CO2_2016 - page 10

8
Chimica Oggi - Chemistry Today
- vol. 34(2) March/April 2016
(10). Specifically, LP conditions were
generated across a mega-bore
2
D
column (10 m × 0.53 mm ID), it being
connected to a mass spectrometer.
Flow modulation was again
performed with reduced gas flows,
namely in the range 6-7 mL min
-1
.
An additional noteworthy feature of
the research was the use of a longer
accumulation loop, a solution that
greatly improved the general post-
modulation peak shape quality.
A description of the beneficial effects
of using a mega-bore
2
D column,
along with a long loop, are herein
given. A 10 m × 0.53 mm ID column
was selected because such
capillaries have been demonstrated
to work ideally under LP conditions
(11). A 1.5 m × 0.25 mm ID restriction
was located before the
2
D column to
avoid sub-ambient pressure
conditions reaching the FM. The
dimensions of the accumulation loop
were 20 cm × 0.51 mm ID. Initial
optimization was performed on linear
C
10
alkane. The
1
D average LV was about 13 cm s
-1
(initial
flow: 0.44 mL min
-1
). During the accumulation step (2.7 s), the
loop LV was 3.7 cm s
-1
. With regard to the re-injection pulse
(400 ms), the flow exiting the modulator was calculated to be
approx. 7 mL min
-1
, leading to a
2
D average LV of ≈ 170 cm
s
-1
. A sequence of modulated peaks relative to the C
10
alkane is shown in Figure 5a.
The first impression was that the modulation process worked
well. However, after a more careful inspection of the three
peaks it can be easily observed that the second and third
modulated peak are not symmetrical; the asymmetry factor
was measured for the third peak at 10% height, and a value
of 0.75 was derived. Such poor peak shapes have been
related to the preservation of
1
D analyte concentration
profiles within the accumulation loop (12). Thermal
modulation, on the contrary, nullifies concentration gradients
through peak focusing.
The drawback was resolved by using a longer loop, in this
case 46 cm. It is assumed that such a modification works
because it leads to a double accumulation/re- injection step,
or four-stage process. In essence, modulation was performed
on a specific chromatography band through accumulation I,
re-injection I, accumulation II (these steps occurred inside the
loop), and then re-injection II (transfer of the analyte band
onto the second dimension). The long-loop experimental
result is illustrated in Figure 5b; the improvement in peak shape
is evident: an asymmetry factor of 1 was measured for the
third peak. It was hypothesized that the four-stage process
allowed more time for the solute to reach a more uniform
intra-loop concentration distribution. The occurrence of a
four-stage process was confirmed by the retention time delay
equivalent to one modulation period.
An FM GC×LP GC-qMS method was developed for the
analysis of fish (menhaden) oil fatty acid methyl esters
(FAMEs). The satisfactory chromatographic performance
attained on the mega-bore capillary can be seen in Figure
6, which reports a single modulation at about 38 min.
In the schematic, the first (apolar - 30 m × 0.25 mm ID) and
second (medium polarity - 8 m × 0.32 mm ID) dimension, and
loop, are shown with the same apparent ID. During the
passage from the first dimension to the loop, the length of the
analyte band will decrease due to the differences in ID. The
same concept, namely an increase in the length of the
analyte band, will occur during the passage from the loop to
the second column. The isothermal experiment was carried
out at a temperature of 120°C. The injector (gauge) pressure
was 84.1 kPa, while the auxiliary one was 40 kPa. During
accumulation, the
1
D + loop flow was calculated to be
approx. 0.44 mL min
-1
, or 7.3 μL s
-1
. The intra-loop LV was
calculated to be about 3.4 cm s
-1
. Considering a duration of
the accumulation period of 4.5 s, then at the end of the
accumulation step the chromatography band should occupy
a loop length of just over 15 cm (Figure 4A). During the
re-injection process, the loop +
2
D flow rate, was about 6.1 mL
min
-1
, corresponding to an intra-loop LV of approx. 47 cm s
-1
,
and a second-dimension one of about 180 cm s
-1
. The
application of a re-injection step of 500 ms was sufficient to
push the chromatography band for a length of just over 23
cm, hence entirely out of the loop (Figure 4B). The peak base
width (4
σ
), for the most intense modulated peak, was 780 ms.
An FM GC×GC-qMS experiment was carried out on a
commercial fragrance; the
2
D flow rate was calculated to be
nearly 8 mL min
-1
, at the beginning of the analysis. A side-by-
side comparison was made by connecting the primary
column directly to the MS ion source. The perfume was
analyzed under the same injection (sample volume and split
ratio) and temperature program conditions. The applied
pressure generated an average LV of 30 cm s
-1
(the GC×GC
1
D average LV was approx. 11 cm s
-1
). For 22 compounds
(with different polarities) it was found that the FM GC×GC-
qMS signal intensities were generally higher (on average by a
factor of 1.5). Such an outcome was a great improvement
compared to previous research (8).
In a later research, an approach defined as FM low-pressure
GC×GC-MS (abbreviated as GC×LP GC-MS), was introduced
Figure 4.
Schematic illustrating an efficient FM process: (A) accumulation; (B) re-injection; (C)
modulated C
9
alkane. Reproduced with permission from Elsevier [Tranchida P.Q., Franchina F.A.,
et al., J. Chromatogr. A 2014, 1359, 271-276].
1,2,3,4,5,6,7,8,9 11,12,13,14,15,16,17,18,19,20,...68
Powered by FlippingBook