1792
Herbert H. P. Fang and Xiao-Shan Jia
C5H7O2N, each gram of biomass contains 0.124 g of N
and 0.531 g of C. Because the average nitrogen content in
protein is 16.0% (Speece and McCarty, 1964; Noguera et
al., 1994; Kuo et al., 1996), each gram of protein in the
mixed liquor represents 1.29 g (1.29 = 0.160/0.124) of bio-
mass.
After the addition of butyrate and formate and adjust-
ing the pH to 7.9, each vial was submerged in a 378C
shaking water bath. The vigorous shaking motion
(35 mmÂ125 strokes min 1) ensured complete mixing. At
given time intervals, the volume of biogas produced was
measured using a syringe, and the contents of the biogas
and mixed liquor were analyzed. All sludge handling and
solution transfers were conducted inside an anaerobic
workstation (Forma Scienti®c, Model 1029).
substrate; according to Boone and Whitman (1988),
out of the 43 species of methanogens identi®ed 19
can use either formate or H2/CO2 as substrate; (3)
formate and H2/CO2 are of similar levels of energy
of formation (Thauer et al., 1977), and can be
transferred rather freely from one to another
through enzymatic reactions.
MATERIALS AND METHODS
Enrichment culture
The anaerobic butyrate-utilizing sludge was enriched in
a 5.0-l water-jacketed chemostat reactor at 378C. The reac-
tor was seeded with the anaerobic sludge obtained from a
UASB reactor (Fang and Chui, 1993) treating dairy waste-
Analytical
The protein content in the mixed liquor was measured
by the folin method (Lowry et al., 1951), and the COD by
the standard methods (APHA, 1989). The biogas contents,
including methane, carbon dioxide and hydrogen, were
water. The granular sludge was disintegrated using
Waring blender before the seeding. Feed solution for the
a
1
chemostat reactor contained 5500 mg l
of butyrate,
1
equivalent to 10000 mg l of COD, as the sole substrate.
The solution also comprised nutrient and trace metals,
plus bicarbonate as the buering chemical, following a
formula used in previous studies (Kwong and Fang, 1996)
with the exception of excluding sulfate in the formulation.
The content in the reactor was completely mixed by recir-
culating the biogas using a compressor. The hydraulic
retention time was 10 days. The reactor was operated for
over 90 days to ensure reaching a steady state before
sludge was sampled for the batch tests. During the period
of day-60 to day-90, the VSS (volatile suspended solids)
analyzed by
a gas chromatograph (Hewlett-Packard,
Model Series II) equipped with a thermal conductivity
detector, and a 25 m  0.53 mm CarboPLOT P7 column
with a ®lm thickness of 25 mm. Argon was used as the car-
rier gas at a ¯owrate of 30 ml min 1. The column was
operated at a temperature program of 508C for 2.5 min
and then 1108C for 1.3 min. The injection port and detec-
tor were both kept at 1808C. Hydrogen content could be
detected down to the 10 6 atm level.
Concentrations of VFAs and alcohols were measured by
a second gas chromatograph of the same model equipped
with a ¯ame ionization detector and a capillary column
(Alltech, Econo-Cap FFAP, 30 m in length, 0.53 mm i.d.
and 1.2 mm ®lm thickness). The column was operated at a
temperature program of 708C for 4 min and then 1408C
for 3 min. The temperature program of injection port and
detector were the same, 2008C. Helium was used as the
carrier gas at a ¯ow rate of 40 ml min 1. Formate was
analyzed by an ion chromatograph (Shimadzu HPLC
10A) equipped with a CDD-6A conductivity detector and
1
average 361 mg l in the reactor, and the average COD
removal was 91%. The sludge yield was thus estimated as
0.040 mg VSS mg COD 1, which is comparable to the lit-
1
erature data of 0.037 mg VSS mg COD
1995b) and 0.047 mg VSS mg COD
McCarty, 1969). The mixed liquor on average contained
(Fang et al.,
(Lawrence and
1
1
1
150 mg l of residual butyrate and 610 mg l of acetate
with a pH ranging from 7.6 to 7.8. There were no fatty
acids detected other than residual butyrate and acetate.
The partial pressure of hydrogen (PH ) in the biogas was
10 5 atm.
a
Shim-Pack IC-A3 column. A solution containing
2
8.0 mM of 4-hydroxybenzoic acid and 3.2 mM of bis-[2-
hydroxyethyl]-iminotris-[hydroxyethyl]-methane was used
as the mobile phase. The ¯ow rate of the mobile phase
was 1.0 ml min 1, the oven temperature 408C and the
detector temperature 438C. Individual VFA (from C1 to
C7) could be detected at the 1 mg l level, whereas alco-
hols (from C1 to C4) could be detected at the 5 mg l
level.
Batch experiments
Two series of batch experiments on the by-product for-
mation of butyrate-degrading sludge were conducted in
120 ml glass serum vials. The ®rst series was to investigate
the eect of butyrate concentration. Five batches were run
using butyrate as the sole substrate at 500, 900, 3000, 6000
and 9000 mg l 1, respectively. The second series was to in-
1
1
vestigate the eect of increased PH by using formate as
the co-substrate for the degradation of butyrate. Two
batches were run using feed solutions containing 800 mg
2
RESULTS AND DISCUSSION
1
1
l
of butyrate plus 3000 mg l of formate in one, and
1 1
Degradation of butyrate
960 mg l of butyrate and 6000 mg l of formate in the
other. Formulation of the nutrient and trace elements for
the feed solution followed that used in several previous
studies (Jia et al., 1996). The nutrient stock solution was
sparged with nitrogen to strip o any dissolved oxygen.
The butyrate-degrading sludge from the chemostat reac-
In the ®rst series of experiments, ®ve batches
were run using butyrate as the sole substrate at 500,
900, 3000, 6000 and 9000 mg l 1, respectively. The
results of the ®ve batches were similar. Figure 1 il-
tor was used to seed each batch reactor. About 200 ml of lustrates the variations of butyrate and products in
mixed liquor was obtained from the chemostat reactor.
the mixed liquor, using the batch degrading 3000
1
mg l of butyrate for exempli®cation, throughout
After centrifugation, the biomass was washed with the
stock nutrient solution. After another centrifugation, the
biomass was re-suspended in 800 ml of fresh stock nutrient
the experiment, including (a) butyrate and acetate,
(b) methane, (c) PH , (d) i-butyrate, (e) propionate,
and (f) i-valerate and caproate.
solution. Each 120 ml vial was added with 60 ml of this
solution containing 70 mg l
2
1
of protein, equivalent to
1
90.3 mg l of VSS. Because of dilute concentration, the
biomass content in the solution could not be measured
directly by the VSS content. In this study, biomass was
estimated from the protein concentration of the mixed
Figure 1a illustrates that, after an acclimation
period of about 100 h, butyrate degraded rapidly
and became depleted after 200 h. During the period,
liquor. Assuming biomass has
a chemical formula of acetate, which was absent in the original feed