a 1200 mass spectrometer (electron ionisation mode, 70 eV).
Separation was achieved on a VF-1MS fused-silica capillary
column (Varian; 50 m ꢃ 0.25 mm i.d. ꢃ 0.25 mm film thick-
ness), with helium as the carrier gas, and an oven programme
of 508C (hold for 2 min) to 3008C (hold for 6.75 min) at
ꢂ1
88C min
.
GC/C/IRMS of hydropyrolysis products
plus
The results for 5a-cholestan-3-one were very similar to
those for 5a-cholestanol (Table 1), and suggested that the
substitution of the alcohol functional group with a ketone
had no effect on the efficiency or selectivity of the
defunctionalisation promoted by hydropyrolysis over a
platinum catalyst.
A GC/C/IRMS instrument (DELTA XP, Thermo Scien-
tific) was used for the determination of compound-specific
1
3
d C isotope ratios of the products from the hydropyrolysis of
cholestanol, cholesterol and cholestanone (preformed in
triplicate). Separation was performed on a fused-silica
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 501–505
DOI: 10.1002/rcm
Meredith, Will
