Agilent FPD-Low Level Sulfur Detection in CO2
Agilent FPD-Low Level Sulfur Detection in CO2
Agilent FPD-Low Level Sulfur Detection in CO2
Gas Chromatography
Authors Introduction
Roger L. Firor
Agilent Technologies, Inc. The food and beverage industry has strict guide-
2850 Centerville Road lines on the quality of CO2 required for use in
Wilmington, DE 19808-1610 products for human consumption. Food grade CO2
USA
(> 99.95%) can contain a number of impurities.
Bruce D. Quimby Impurities can produce off-odors or flavors and in
Agilent Technologies, Inc. some rare instances a health hazard can result.
2850 Centerville Road
Wilmington, DE 19808-1610
Guidelines1 have been established recently for
USA allowable levels of impurities by chemical type.
These have been published by the International
Society of Beverage Technologists (ISBT).
Abstract
The most common impurities are acetaldehyde,
An inert valving system using the volatiles interface (VI)
methanol, ethanol, hydrogen sulfide, carbonyl
on the Agilent 6890 is described for the analysis of ppb
sulfide, and sulfur dioxide. For example, the guide-
level volatile sulfur compounds in food grade carbon
dioxide. Examples using the Agilent AED, FPD, and line concentrations for total sulfur and total
Sievers SCD are shown. The blending of CO2 with an eight volatile hydrocarbons is 0.1 ppm v/v and
component sulfur mix provides an easy to use means of 50 ppm v/v, respectively. A chromatographic
calibrating the system for low level sulfur determinations. system based on the Agilent 6890 equipped with a
sulfur selective detector can easily quantify sulfur
at levels well below those recommended for food
grade CO2. Strict adherence to sample introduction
hardware recommendations is required for
reliable operation.
Experimental DMDS SO2 Mix
30 m × .25 mm capillary
The sulfur selective detectors (one per GC) were column (flow restrictor)
mounted on the Agilent 6890 gas chromatograph
equipped with the volatiles interface. The GC was
configured for 220 V operation in order to achieve Sulfur standards
high oven ramp rates. All GC gas flows and pres- To GSV
sures were controlled electronically. A 6-port CO2 He
gas-sampling valve (GSV) in Hastelloy C was inter- Point of use mixing occurs
3
in Swagelock or Valco Tee
faced directly to the Silcosteel treated VI with 4
5
1/16-inch Silcosteel tubing, Restek part
no. 20595. All lines coming in contact with the 3 Channel aux flow module
sample including the 1 cc sample loop were (installed in 6890, 2 channels free for other use)
Typical oven: -20 °C (9.5 min), 10 °C/min Agilent G2350A Cavity 245 °C O2 55 psi
to 250 °C, 80 °C/min to 280 °C Atomic Emission Transfer line 245 °C H2 45 psi
Sulfur 181 nm He makeup 100 mL/min
VI Inlet: 5 mL/min, constant flow mode
Sievers 355 803 °C Furnace P1 5.4 Detector
Temp 150 °C, Split ratio 0.1:1 Sulfur Chemiluminescent P2 171 Furnace
H2 100 mL/min
A dynamic blending system was used to add sulfur O2 8 mL/min
calibration mix, in a controlled fashion, to the pure O3 5.6 psi
CO2 matrix, producing a stream of known sulfur
concentrations for delivery to the GSV. This
system is shown in Figure 2. The sulfur calibration Results
mix (DCG Partnership I, LTD., Pearland, TX,
The FPD demonstrated excellent sensitivity to a
281-648-1894) consisted of the components listed
25 ppbv/component sulfur mix in pure CO2 as
in Table 1 at certified concentration levels of
shown in Figure 3. This level of sample mix was
5 ppm each.
prepared using the dynamic blending system by
2
mixing 1 mL/min sulfur calibration standard with components present in the CO2. Helium was used
199 mL/min CO2. Quantitation of all components to purge the lines thoroughly for approximately
except for H2S is possible at this level. Hydrogen 1 hour. Injections of helium were made into the
sulfide at levels of approximately 50 ppbv should GC-AED system to monitor any potential carry-
be quantifiable with the FPD when using the over. It is not uncommon to see residual levels of
sample introduction system described here. some sulfur mix compounds present for one or two
GSV injections after the mix is disconnected.
In Figure 4, AED chromatograms of the sulfur
181 nm line are shown at sulfur mix concentra- Figure 5 shows a SCD chromatogram of pure CO2
tions of 5 ppm (no dilution) and 35 ppb (dilution blended with the sulfur mix to 25 ppb. The SCD
in CO2). All eight sulfur compounds show good was able to detect down to levels of approximately
signal to noise at the 35 ppb level. 2 ppb for COS when using a 5 mL sample loop.
Carbonyl sulfide and carbon disulfide in pure CO2
Next, pure CO2 was analyzed without addition of can be seen at levels of 2 ppb and 1 ppb,
the sulfur calibration mix. Care was taken to respectively, on the AED in Figure 6.
ensure that carryover of residual sulfur com-
pounds was not mistakenly identified as
CS2
MeSH THT
H2S COS
t-BuSH
6 7 8 9 10 11 12 13 14 min
Figure 3. Carbon dioxide with 25 ppb sulfur calibration mix. Detector: Agilent Flame Photometric. One cc sample loop, no split.
5 ppm in CO2
5 8
4
3 7
2
1
35 ppb in CO2
5 10 15 20 25 30 35
Figure 4. Sulfur calibration mix in CO2 at levels of 5 ppm and 35 ppb. Detector: Agilent Atomic Emission.
3
www.agilent.com
CS2
COS THT
MeSH
H2S
2 4 6 8 10 12 min
AED: 1 cc loop
CS2
COS
~2 ppb ~1 ppb
7 8 9 10 11 12
A number of non-sulfur impurities can also be pre- cost and somewhat more difficult setup/tuning
sent as outlined in the ISBT guide. Although not compared to the FPD. The AED offers the most
shown in this application note, the AED could be complete solution when analysis of hydrocarbons
used to look at the S 181 nm and C 193 nm lines is also desired with a single valve-column-detector
simultaneously. Thus, many of the hydrocarbon combination.
impurities could easily be separated and quanti-
fied at low ppb levels using the 105 m × 0.53 mm ×
5.0 µm DB1 column in an Agilent 6890/AED References
system. The carbon 193 nm line is approximately 1. Carbon Dioxide Quality Guidelines and
five times more sensitive than the FID for those Analytical Procedure Bibliography, Interna-
compounds that respond well on a FID. Therefore tional Society of Beverage Technologists,
detectability of the hydrocarbon impurities would Homosassa, FL, (1999)
not be difficult at the levels required.
Agilent shall not be liable for errors contained herein or for incidental or consequen-
Conclusions tial damages in connection with the furnishing, performance, or use of this
material.
For detection of sulfur impurities in beverage Information, descriptions, and specifications in this publication are subject to change
without notice.
grade CO2 at levels down to 25 ppb, the FPD offers
an easy-to-use low cost solution when coupled to Copyright© 2001
the optimized sample introduction system Agilent Technologies, Inc.