Homogeneous catalysis of valeronitrile hydrolysis under supercritical conditions

Article abstract of DOI:10.1002/cssc.201100443

Supercritical nitrile hydrolysis can be used for both, amide and acid production as well as waste water treatment, as the hydrolysis products show good biodegradability. The conventional process at ambient conditions requires large amounts of mineral acid or base. Approaches that use supercritical water as a green solvent without a catalyst have been investigated over recent years. Findings for valeronitrile hydrolysis presented recently showed promising reaction rates and valeric acid yields. In an attempt to further maximize product yield and to better understand the impact of the pH, reactions in dilute sulfuric acid (0.01 mol L^-1) were performed in a continuous high-pressure laboratory-scale apparatus at 400-500 °C, 30 MPa, and a maximum residence time of 100 s. Results from both reaction media were compared with regard to productivity and sustainability. Supercritical tuning: The outstanding characteristic of water as a reaction medium is the possibility of tuning its properties by changing the temperature and pressure. In combination with small amounts of sulfuric acid, even the most resistant compounds show significant conversion within seconds. The kinetics of the hydrolysis of valeronitrile are investigated under supercritical conditions, which results in remarkable yields of valeric acid. Copyright

Full text of DOI:10.1002/cssc.201100443

DOI: 10.1002/cssc.201100443
Homogeneous Catalysis of Valeronitrile Hydrolysis under
Supercritical Conditions
Michael Sarlea, Sabine Kohl, Nina Blickhan, and Herbert Vogel*[a]
Supercritical nitrile hydrolysis can be used for both, amide and
acid production as well as waste water treatment, as the hy-
drolysis products show good biodegradability. The convention-
al process at ambient conditions requires large amounts of
mineral acid or base. Approaches that use supercritical water
as a green solvent without a catalyst have been investigated
over recent years. Findings for valeronitrile hydrolysis present-
ed recently showed promising reaction rates and valeric acid
yields. In an attempt to further maximize product yield and to
better understand the impact of the pH, reactions in dilute sul-
furic acid (0.01 molLꢀ1) were performed in a continuous high-
pressure laboratory-scale apparatus at 400–5008C, 30 MPa, and
a maximum residence time of 100 s. Results from both reaction
media were compared with regard to productivity and sustain-
ability.
Introduction
Organic amides and acids are versatile intermediates used in
the chemical and pharmaceutical industries.[1,2] The well-known
nitrile hydrolysis process can deliver both products in satisfac-
tory yields if conducted accordingly.[3] However, nitrile hydra-
tion usually shows low reaction rates at ambient conditions,
and, therefore, strong acids, bases, or enzymes are used as cat-
alysts.[4,5] As a result, large amounts of environmentally un-
friendly brine are generated during the extraction of the de-
sired products.[3,6]
stability of valeronitrile (VN) as significant conversion took
place above 4008C in the given residence time. Nevertheless,
significant valeronitrile conversion and valeric acid (VS) selec-
tivity were achieved at high temperatures.[21]
An interesting approach is the combination of both meth-
ods—hydrolysis of valeronitrile in dilute sulfuric acid under su-
percritical conditions. Thus, an assessment of homogeneous
catalysis on the reaction rates will be performed, which should
deliver similar yields and selectivities at shorter reaction times.
For a valid comparison, catalyzed kinetic experiments on valer-
onitrile, valeramide (VA), and valeric acid were performed
under the same reaction conditions as the tests performed in
pure water.
Over the last two decades, organic reactions have been per-
formed in supercritical water as it is inexpensive, nonflamma-
ble, nontoxic, and abundant.[7–13] In particular, the possibility of
tuning properties such as polarity and ionic product by chang-
ing the temperature (T) and pressure (p) qualifies supercritical
water as an outstanding solvent.[14–20] Usually, hydrolysis reac-
tions are performed at near-critical conditions to take full ad-
vantage of the high acidity possessed by water (Figure 1). Fur-
thermore, supercritical water is suitable for reactions that re-
quire high temperatures to deliver satisfactory yields. Previous
experiments in pure water have demonstrated the remarkable
Kinetic results were gathered and evaluated by using the
Presto-Kinetics program. All findings were compared in terms
of productivity and sustainability with the results of the reac-
tion in pure water.
Results and Discussion
In general, dissociation above the critical point of water
(3748C, 22 MPa) decreases rapidly with temperature
(Figure 1).[22,23] The reduced ability to stabilize ions has been in-
vestigated by Xiang et al., who observed a strong dependence
between density and dissociation above 3508C.[24] The low die-
lectric constant has an impact on salt solubility and polar reac-
tion pathways.[25,26]
Acid-catalyzed nitrile hydrolysis is well-known, and the con-
ventional approach demands 70% sulfuric acid at 1008C and
[a] Dr. M. Sarlea, Dr. S. Kohl, N. Blickhan, Prof. H. Vogel
Ernst-Berl-Institut fꢀr Technische und Makromolekulare Chemie
Technische Universitꢁt Darmstadt
Petersenstrasse 20, 64287 Darmstadt (Germany)
Fax: (+49)6151-163465
~
&
*
Figure 1. Plot of ion product ( ), density ( ), and dielectric constant ( ) vs.
temperature at p=30 MPa.
E-mail: vogel@ct.chemie.tu-darmstadt.de
200
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2012, 5, 200 – 205

Valeronitrile Hydrolysis under Supercritical Conditions
several hours residence time.[3] Similar conversion levels were
attained within minutes in pure supercritical water.[21]
The behavior of sulfuric acid under supercritical conditions is
quite different compared to that at ambient temperature. The
second dissociation constant is very small; thus, the first disso-
ciation reaction may be considered as the only source of hy-
drogen ions. At comparable conditions of 4008C and 30 MPa, a
dilute sulfuric acid solution of approximately 0.013 molLꢀ1 ex-
hibits a pH of 3.32.[24]
gies and prefactors for each reaction step, which were com-
pared with the data obtained in pure water.
The parameters were modeled by using Presto-Kinetics. The
systematic error for plug-flow idealization of tubular-flow reac-
tor data was assessed and resulted in a 2% deviation, which is
low compared to literature values.[33] Both, irreversible and
equilibrium reactions between nitrile, amide, and acid were
tested, which displayed only slight differences in fit perfor-
mance. The results with the best fit quality depicted below
were obtained with the reaction pathway described in the lit-
erature. The degradation products (DP) of valeric acid in the re-
action network are shown in Scheme 3.
The reaction pathway consists of the consecutive addition of
two water molecules to the nitrile/amide functional group and
subsequent ammonia elimination (Scheme 1).[27–29]
Scheme 3. Reaction network used to model the experimental data.
The experimental concentration–time data were processed
by application of the abovementioned reaction network using
the experimental rate constants as starting parameters. The
procedure was repeated for all temperature and reactant data
sets. A close match between the least-square fit and measured
values for all data sets was attained (Figure 2). New reaction
Scheme 1. Valeronitrile hydrolysis pathway via valeramide to yield valeric
acid.
The yield of valeric acid was diminished under the afore-
mentioned conditions because of parasitic decarboxylation re-
actions (Scheme 2).[30,31] As such, the desired outcome of the
Scheme 2. Decarboxylation of valeric acid to butane and CO2.
experiments would be a substantial enhancement of the rate
of reaction of the hydrolysis steps along with unchanged or
only slightly increased decarboxylation rates. Consequently,
higher valeramide and valeric acid yields would be achieved
compared to the levels attained in pure supercritical water.
Hence, the impact on each reaction step was investigated. The
findings are displayed below and start with valeric acid decar-
boxylation followed by valeramide and valeronitrile hydrolysis.
&
Figure 2. Modeled concentrations and experimental data of valeronitrile ( ),
~
!
valeramide ( ), and valeric acid ( ) hydrolysis at T=3758C and p=30 MPa.
rate constants were extracted from the modeled data and
used for the Arrhenius activation energy (EA) calculation and
prefactor (A) determination for each reaction step and were
compared with the experimental results.
Kinetic analysis and reaction modeling
The reaction equilibria described in the literature were investi-
gated and revealed an insignificant formation of reverse reac-
tion products. Given that water is available in excess, the
change in its concentration can be neglected. The level of hy-
drogen ions remained constant and does not appear in the ki-
netic equations. Thus, pseudo-first-order reaction kinetics were
used to evaluate the experimental data.[32] Furthermore, high
amide and acid selectivities occurred without byproduct for-
mation. This allows the evaluation of the kinetic parameters by
using a global approach between 400 and 5008C. The result-
ing reaction rate constants delivered Arrhenius activation ener-
Valeric acid stability
Significant valeric acid conversion takes place above 4608C in
the predetermined residence time. At 4608C, no difference be-
tween the catalyzed and uncatalyzed reaction was observed;
at 53 s residence time, 8% conversion was registered for both
media. However, for the same residence time at 5008C, an in-
crease in conversion of 22% in water and 28% in dilute acid
was recorded (Figure 3). Thus, the impact on valeric acid decar-
ChemSusChem 2012, 5, 200 – 205
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
201

H. Vogel et al.
Figure 3. Comparison of valeric acid conversion against residence time at
Figure 4. Comparison of valeramide conversion against residence time at
ꢀ1
~
!
~
!
&
&
*
*
T=460 ( ), 475 ( ), 490 ( ), and 5008C ( ), p=30 MPa; concentration of va-
T=400 ( ), 425 ( ), 450 ( ), and 4758C ( ), p=30 MPa; a) c=0.03 molL
leronitrile: a) c=0.07 molLꢀ1 in water and b) 0.01 molLꢀ1 sulfuric acid.
in water and b) 0.01 molLꢀ1 sulfuric acid.
boxylation is rather small, which is in accordance with the ex-
periments conducted by Belsky et al. on the decarboxylation
of acetic acid derivatives.[34]
Activation energies sustain the high impact of temperature
on valeramide hydrolysis in dilute acid. The uncatalyzed reac-
tion displayed a 27 kJmolꢀ1 higher activation energy, and the
42 kJmolꢀ1 difference in the modeled results is even larger
(Table 2).
Both the experimental and modeled activation energies of
the catalyzed reaction show an increase of 12 kJmolꢀ1. As both
media show similar conversions at low temperatures and by
taking into consideration that the temperature has a higher
impact on the acid-catalyzed reaction, the resulting activation
energy seems plausible. The Arrhenius prefactor is one order
of magnitude higher than the prefactor of the uncatalyzed re-
action (Table 1).
Table 2. Comparison of the activation energies and Arrhenius prefactors
obtained from the experimental and modeled data for valeramide hydrol-
ysis.
Reaction step
E
A [kJmolꢀ1
]
Prefactor A [sꢀ1
]
experimental
modeled
experimental
modeled
The findings appear beneficial to preserve high valeric acid
yields throughout the given temperature and residence time
ranges.
water: VA!VS
acid: VA!VS
114
87
114
72
1.08ꢁ107
4.70ꢁ105
1.03ꢁ107
5.21ꢁ104
Table 1. Comparison of the activation energies and Arrhenius prefactors
obtained from the experimental and modeled data for valeric acid con-
version.
As expected, a significant increase in valeric acid yield was
attained, as the enhancement of the hydrolysis step exceeded
the decarboxylation acceleration already presented. Valeramide
hydrolysis in pure water resulted in a maximum yield of 90%
valeric acid at 4508C and 53 s residence time. At higher tem-
peratures, a definite limit to the yield was met owing to the
consecutive decarboxylation reaction. In dilute sulfuric acid,
higher yields and selectivities were attained, which is remark-
able considering the high reaction temperatures. Thus, yields
of 99% valeric acid were recorded after 18 s at 4508C and
after 9 s at 4758C. At longer residence times, the yields drop-
ped because of the subsequent decarboxylation of valeric acid
(Figure 5).
Reaction step
EA [kJmolꢀ1
]
Prefactor A [sꢀ1
]
experimental
modeled
experimental
modeled
water: VS!DP
acid: VS!DP
143
155
143
154
2.03ꢁ107
1.70ꢁ108
1.83ꢁ107
1.31ꢁ108
Valeramide hydrolysis
Valeramide is an intermediate of the valeronitrile conversion
and was separately investigated, as one of the two hydrolysis
steps would act as a bottleneck for the consecutive reaction.
Valeramide shows higher temperature sensitivity than valeric
acid. Reaction conditions were aligned with the previous ex-
periments in pure water. The acid-catalyzed reaction shows
considerable enhancement between 400–4758C (Figure 4).
Valeramide conversion reaches 81% after 9 s at 4008C, where-
as only 24% conversion is attained in pure water under the
same conditions. At higher temperatures, all of the valeramide
is consumed within seconds.
Valeronitrile hydrolysis
The final step concerning the catalytic effect of acid on the re-
action pathway is the hydrolysis of valeronitrile. Experiments in
pure water showed a higher stability of valeronitrile than valer-
amide. According to Figure 6, the impact of sulfuric acid on va-
leronitrile hydrolysis was not as tremendous as it was on valer-
amide conversion. A considerable effect can be observed at
202
www.chemsuschem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2012, 5, 200 – 205

Valeronitrile Hydrolysis under Supercritical Conditions
Table 3. Comparison of the activation energies and Arrhenius prefactors
obtained from the experimental and modeled data for valeronitrile hy-
drolysis.
Reaction step
EA [kJmolꢀ1
]
Prefactor A [sꢀ1
]
experimental
modeled
experimental
modeled
water: VN!VA
acid: VN!VA
98
79
89
79
4.81ꢁ105
1.70ꢁ104
8.64ꢁ104
2.16ꢁ104
Figure 5. Comparison of valeric acid yield against residence time at T=400
~
!
&
*
( ), 425 ( ), 450 ( ), and 4758C ( ), p=30 MPa; concentration of valera-
mide: a) c=0.03 molLꢀ1 in water and b) 0.01 molLꢀ1 sulfuric acid.
Figure 7. Comparison of valeric acid selectivity against residence time at
~
!
&
*
*
T=400 ( ), 425 ( ), 450 ( ), 475 ( ), and 5008C ( ), p=30 MPa;
a) c=0.07 molLꢀ1 valeronitrile in water and b) 0.01 molLꢀ1 sulfuric acid.
theless, a positive impact on valeric acid yields at short resi-
dence times over the whole temperature range was noted
(Figure 8). In combination with the aforementioned high selec-
tivity, technical applications are conceivable.
Figure 6. Comparison of valeronitrile conversion against residence time at
~
!
&
*
*
T=400 ( ), 425 ( ), 450 ( ), 475 ( ), and 5008C ( ), p=30 MPa; concentra-
tion of valeronitrile: a) c=0.07 molLꢀ1 in water and b) 0.01 molLꢀ1 sulfuric
acid.
Conclusions
The activation energies extracted from experimental and
modeled data exhibit a good correlation. The rates of reaction
for hydrolysis and decarboxylation were increased by the addi-
low temperatures and short residence times. For instance, a va-
leronitrile conversion of 15% at 4008C after 9 s was recorded,
whereas a conversion of 35% was observed in dilute acid
under the same experimental conditions. At higher tempera-
tures and longer residence times, the effect is less significant
inasmuch as the enhancement of the conversion remained
under 10%.
The catalytic effect resulted in an activation energy that was
approximately 20 kJmolꢀ1 lower (Table 3). Thus, the lower
impact compared to valeramide is mirrored by the kinetic
results.
A key aspect is the increase in selectivity (Figure 7). After 9 s
at 4508C, a remarkable 98% valeric acid selectivity was ob-
served. In addition, maximum selectivity shifted to shorter
times, especially at temperatures above 4508C.
However, valeric acid yields attained in dilute sulfuric acid
do not exceed the results for pure supercritical water. For ex-
ample, after 53 s at 4508C in water, a maximum yield of 80%
was obtained, whereas only 59% was observed in acid. Never-
Figure 8. Comparison of valeric acid yield against residence time at T=400
~
!
&
*
^
( ), 425 ( ), 450 ( ), 475 ( ), and 5008C ( ), p=30 MPa; concentration of
valeronitrile: a) c=0.07 molLꢀ1 in water and b) 0.01 molLꢀ1 in sulfuric acid.
ChemSusChem 2012, 5, 200 – 205
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
203

H. Vogel et al.
tion of sulfuric acid. Nitrile and, in particular, amide hydrolysis
show a decrease in activation energy at low pH values. In con-
trast, decarboxylation exhibits a slight increase; this finding, as
well as the calculated activation energies and prefactors, is
supported by the literature.[34,35] Experiments conducted with
trifluoric acetic and propionic acid exhibit an increase in activa-
tion energy of 14–30 kJmolꢀ1 when switching from the anionic
to the acidic form. The calculated difference of 12 kJmolꢀ1
matches this trend; the lower margin is caused by the lack of
halo groups that stabilize the negative charge on the anion,
which results in a lower dissociation constant.
tures with acid addition, a significant increase in selectivity was
attained for short residence times.
The substantial enhancement of valeramide conversion was
observed, which is supported by the calculated activation en-
ergies. The valeric acid yields reached 99%, which is a remark-
able improvement considering the short residence times in
comparison to the levels attained in pure water. As only a little
acid was required, the amounts of waste salt produced during
product purification should be acceptable from a technical
point of view. However, even low concentrations of sulfuric
acid enhance corrosion because sulfate anions act as oxidants
in high-temperature water.[36] The impact is particularly distinct
at temperatures around 2008C and results in a potential ho-
mogeneous active dissolution of the reactor material, which
occurs during heat-up or cool-down of the reaction solution.[37]
At higher temperatures, passivation of the alloy takes place.[38]
It might be beneficial to switch to phosphorous acid as a cata-
Belsky et al. have performed similar alkyl nitrile hydrolysis ex-
periments [R=CH3ꢀ, CH3CH2ꢀ, and (CH3)2CHꢀ] at near-critical
conditions in hydrochloric acid solutions and reported a de-
crease in activation energy with an increasing electron-donat-
ing character of R.[6] The findings presented in this work
(79 kJmolꢀ1 for R=CH3CH2CH2ꢀ) match Belsky’s trend very well
[127, 90, and 83 kJmolꢀ1 for R=CH3ꢀ, R=CH3CH2ꢀ, and R= lyst because it exhibits less aggressive properties at low con-
(CH3)2CHꢀ, respectively].
centrations. Conversely, reactor plugging attributable to poorly
soluble phosphates can occur.[36]
The impact of a low sulfuric acid concentration on the con-
version of valeronitrile in supercritical water has to be evaluat-
ed by taking each reaction step into consideration. Although
the nitrile conversion rate was only increased at low tempera-
Almost no catalytic impact was observed for valeric acid de-
carboxylation, which is a positive result, as no acceleration of
the degradation of the commercial product occurs.
Scheme 4. Schematic representation of the high-pressure plant used for nitrile hydrolysis under supercritical conditions.
204
www.chemsuschem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2012, 5, 200 – 205

Valeronitrile Hydrolysis under Supercritical Conditions
The overall yield of valeric acid was not improved signifi-
cantly; hence, pure supercritical water would still be the pre-
ferred solvent for industrial scale-up. The key benefit is a lower
environmental impact, as no waste brine is generated. Further-
more, corrosion issues, which could be induced by sulfuric
acid, and potential plugging caused by salts with poor solubili-
ty are avoided. Sulfate ions, which are not considered as an ox-
idizing agent at room temperature, may act as strong oxidizers
in high-temperature water. Thereby, metals might undergo fast
active dissolution.[36] Nickel-based alloys are widely used for
high-temperature applications under corrosive conditions.
These materials are attacked by most acidic solutions of high
density, regardless if sub- or supercritical. However, their corro-
sion rates are low in low-density supercritical solutions; hence,
a reactor made from these alloys is conceivable. However, pre-
heat and cool-down sections should be made from a different
material. Niobium and tantalum are resistant to H2SO4 solu-
tions below approximately 3508C.[39]
Keywords: homogeneous catalysis · hydrolysis · kinetics ·
supercritical fluids · sustainable chemistry
[1] B. Izzo, C. L. Harrell, M. T. Klein, AIChE J. 1997, 43, 2048–2058.
[2] H. Rçmpp, J. Falbe, M. Regitz, RꢂMPP Chemie Lexikon 9th Ed., Thieme
Verlag, Stuttgart, 1996.
[3] R. M. Lincoln (Atlantic Richfield Company), US 3876691A, 1975.
[4] M. Ress-Loeschke, B. Hauer, R. Mattes, D. Engels (BASF AG), DE
10010149, 2003.
[5] C. O’Reilly, P. D. Turner, J. Appl. Microbiol. 2003, 95, 1161–1174.
[6] A. J. Belsky, T. B. Brill, J. Phys. Chem. A 1999, 103, 3006–3012.
[7] B. Subramaniam, M. A. McHugh, Ind. Eng. Chem. Process Des. Dev. 1986,
25, 1–12.
[8] M. Perrut, Ind. Eng. Chem. Res. 2000, 39, 4531–4535.
[9] A. R. Katritzky, D. A. Nichols, M. Siskin, R. Murugan, M. Balasubramanian,
Chem. Rev. 2001, 101, 837–892.
[10] P. E. Savage, S. Gopalan, T. I. Mizan, C. J. Martino, E. E. Brock, AIChE J.
1995, 41, 1723–1778.
[11] H. Weingꢂrtner, E. U. Franck, Angew. Chem. 2005, 117, 2730–2752;
Angew. Chem. Int. Ed. 2005, 44, 2672–2692.
[12] P. G. Jessop, W. Leitner, Chemical Synthesis Using Supercritical Fluids
(Eds.: P. G. Jessop, W. Leitner), Wiley-VCH, Weinheim, 1999, Chapter 1.
[13] S. E. Hunter, P. E. Savage, Chem. Eng. Sci. 2004, 59, 4903–4909.
[14] E. Dinjus, R. Fornika, M. R. Scholz, Chemistry under Extreme or Non-Clas-
sical Conditions, 1st Ed. (Eds.: R. van Eldik, C. D. Hubbard), Wiley-VCH,
Weinheim, 1996, Chapter 6.
Nevertheless, the addition of catalytic amounts of acid is rea-
sonable, if valeric acid is produced directly from valeramide be-
cause of a significant impact on reaction rates as well as in-
creased yields and valeric acid selectivities.
Experimental Section
[15] X.-Y. Zheng, Y. Arai, T. Furuya, Trends Chem. Eng. 1996, 3, 205–217.
[16] N. Akiya, P. E. Savage, Chem. Rev. 2002, 102, 2725–2750.
[17] A. Kruse, E. Dinjus, J. of Supercritical Fluids 2007, 39, 362–380.
[18] S. Mittelstꢂdt, Anwendung der Chemie unter ꢀberkritischen Bedingungen
auf technisch interessante Synthesen, Dissertation, Technische Universitꢂt
Darmstadt (D), 1997.
[19] S. D. Iyer, M. T. Klein, J. of Supercritical Fluids 1997, 10, 191–200.
[20] B. Izzo, M. T. Klein, Ind. Eng. Chem. Res. 1999, 38, 1183–1191.
[21] M. Sarlea, S. Kohl, N. Blickhan, H. Vogel, ChemSusChem 2010, 3, 85–90.
[22] M. Uematsu, E. U. Frank, J. Phys. Chem. Ref. Data 1980, 9, 1291–1306.
[23] W. L. Marshall, E. U. Frank, J. Phys. Chem. Ref. Data 1981, 10, 295–304.
[24] T. Xiang, K. P. Johnston, W. T. Wofford, E. F. Gloyna, Ind. Eng. Chem. Res.
1996, 35, 4788–4795.
[25] F. J. Armellini, J. W. Tester, J. Supercrit. Fluids 1991, 4, 254–264.
[26] F. J. Armellini, J. W. Tester, J. Supercrit. Fluids 1994, 7, 147–158.
[27] J. Clayden, N. Greeves, S. Warren, P. Wothers, Organic Chemistry, Oxford
University Press, Oxford, 2001.
[28] R. Brꢃckner, Reaktionsmechanismen 2nd Ed., Spektrum Akademischer
Verlag, Heidelberg, 2002.
Testing plant
The schematic setup of the high-pressure plant is shown in
Scheme 4. The feed was pumped through the preheater into the
reactor by using an HPLC pump. The tubular reactor was made of
Inconel C276 (2.83 cm3 volume) and was spirally embedded into
an aluminum cylinder. The cylinder was enfolded by using an alu-
minum coat and heated by using a 500 W heating device. After
passing the reactor, the reaction mixture was cooled down and
depressurized. The plant was designed for a maximum tempera-
ture of 5608C and a pressure of 40 MPa. Residence times between
8–120 s were adjusted by flow variation.
Analysis
The reaction mixture was analyzed by using GC [SHIMADZU GC-
17 A, capillary column (CW 20M MN Fused Silica, L=19 m, inner di-
ameter (ID): 0.32 mm, df =0,5 mm), nitrogen as carrier gas, flame
ionization detector] and HPLC [KNAUER HPLC equipped with a
Knauer smartline RI detector 2300 and a Knauer smartline UV de-
tector 2500 (@210 nm), column (ProntoSIL C18 ACE-EPS, L=
250 mm, ID=3 mm, particle size=3.0 mm) and 2.5 mm sulfuric acid
in 9:1 water/acetonitrile as eluent].
[29] L. A. M. M. Barbosa, R. A. van Santen, J. Mol. Catal. A: Chem. 2001, 166,
101–121.
[30] S. Matsubara, Y. Yokota, K. Oshima, Org. Lett. 2004, 6, 2071–2073.
[31] M. Okazaki, T. Funazukuri, J. Mater. Sci. 2006, 41, 1517–1521.
[32] N. A. Bhore, M. T. Klein, K. B. Bischoff, Ind. Eng. Chem. Res. 1990, 29,
313–316.
[33] S. V. Ramayya, M. J. Antal, Energy Fuels 1989, 3, 105–108.
[34] A. J. Belsky, P. G. Maiella, T. B. Brill, J. Phys. Chem. A 1999, 103, 4253–
4260.
[35] D. Miksa, J. Li, T. B. Brill, J. Phys. Chem. A 2002, 106, 11107–11114.
[36] P. Kritzer, J. Supercrit. Fluids 2004, 29, 1–29.
Chemicals: valeronitrile (technical grade, purchased from Lonza),
valeramide (97%, Acros), and valeric acid (99%, Acros).
[37] P. Kritzer, N. Boukis, E. Dinjus, Corrosion 1998, 54, 689–699.
[38] P. Kritzer, N. Boukis, E. Dinjus, Corrosion 2000, 56, 1093–1104.
[39] P. Kritzer, N. Boukis, E. Dinjus, J. Supercrit. Fluids 1999, 15, 205–227.
Acknowledgements
The authors would like to thank Lonza AG, Visp, for their finan-
cial support of this work.
Received: August 9, 2011
Published online on December 20, 2011
ChemSusChem 2012, 5, 200 – 205
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
205