WORLD INTELLECTUAL PROPERTY ORGANIZATION International Bureau
PCT
INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION TREATY (PCT)

(51) International Patent Classification 7 : B01J 21/08, 23/40, C07C 27/00

Al

(11) International Publication Number: (43) International Publication Date:

WO 00/30744
2 June 2000 (02.06.00)

(30) Priority Data:
60/109,712
(21) International Application Number: PCT/US99/27421
(22) International Filing Date:      19 November 1999 (19.11.99)

24 November 1998 (24.11.98)    US

(71) Applicant: MICHIGAN STATE UNIVERSITY [US/US]; 238
Administration Building, East Lansing, MI 48824 (US).
(72) Inventors: ZHANG, Zhigang; 1633 F Spartan Village, East
Lansing, MI 48823 (US). MILLER, Denis, J.; 5214 Whitetail Circle, Okemos, MI 48864 (US). JACKSON, James, E.; 6232 W. Reynolds Road, Haslett, MI 48840 (US).
(74) Agent: McLEOD, Ian, C; 2190 Commons Parkway, Okemos, MI 48864 (US).
(81) Designated States: AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY, CA, CH, CN, CU, CZ, DE, DK, EE, ES, FI, GB, GE, GH, GM, HR, HU, ID, IL, IN, IS, JP, KE, KG, KP, KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR, TT, UA, UG, UZ, VN, YU, ZW, ARIPO patent (GH, GM, KE, LS, MW, SD, SL, SZ, TZ, UG, ZW), Eurasian patent (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), European patent (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE), OAPI patent (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG).

Published
With international search report.

(54) Title: CONDENSED PHASE CATALYTIC HYDROGENATION OF LACTIC ACID TO PROPYLENE GLYCOL
(57) Abstract
The present invention provides a process for production of propylene glycol with high yield and selectivity in an aqueous reaction mixture of lactic acid and hydrogen with an essentially pure elemental ruthenium catalyst on an inert support at elevated pressure and temperature. In particular, the present invention provides a process wherein the catalyst is a ruthenium salt deposited on a microporous support, reduced to ruthenium on the support with hydrogen, and oxidized in the presence of oxygen to provide a ruthenium oxide surface on the surface of the ruthenium metal and wherein the catalyst is maintained in the surface oxidized state until it is reduced with hydrogen prior to the reaction process.

FOR THE PURPOSES OF INFORMATION ONLY
Codes used to identify States party to the PCT on the front pages of pamphlets publishing international applications under the PCT.

AL Albania ES Spain LS Lesotho SI Slovenia
AM Armenia FI Finland LT Lithuania SK Slovakia
AT Austria FR France LU Luxembourg SN Senegal
AU Australia GA Gabon LV Latvia sz Swaziland
AZ Azerbaijan GB United Kingdom MC Monaco TD Chad
BA Bosnia and Herzegovina GE Georgia MD Republic of Moldova TG Togo
BB Barbados GH Ghana MG Madagascar TJ Tajikistan
BE Belgium GN Guinea MK The former Yugoslav TM Turkmenistan
BF Burkina Faso GR Greece    Republic of Macedonia TR Turkey
BG Bulgaria HU Hungary ML Mali TT Trinidad and Tobago
BJ Benin IE Ireland MN Mongolia UA Ukraine
BR Brazil IL Israel MR Mauritania UG Uganda
BY Belarus IS Iceland MW Malawi US United States of America
CA Canada IT Italy MX Mexico UZ Uzbekistan
CF Central African Republic JP Japan NE Niger VN Viet Nam
CG Congo KE Kenya NL Netherlands YU Yugoslavia
CH Switzerland KG Kyrgyzstan NO Norway ZW Zimbabwe
CI Cdte d'lvoire KP Democratic People's NZ New Zealand   
CM Cameroon    Republic of Korea PL Poland   
CN China KR Republic of Korea PT Portugal   
CU Cuba KZ Kazakstan RO Romania   
CZ Czech Republic LC Saint Lucia RU Russian Federation   
DE Germany LI Liechtenstein SD Sudan   
DK Denmark LK Sri Lanka SE Sweden   
EE Estonia LR Liberia SG Singapore   

CONDENSED PHASE CATALYTIC HYDROGENATION OF LACTIC ACID
TO PROPYLENE GLYCOL

CROSS-REFERENCE TO RELATED APPLICATION
The application claims the benefit of U. S. Provisional Patent Application Serial No. 60/109,712, filed November 24,   1998.
5
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
This   invention was   sponsored by  the  United
States Department of Agriculture under Contract Nos. 93-
10 37500-9585  and  98-35504-6356  CSRS.     The  United  States
Government has certain rights in this invention.

BACKGROUND OF THE INVENTION

15 (1)    Field of the Invention
The present invention relates to a process for production of propylene glycol in high yield and selectivity in an aqueous reaction mixture of lactic acid and hydrogen with an essentially pure  elemental
20 ruthenium   catalyst   on   an   inert   support   at   elevated
pressure and temperature. In particular, the present invention relates to the process wherein the catalyst is a ruthenium salt deposited on a microporous support, reduced to ruthenium metal on the support with hydrogen,
25 and oxidized with oxygen to provide a ruthenium oxide
surface on the ruthenium metal and wherein the catalyst is maintained in the surface oxidized state until it is reduced with hydrogen prior to the reaction process.

(2)    Description of Related Art
Propylene glycol is a nontoxic chemical used in  polymer  production;   as   a  nontoxic   antifreeze;   in food,      drinking,       cosmetics,       and      pharmaceutical
applications.    In the United States, the production rate
of propylene glycol is approaching one billion pounds
per year, with growth projected at 5 to 7% per year.
5 Commercial production of propylene glycol is
currently petroleum-based and involves the high pressure and high temperature hydrolysis of propylene oxide. Propylene oxide is manufactured by either the chlorhydrin    process    or    the    per-oxidation    process
10 (Szmant, H.H., Organic Building Blocks of the Chemical
Industry,    Wiley,    New   York,    New   York.    pp.    281-283 (1989)).    Making either propylene glycol precursor is a multi-step   process that involves a variety of organic solvents   and   results   in   the   formation   of   multiple
15 byproducts.    The selling price of propylene glycol using
these petroleum-based precursors is about $0.60 to $0.65 per pound.
Alternative processes for producing propylene
glycol    have    been    reported. For    example,     the
2 0 hydrogenation of organic esters to alcohols and glycols
was reported by Adkins and co-workers who were able to achieve 80% yields of propylene glycol from methyl lactate over copper/chromium oxide and Raney nickel catalysts at  150° to 25CPC and extremely high hydrogen
25 pressures of about 20 to 30    MPa   (2,900 to 4,350 PSIG)
(Bowden and Adkins, J. Am. Chem. Soc. 56: 689 (1934); Adkins and Billica, J. Am. Chem. Soc. 70: 3118 (1948); Adkins and Billica, J. Am. Chem Soc. 70: 3121 (1948)). In addition to high pressures,   high catalyst  loadings
30 were necessary as well to achieve these relatively high
yields.      Broadbent   et   al.    (J.   Org.    Chem.    24:    1847 (1959))   was able to obtain propylene glycol yields  as high as  80%  from ethyl  lactate  at   150°C  over  rhenium black catalysts,  but at very high hydrogen pressures of
35 about 25 MPa  (3,626 PSIG).    More recently,  U.S.  Patent
No. 5,731,479 to Antons discloses a process for
preparing optically active alcohols from optically
active carboxylic acids with hydrogen in the presence of
ruthenium catalysts. An example is provided for the
5 conversion of optically active L-(+)-lactic acid to L-
(+)-propane-1,2-diol. The yields obtained were about 88% over bulk ruthenium powder, and 74% over ruthenium supported on carbon. To achieve efficient conversion of an optically active carboxylic acid to its corresponding
10 optically active alcohol, the process is performed at a
relatively high pressure (10 to 20 MPa) and the reaction is allowed to proceed to completion.
While the aforementioned relate to the conversion of lactic acid and derivatives to propylene
15 glycol,  there    are several U.S.  Patents related to the
aforementioned which disclose various mixed bed catalysts and processes for their use in hydrogenation reactions for converting organic acid esters to their alcohols.
20 U.S.   Patent  Nos.   4,985,572   and  5,149,680   to
Kitson et al discloses a process for catalyzed hydrogenation of carboxylic acids and their anhydrides to produce the corresponding alcohol and/or carboxylic acid ester and catalysts for achieving the conversion.
25 The     process     is     particularly     applicable     to     the
hydrogenation of dicarboxylic acids and their anhydrides using catalysts that comprise an alloy of a noble metal from Group VIII of the Periodic Table of Elements and at least one metal capable of alloying with the Group VIII
3 0 noble metal.
U.S. Patent No. 5,478,952 to Schwartz discloses improved hydrogenation catalysts consisting essentially of highly dispersed, reduced ruthenium and rhenium on carbon supports.    Specifically,  the catalyst
3 5 relates   to  the  production  of  tetrahydrofuran,   gamma-
butyrolactone,    1,4-butanediol    and   the    like    from   a
hydrogenatable precursor such as maleic acid,  succinic
acid in an aqueous solution in the presence of hydrogen
and one of the above catalyst.
5 Finally,  U.S.   Patent No.   5,536,879 to Antons
et al. discloses a hydrogenation process which uses ruthenium catalysts to prepare optically active amino alcohols from optically active amino acids. The process is  performed at  relatively high pressures  to achieve
10 efficient    conversion    of    the    amino    acid    to    its
corresponding alcohol.
The catalysts used in the above related patents can be adapted to convert lactic acid to propylene glycol.    However,  the disclosed hydrogenation
15 reactions     generally     have     inadequate     yields     and
selectivities     to    propylene     glycol. Inadequate
selectivity results in the production of unwanted alkanes, such as methane and ethane, by side reactions. Because   these   byproducts   reduce   overall   yield   and
20 contaminate the final product,  the byproducts increase
the overall costs associated with producing propylene
glycol    from    lactic    acid. Therefore,     commercial
production of propylene glycol remains dependent on petroleum-based routes for its synthesis.
25 Thus, there remains a desire to provide a low-
cost process for producing propylene glycol from renewable resources. In particular, it would be desirable that the process enable the bulk production of propylene  glycol   in  a  process   that   can  be  performed
30 under relatively mild conditions which results in high
yields and selectivity for propylene glycol.

SUMMARY OF THE INVENTION

The present invention provides a process for
35 the low-cost,  bulk production of propylene glycol from
lactic acid. The process produces high yields of
propylene glycol from "crude or unrefined" lactic acid
under reaction conditions with high selectivity and,
therefore, low byproduct formation. Thus, the present
5 invention   provides   a   process   for   the   production   of
propylene glycol, which comprises: reacting a reaction mixture of lactic acid and hydrogen in water with a ruthenium catalyst on an inert support with a BET surface area between about 1 to 1,00 0 m2 per gram at a
10 hydrogen pressure between about 3.4 to 16.5 MPa  (500 to
2,400 PSIG) and a temperature between about 50° to 200°C, more preferably between 100° and 170°C; and removing the propylene glycol from the reaction mixture.
The    present     invention    also    provides    a
15 composition    which    comprises    an    essentially    pure
elemental   ruthenium on  an  inert  microporous   support, wherein the  ruthenium is  deposited on  the  support by drying a water solution of  the ruthenium salt on the support, reducing the salt to the elemental ruthenium on
20 the   support   with   hydrogen,    and   then   reacting   the
ruthenium on the support with oxygen to provide oxide of the ruthenium oh surfaces of the ruthenium. Preferably, the ruthenium is deposited on an inert microporous support selected from the group consisting of alumina,
25 titania,    silica,    alumina   silicate,    and   microporous
carbon.
The process of the present invention is preferably conducted at hydrogen pressures of 3.4 to 8.3 MPa    (500   to   1,200   PSIG)    in   a   continuous   or   batch
30 process.      When   the  process   is   a   batch  process,   the
reaction mixture is in a closed vessel provided with a stirrer which agitates the reaction mixture. When the process is a continuous process, the continuous process is  a preferably a  fixed bed  reactor  and  the  reaction
35 mixture   is   trickled   through   a   bed   of   the   catalyst.
Further, in the continuous process the gas flow can
either be up or down. In a preferred embodiment, the
hydrogen gas mixture is passed upward through the bed
and the aqueous reaction mixture flows downward. In the
5 continuous process it  is preferable that the reaction
mixture has a space velocity of between about 0.5 and 5 grams of lactic acid per gram of the catalyst per hour. In either the batch or continuous process, the reactor temperatures are preferably between about 50° and 2 00°C,
10 and more preferably between 100° and 170°C,   for maximum
results. Finally, to reduce the formation of reaction byproducts, the reaction mixture can further comprise sulfur added either to the liquid or gas phase.
The present  invention  further provides  that
15 the  lactic acid can be provided to the reaction as a
fermentate produced by a microorganism or other crude or unrefined preparation. In general, when the lactic acid is produced be a microorganism, the lactic acid may be converted    to    its    salt    during    processing    of    the
2 0 fermentate which can  then be  acidulated.     Therefore,
when the lactic acid is provided in a solution as its
salt or a partially acidulated salt, this "crude or
unrefined" reaction mixture may further comprise H2S04 or
other acid. Preferably, the H2S04 is added to the lactic
25 salt solution to provide a 1 to 1 molar ratio of lactate
salt to acid equivalent and the anion, other than lactate is substantially removed. In general, the lactic acid so produced is provided at a concentration between   about   5%   to   90%   by  weight   of   the   reaction
3 0 mixture.
Thus,   in a preferred embodiment,   the present
invention    provides    a    process    for    preparation    of
propylene glycol comprising:   (a)  providing a ruthenium
metal catalyst prepared by reacting hydrogen gas with a
3 5 ruthenium    salt    deposited    and    dried    on    an    inert
microporous  support having a BET surface area between
about 1 and 1,000 m2 per gram;   (b) drying the catalyst in
an oxygen containing atmosphere so as to provide oxide
of   the   ruthenium   on   surfaces   of   the   ruthenium;    (c)
5 introducing the dried catalyst into a reaction vessel;
(d) reacting the catalyst with the oxide on the surface with hydrogen in the vessel; (e) reacting in the vessel at elevated pressures and at elevated temperatures, a reaction  mixture   of   a   lactic   acid   in  water   in  the
10 presence  of  the  catalyst,   wherein  the  temperature  is
between about 50° to 200°C, preferably between 100° and 170°C and the hydrogen pressure is between about 3.4 to 16.5 MPa (500 to 2,400 PSIG) to produce the propylene glycol in the reaction mixture;  and   (f)   recovering the
15 propylene   glycol   from   the   reaction   mixture.      In   a
preferred embodiment, the ruthenium is deposited on a support selected from the group consisting of alumina, titania, silica, alumina silicate, and microporous carbon.    To reduce the formation of reaction byproducts,
20 the reaction mixture can further comprise sulfur.
In the foregoing processes of the present invention, the propylene glycol is recovered in a yield of at least 80%, preferably in excess of 90%, and a selectivity   to   propylene   glycol   of   at    least   80%,
25 preferably in excess of 90%.
The present invention further provides a composition which comprises an essentially pure elemental ruthenium on an inert microporous support, wherein  the  ruthenium is  deposited on the  support by
3 0 drying   a   water   solution   of   ruthenium   salt   on   the
support, reducing the salt to elemental ruthenium on the support with hydrogen, and passivating the composition in an oxygen containing atmosphere so as to provide an oxide  of   the  ruthenium on  surfaces  of   the  ruthenium.
35 Preferably,   the  inert microporous  support  is  selected
from the group consisting of alumina, titania, silica,
alumina silicate, and microporous carbon. In a
preferred embodiment, the inert microporous support has
a BET surface area between about 1 to 1,000 m2 per gram,
5 more preferably between about 100 to 1,000 m2 per gram.
Preferably, drying the ruthenium salt on the catalyst is at about 25°C for about 5 hours and then under a vacuum of about 3 0 inches of mercury at about 100°C for 12 hours.      Preferably,   reducing   the   ruthenium   salt   to
10 elemental ruthenium comprises the steps of   (a)  heating
the catalyst from 25°C to 400°C at a rate of about 0.5°C per minute under a flow of a gas consisting of 10 volume percent of hydrogen in helium at a rate of about 3 0 ml per minute;   (b)   maintaining the catalyst at 4 00°C and
15 changing the gas  to pure hydrogen;      (c)   reducing the
catalyst in the pure hydrogen for about 16 hours; and cooling the catalyst under a helium flow to room temperature. Preferably, passivating the catalyst by placing the reduced catalyst in a stream of about 1 to
20 10, preferably between 1 and 3, volume percent of oxygen
in argon or other inert gas at room temperature for about 1 hour. For using the composition for the conversion of lactic acid to propylene glycol, it is preferable that the ruthenium is reduced in a reaction
25 vessel with hydrogen prior to contact with a reaction
mixture.
Therefore,   it   is   an  object   of   the  present
invention to provide a process for producing propylene
glycol in a reaction mixture using a substrate that is
30 derived   from   renewable   resources.      In  particular,   a
process which uses lactic acid which can be produced inexpensively in fermentation cultures.
It is also an object of the present invention
to provide a process for producing propylene glycol from
35 lactic acid which has high selectivity and yield.
-Si-
It    is    further   an   object   of    the   present
invention to provide a composition that is a catalyst
for the conversion of lactic acid to propylene glycol
that    effects    the    conversion    at    high    yields    and
5 selectivities.

DESCRIPTION OF THE DRAWINGS

Figure 1 is a flow diagram for a process for
converting lactic acid to propylene glycol in a
10 continuous trickle-bed reactor.
Figure 2 is a flow diagram for a process for converting lactic acid to propylene glycol in a batch reactor.
Figure    3    is    a    bar    chart    that    shows    a
15 comparison of propylene glycol  (PG) yields as a function
of reaction time for several prepared catalysts and commercial catalysts.
Figure  4   is  a  three-dimensional   graph  that
shows   propylene   glycol   (PG)   yield   as   a   function   of
20 pressure and temperature.
Figure 5 is a graph that shows propylene glycol (PG) yield over time as a function of different initial lactic acid concentrations.
Figure б is a graph that shows the results of
25 a comparison of propylene glycol yields as a function of
time between neat lactic acid and acidulated calcium lactate.
Figure 7 is a graph that shows the effect of
liquid feedstock flow rate on catalyst performance.    The
3 0 graph   measures   the   percent   lactic   acid   conversion
(con%), percent propylene glycol yield (yield%) and selectivity at various lactic acid feedstock flow rates (WHSV).
Figure 8 is a bar graph that shows the percent
35 lactic acid   (LA)   conversion,   percent propylene glycol
-10-
(PG) yield, and percent dispersion for catalysts 1 through 10 which have the properties of catalysts 1 through 10 shown in Table 1.

5 DESCRIPTION OF PREFERRED EMBODIMENTS
As used herein, the terra "MPa" refers to mega-
Pascal which is the SI unit of measure for pressure or
stress.     1 MPa is equivalent to 1 PSIG x 6895.757/106;
or,    1   bar   x   100,000/106   or   1   atmosphere    (atm)    x
10 101, 325/106; or,  1 inches of mercury x 3386.388/106.
BET surface area of the inert supports was calculated according to the method of Brunauer, Emmett, and Teller (BET) disclosed in J. Am. Chem. Soc, 60: 309 (1938).
15 The   present   invention  provides   an   improved
process for the conversion of lactic acid in a feedstock to propylene glycol with high selectivity and yield. A key component of the process is a catalyst comprising ruthenium on  an  inert  support  which  includes  carbon,
20 alpha-alumina,      gamma-alumina,      alumina     siliconate,
silica, or titania.
In a preferred embodiment, the catalyst comprises ruthenium metal deposited on an inert support such   as    carbon   or   alumina   which    is    preferably   a
25 microporous support that has a BET surface area between
about 1 to 1,000 m2 per gram, preferably between about 100 to 1,000 m2 per gram, so as to facilitate high selectivity of the reaction to propylene glycol. In a preferred   embodiment,    the   support   is   a   microporous
30 carbon support.    The ruthenium metal is deposited on the
support by mixing an aqueous solution of a precursor ruthenium salt in an amount just sufficient to fill the pore volume of the support. Preferably, the ruthenium salt is ruthenium chloride and the ruthenium loading is
35 between   about   4   to   5   wt%.       After   depositing   the
ruthenium in and/or on the support, the ruthenium salt
impregnated support is dried. In particular, the
catalyst is dried at room temperature for between about
5 to 16 hours, preferably about 5 hours, and then under
5 a vacuum at between about 50°C to 100°C for between about
4 to 12 hours, preferably 100°C for 12 hours. Preferably, the vacuum is about 3 0 inches of mercury. Then the ruthenium salt is reduced by heating at an elevated temperature under hydrogen.    In particular, the
10 catalyst is heated from 25°C to 400°C under a hydrogen
containing gas at a flow rate about 0.5°C per minute wherein the gas is between about pure hydrogen to 10 volume percent hydrogen in helium, preferably 10 volume percent hydrogen in helium.    When 400°C has been reached,
15 the gas is switched to pure hydrogen and the reduction
is continued for 8 to 16 hours, preferably 16 hours. After reduction, the catalyst is cooled to room temperature and passivated in an oxygen containing atmosphere which coats the surface of the ruthenium to
20 provide an oxide of the ruthenium on the surface of the
ruthenium. In particular, the catalyst is cooled under an argon, helium or other inert gas flow, preferably helium, and then passivated in a stream of oxygen between about 1 and 10, preferably between about 1 to 3
25 volume percent of oxygen in argon.    The oxidized metal
surface enables the catalyst to be maintained without
using elaborate storage conditions. Prior to using the
passivated catalyst in the improved process reaction,
the catalyst is reduced with hydrogen. After
30 reducing   the   catalyst,   the   lactic   acid   feedstock  is
added to the catalyst and the reaction is performed under pressure in a hydrogen atmosphere and at an elevated temperature. The catalyst which is prepared in this manner is  able to perform at high efficacy for a
35 significantly   longer   period   of   time   than   prior   art

catalysts. In particular, the catalyst facilitates the
efficient conversion of lactic acid to propylene glycol
at temperatures and hydrogen pressures that are
significantly milder than conditions reported in the
5 prior  art.     For  example,   the  process  of  the present
invention enables the efficient conversion of lactic acid to propylene glycol at 150°C and hydrogen pressures as low as 3.3 MPa   (479 PSIG).
Examples   of   catalysts   that   were   prepared
10 according to the present invention are shown in Table 1.
Of the prepared catalysts, the following catalysts are preferred in the process of the present invention which is a catalyst that consists of ruthenium metal on a microporous  support with a ruthenium metal  loading of
15 between about  4  to  5%  is preferred.     An example of a
ruthenium on microporous support catalyst which is preferred is CG6M-F (Table 1, run number 3) wherein the support is a 110+200 carbon support with a BET surface area of 728 m2 per gram (available from Cameron-Yakima,
2 0 Inc., Yakima, Washington).    The ruthenium metal loading
is about 5.0 wt% and is applied to the support as an aqueous solution of ruthenium chloride hydrate with a dispersion of about 13%.
Another example of a ruthenium on microporous
25 support  catalyst  that  is preferred is SG6-D   (Table  1,
run number 5) wherein the support is a 100 mesh microporous support with a BET surface area of 777 m2 per gram (available from Cameron-Yakima, Inc.) . The ruthenium metal loading is about 4.4 wt% and is applied
3 0 as  an  aqueous   solution of  ruthenium chloride  hydrate
with a dispersion of about 10%.
Run Catalyst Support Precursor Dispersion Loading

Name BET (m2/gram)

1 CG5P-G Cameron-Yakima,  Inc. CG5P-200 mesh 648 Aldrich ruthenium (III) chloride hydrate (aq) 6% 5.0%
2 CG5P-H Cameron-Yakima,  Inc. CG5P-200 Mesh 648 Aldrich ruthenium (III) chloride hydrate (EtOH) 3% 5.0%
3 CG6M-F Cameron-Yakima,  Inc. macroporous-100 +200 728 Aldrich ruthenium (III) chloride hydrate (aq) 13% 5.0%
4 SA135-C Aldrich silica-alumina grade 135,  -100 mesh 440 Aldrich ruthenium (III) chloride hydrate (aq) 14% 5%
5 SG6-D Cameron-Yakima,  Inc. microporous 100 mesh 777 Aldrich ruthenium (III) chloride hydrate (aq) 10% 4.4%
6 CG5P-N01-I Cameron-Yakima,  Inc. CG5P-200 mesh 648 Alfa ruthenium nitrosyl nitrate hydrate (aq) 38% 5%
7 CG5P-A Cameron-Yakima,  Inc. CG5P (20*50 mesh) 648 Aldrich ruthenium (III) chloride hydrate (aq)
10% 5.4%
8 AL100-B Aldrich alpha aluminum oxide (-100 +200 mesh) 0.24 Aldrich ruthenium (III) chloride hydrate (aq) 0% 5%
9 ALg-E Alfa gamma aluminum size +100 mesh 45 Aldrich ruthenium (III) chloride hydrate (aq) 13.5% 4.7%
10 TiP25-C19-J Degussa P25 Titania-200 mesh 49 Aldrich ruthenium (III) chloride hydrate (aq) 18% 5%

A    further    example    of     a     ruthenium    on
microporous support catalyst that is preferred is CG5P-
N01-1  (Table 1,  run number 6) wherein the support is a
microporous  CG5P-200  mesh  support  with  a  BET  surface
5 area of 648 m2 per gram  (available from Cameron-Yakima,
Inc.). The ruthenium metal loading is about 5.0 wt% and is applied as an aqueous solution of ruthenium nitrosyl nitrate hydrate with a dispersion of about 38%.
An example further still of a ruthenium on a
10 microporous support catalyst that is preferred is CG5P-A
(Table 1, run number 7) wherein the support is a microporous CG5P 20*50 mesh support with a BET surface area of 64 8 m2 per gram (available from Cameron-Yakima, Inc.).    The ruthenium metal loading is about 5.4 wt% and
15 is applied as an aqueous solution of ruthenium chloride
hydrate with a dispersion of about 10%.
A catalyst which consists of a ruthenium metal on a silica-alumina support with a ruthenium metal loading of between about 4 to 5% is also preferred.     An
20 example is SA135-C  (Table 1,  run number 4)  wherein the
ruthenium metal is on a Silica-alumina grade 135,   -100 mesh support with a BET surface area of 44 0 m2 per gram (available   from Aldrich  Chemical   Company,   Milwaukee, Wisconsin) .      The   ruthenium  metal   is   applied   to   the
25 support  as  an  aqueous  solution of   ruthenium chloride
hydrate with a dispersion of about 14%.
A catalyst which consists of a ruthenium metal on titania support with a ruthenium metal loading of between about 4 to 5% is also preferred.    An example is
30 TiP25-C19-J    (Table    1,    run   number    8)    wherein    the
ruthenium metal is on a titania 200 mesh support with a BET surface area of 49 m2 per gram (available from Degussa, Inc.). The ruthenium metal is applied to the support  as   an  aqueous   solution  of   ruthenium  chloride
3 5 hydrate.
Further still, a catalyst which consists of a
ruthenium metal on a gamma aluminum oxide support with
a ruthenium metal loading of between about 4 to 5% is
also  preferred.      An   example   is   ALg-E   (Table   1,   run
5 number   9)   wherein  the   ruthenium metal   is   on  a  gamma
aluminum oxide +2 00 mesh support with a BET surface area of 45 m2 per gram (available from Johnson-Matthey Chemicals, Inc.). The ruthenium metal is applied to the support   as   an  aqueous   solution  of   ruthenium  chloride
10 hydrate with a dispersion of about 13.5%.
An important aspect of the present invention is that "unrefined" or crude lactic acid preparations can be efficiently converted to propylene glycol in the foregoing  process.      "Unrefined"   lactic  acid  as  used
15 herein   can   contain   more   than   100   parts   per   million
alkali metal, alkaline earth metal, or ammonium salts of lactic acid. Therefore, "unrefined" lactic acid includes bacterial fermentates and other crude or impure preparations of lactic acid.    The conversion processes
20 of  the prior art  require pure preparations of  lactic
acid or its esters as the substrate for conversion to propylene glycol whereas the present invention can efficiently convert the lactic acid in crude preparations to propylene glycol.    This is an important
25 improvement over the processes of the prior art.    Thus,
in the present invention, the "unrefined" or impure lactic acid is either a fermentate or lactic acid containing lactate salts. Impure lactic acid preferably contains 50 to 100% lactic acid.
30 In    general,     "unrefined"     lactic    acid    is
produced by microorganisms such as the homolactic microorganisms Lactobacillus delbruceckii, L. bulgarcius, and L. leichmanii in a fermentor. The microorganisms produce lactic acid which is converted to
35 its    salt    during   the    fermentation   process,    through
addition of a base in order to control pH. Therefore,
in fermentation, lactic acid is produced as lactate
salt, which may be partially acidulated to produce a
lactic acid and lactate salt mixture. This mixture can
5 be treated by the process of the present invention which
provides an economic source of lactic acid. Residual lactate can be, further acidified and/or can be recycled into the crude broth. Lactate salt could be in the form as   a   fermentate,   alkali  metal,   and  alkali   earth,   or
10 ammonium   lactate.        The    lactic    fermentate   can   be
partially or fully acidulated. The conversion of lactate to its acid can be achieved by addition of sulfuric or other acid to the fermentate.
Since producing lactic acid by fermentation is
15 less  expensive than producing lactic acid by chemical
synthesis, it would be desirable to use fermentation methods to produce the lactic acid for the hydrogenolysis reaction. Furthermore, the fermentation process allows renewable resources to be used to produce
20 propylene glycol.   The following materials are renewable
resources which among other renewable resources can be used in lactic acid fermentation: products containing starch (q.v.), i.e., corn starch, potato starch, or rice starch;   whey;   dextrose;   corn  sugar;   cane  sugar,   beet
25 sugar,   or  molasses   of  beet   sugar;   and  sulfate   lyes.
Ammonium salts or other nutrient sources are used as nitrogen sources. The fermentation is performed according to methods well known in the art to produce lactic acid.
3 0 After      fermentation,      biomass,      magnesium
hydroxide, limestone, and other nutrients and ingredients that may have been in the water are removed by filtration. Optionally, the filtrate is concentrated by   evaporation,    then   the   filtrate   is   acidified   with
35 sulfuric  acid,   usually at  a hydrogen to  lactate molar
ratio of 1 to 1 to regenerate the lactic acid and
precipitate the calcium as calcium sulfate. The
resulting filtrate consists of about 10% crude lactic
acid which may be concentrated to greater than 50%. It
5 is this "unrefined" lactic acid which can be used in the
process of the present invention to make propylene
glycol. In a preferred process for the conversion of
"unrefined" lactic acid, the reaction is performed with
a 5% ruthenium catalyst at 150°C and 14.5 MPa hydrogen
10 pressure.
Alternatively, the "unrefined" lactic acid can
be further purified in a refining process well known to
those  skilled  in the art which  includes  treating the
"unrefined"     lactic    acid    solution    with    activated
15 vegetable carbon to remove organic impurities,  treating
the solution with ferrocyanide to remove heavy metals, and finally filtrating the solution to remove impurities that have coagulated during purification. The solution may also be passed through ion-exchange resins to remove
2 0 the last traces of contamination.   This further purified
lactic acid may be used as the substrate in the present invention to make propylene glycol.
The present  invention further comprises  the
conversion of lactic acid to propylene glycol as shown
25 above    except    that    the    feedstock    mixture    further
comprises a small quantity of sulfur. In U.S. Patent No. 5,600,028 to Gubitosa and U.S. Patent No. 4,430,253 to Dubeck, which are hereby incorporated herein by reference,  in the hydrogenolysis of sugar over ruthenium
3 0 catalysts   it   was   shown   that   the   addition   of   small
quantities of sulfur partially poisoned the ruthenium
catalyst. While the poisoned catalyst had a lowered
conversion rate, the formation of methane and ethane
byproducts in the reaction was essentially eliminated.
3 5 Therefore,    in   an   embodiment   of   the   process   of   the
present invention that has reduced formation of methane
and ethane byproducts, the lactic acid feedstock further
comprises sulfur which can enhance the selectivity of
the reaction to propylene glycol by reducing lactic acid
5 conversion  to   the  hydrocarbon  byproducts   methane   and
ethane. The decrease in catalyst activity stemming from the addition of the sulfur can be compensated for by preforming the process at slightly higher reaction temperatures or longer space velocities.
10 Figure 1 shows a process for the conversion of
lactic acid to propylene glycol in a continuous flow reactor. The process comprises a reactor 10 which is a three-phase reactor commonly run as a trickle bed that contains a fixed bed of catalyst  12.     Lactic acid is
15 provided as a water solution at 14.    When the lactate is
"unrefined" lactate produced by a microorganism in a fermentor, the "unrefined" lactate is mixed with H2S04 provided at 16 to produce preferably a 1 to 1 molar ratio  of   lactic  acid  to  H+  solution.      The   resulting
20 solution   mixed   at   18   is   pressurized   at   20   and   the
solution feedstream 22 mixed with fresh hydrogen stream 24 before being fed to the reactor 10. Before the hydrogen stream 24 is mixed with the solution feedstream 22, hydrogen at 26 is optionally passed through a pre-
25 saturator 28  and the resulting hydrogen  stream 24  is
pressured by compressor 30. Optionally, the feedstock stream 22 and hydrogen stream 24 are heated in a preheater 32 to the desired temperature and passed downwardly   or   upwardly   through   the   bed   in   intimate

contact with the catalyst 12.    Thus, the gas flow can be
cocurrent or countercurrent to the feedstock flow.
Thus, in a typical process of the present invention, the passivated catalyst 12 is introduced into the reactor 10  and re-reduced with hydrogen  stream 24 without feedstream 22 for about 60 minutes under
pressure of about 1.7 MPa (250 PSIG). A volume of
hydrogen (STP) equal to 50 times the volume of catalyst
is used to ensure complete catalyst reduction.
5 Preferably,   the passivated catalyst is 5% ruthenium on
an inert support such as alumina, titania, silica, aluminosilicate or microporous carbon. After re-reducing the catalyst 12, the reactor 10 is allowed to cool   below  150°C  and  vented  to   low pressure.     Then,
10 feedstream   22   containing   the   lactic   acid   feedstock
solution is introduced in mixture with hydrogen stream 24 into the reactor 10 without opening the reactor. The feedstream rate is adjusted to give a weight hourly space velocity of about 0.5 gram to 5.0 gram lactic acid
15 per gram catalyst per hour.    The hydrogen to lactic acid
feed molar ratio ranges from about 1.4 to 1 to 8.7 to 1, with a preferred value of about 4.2 to 1. The reaction is carried out at a temperature between 50°C to 200°C, preferably between about 70° to 150°C,  and the hydrogen
2 0 pressure is increased to a desired value, usually 3.4 to
16.5 MPa (500 PSIG to 2,400 PSIG), and preferably between 3.4 MPa to 10.3 MPa  (500 to 1,500 PSIG).
After   the   reaction,    the   reactor   effluent stream 34 is cooled to below 40°C in a heat exchanger 36
25 against  a  suitable  fluid  such as  water,   and/or other
process stream that requires heating, e.g., a process hydrogen stream. The cooled effluent stream is passed through a gas-liquid separator 38, wherein the fluid is separated into an overhead gas stream 4 0 and a bottom
30 liquid   stream   42.      The   overhead   stream   40   contains
mainly hydrogen and some methane, and is pressure reduced at 44 and passed to hydrogen purification unit 46. Here the gas is purified to about 90 vol% hydrogen and   recycled   as   stream  48   through   compressor   30   to
-20-
reactor 10.
In    a    typical    separation    for    recovering
propylene glycol   (not shown),  the bottom liquid stream
42 is preheated and passed to a low pressure separator
5 from which overhead water vapor is withdrawn.    Then, the
liquid product is passed through a water distillation column which removes water from the liquid product. Finally, the liquid product from the water column is fed to   a   propylene   glycol   recovery   column   wherein   the
10 propylene glycol is recovered.
In some cases the bottom liquid stream 42 may contain monohydroxide alcohols, particularly when the starting material consists of starting polyols produced in    sugar   conversion   reactions.        Therefore,    in   an
15 optional step,  the bottom liquid stream 42 is passed to
an alcohol separation column at about atmosphere pressure wherein monohydroxide alcohols are removed, if present, before being passed to the low pressure separator.   Also, in some cases the bottom liquid stream
20 42 may need to be demineralized,  particularly when the
feedstock is a fermentate wherein the lactate salt has been converted to lactic acid with sulfuric acid. Therefore, in an optional step, the bottom liquid stream 42    is    demineralized    in   a   demineralizer   containing
25 cationic resins to remove ions  from the bottom liquid
stream 42 before being passed to the low pressure separator or alcohol separation column.
The process of the present invention is also efficient when performed in a batch reactor.    A typical
30 batch reaction setup is  shown in Figure 2.     The batch
reactor 110 consists of a vessel 112 and a lid 114 and a quartz liner 116 to prevent corrosion of the metal surfaces of the reactor 110. The reactor 110 is controlled by a controller 118 which provides adjustable
-21-
st irring speed for speeds preferably between about 10
and 1,000 rpm, electric heater power, and automatic
temperature control to maintain a temperature preferably
between about 80°C and 17ffC. The reactor is further
5 equipped  with  a  valve   120   for  charging  gas   into   the
reactor 110, a gas release valve 122 for releasing pressure, a dip tube 124 connected to valve 122 for adding or withdrawing sample from the reactor 110 under pressure,   a   safety   rupture   disk   for  providing   over
10 pressure   protection   (not   shown),   a   thermocouple   for
temperature measurements (not shown), and a stirrer 12 6 consisting of stirring shaft 128 with a gas entrainment propeller 130. The stirrer 126 is regulated by a stirrer drive system which is a packless magnetic drive.
15 The   hydrogen   is   supplied   to   the   reactor
through gas stream 132 from tank 134. The feedstock solution 136 is supplied to the reactor 110 through feedstock stream 138 from vessel 140. Reaction samples are removed through sample stream 142 from the reactor
20 110 for testing in the sampling apparatus 144.
In a typical batch reaction, the desired amount of passivated catalyst is loaded into the reactor 110. In general, between about 0.1 to 5.0 gram of catalyst   is   loaded  for  every  100  grams   of   feedstock
25 solution,  preferably between about 0.5 to 2.5 gram per
100 ml. Then the reactor 110 is sealed and purged three times with hydrogen to 3.3 MPa (479 MPa), and then filled with hydrogen to 2.0 MPa (290 PSIG). The reactor 110 is then heated to 150°C,  the stirrer 126 is
3 0 set at 50 rpm, and held at that temperature and hydrogen
pressure for one-half hour to complete the reduction of the catalyst. Meanwhile, the feedstock vessel 140, which is connected to the reactor by a feedstream 138, is purged with hydrogen twice,  and then filled with the
-22-
desired amount of lactic acid solution, sealed and pressurized.
After the catalyst has been re-reduced,   the
reactor   110   is   depressurized   and   the   feedstock   is
5 introduced into the reactor 110 via feedstream 13 8.    The
reactor 110 is pressurized to between about 3.4 to 16.5 MPa (500 and 2,400 PSIG), preferably between about 3.4 MPa to 10.3 MPa (500 PSIG to 1,500 PSIG), and changing the   reactor   set   point,    if   required.       The   typical
10 feedstream charge to the reactor is  100 ml of  lactic
acid feedstock. The stirring speed is increased to 1,000 rpm. During the reaction, liquid samples can be removed and analyzed by high performance liquid chromatography  (HPLC),  and gas samples can be analyzed
15 by mass  spectrometry.     The  typical  reaction  can  last
from 2 to 8 hours, with 4 to 6 hours being the most common. After the reaction, the valve 120 is closed and the reactor 110 is allowed to cool to below 4 0°C. Then the reactor 110 is depressurized and purged of hydrogen
20 via exhaust stream 146.    Finally,  the reaction solution
is removed and the propylene glycol is recovered by a propylene glycol recovery column as shown previously.
The following examples are intended to promote a further understanding of the present invention.
25
EXAMPLE 1
The catalysts having the properties of the catalysts shown in Table 1 were prepared according as follows.
30 The    carbon    support    was    first    tested    to
determine its pore volume by adding water to incipient wetness,   and then dried for 5 hours at  100°C in 30 Hg vacuum.    An amount of ruthenium  (III)   chloride hydrate (RuCl3«6H20,   available   from  Aldrich   Chemical   Company)
-23-
sufficient to provide a 5 wt% loading of Ruthenium metal
was then added to a quantity of water just sufficient to
fill the pore volume. Then the carbon support was added
to this solution and the mixture stirred. The
5 impregnated carbon was then dried at 25°C for 16 hours
and then 50°C for 4 hours under vacuum. Reduction was accomplished by heating under 100% H2 in a tube furnace at 0.5°C per minute to 400°C and then holding at 400°C for 8 hours.    The catalyst was then cooled in pure argon to
10 room temperature and then passivated in a 2% oxygen in
argon mixture for one hour. Other catalysts were prepared as shown above with the exception that they were reduced at different gas compositions (H2/He mixtures),  at different temperatures,  and for different
15 reduction  time   lengths;   however,   the  above  method  is
preferred.
To determine the dispersion of ruthenium on the support, hydrogen chemisorption was used. Measured dispersions ranged from 1% to 50%.
20
EXAMPLE 2
This example was performed to determine the efficacy of several of the catalysts that were prepared in   Example   1.       The   prepared   catalysts   that   were
25 evaluated in this  example were  5%  ruthenium on carbon
(Ru/C)   (CG6M-F),   5% Ru/C   (CG-5P-G),   and 5% Ru/C   (CG5P-H) .    The batch reactor process was as follows.
The lactic acid material used for most experiments was an 85 wt% aqueous solution of food-grade
30 lactic acid  (Purac,  Inc.,  Lincolnshire,  Illinois).    The
aqueous solution was diluted to 5 to 3 0% by weight which was then used in the reactions. In some experiments, calcium lactate or potassium lactate was used as the feedstock.       These    feedstocks   were   made   by   adding
35 stoichiometric amounts of Ca(OH)2 or KOH to the lactic
acid material from Purac, Inc. which was then diluted to the appropriate concentration between 5 to 3 0% by weight.
The process used a stirred autoclave reactor
5 (Model 4560, Parr Instrument Company, Moline,  Illinois).
The reactor was mounted in a bench top stand for conducting liquid-phase reactions. The reactor used for this example can withstand a maximum temperature of 350°C and a maximum pressure of 20 MPa.    The reactor was made
10 of T316 stainless steel and the interior surfaces of the
reactor was lined in quartz to further protect the reactor from corrosion. The reactor was equipped with a gas inlet valve for charging gas into the reactor, a gas  release valve for releasing pressure,   a dip tube
15 connected   to   a   liquid   sampling  valve   for   adding   or
withdrawing samples from the reactor when under pressure, a safety rupture disk for providing over pressure protection, a thermocouple for temperature measurements,     and    a    stirring    shaft    with    a    gas
20 entrainment impeller.    The stirrer drive is a packless
magnetic drive. The reactor is controlled by a Model 4852 controller (also from Parr Instrument Company), which provides adjustable stirring speeds, electric heater power, and automatic temperature control.
25 The     liquid    phase     reactant     and    product
concentrations were measured by high performance liquid chromatography (HPLC), and the gas phase products were quantified by mass spectrometry. The HPLC was a Spectra Tech P1000   (available from Thermo Separation Products,
3 0 Fremont,     California)     with    both     ultraviolet     and
refractive index detectors and the columns were Aminex HPX 87H ion exchange types (available from Bio-Rad, Inc., Richmond, California). The mass spectrometer was a    model    M100M    Quadrupole    Residual     Gas    Analyzer
35 (available from Ametek,  Inc.,  Paoli,  Pennsylvania).
Liquid samples were withdrawn during the
reaction and filtered using 0.2 /ш syringe microfilters.
Then the samples were mixed with an internal reference
solution (1 wt% sucrose) and injected into the HPLC.
5 Gas phase compositions were monitored continuously for
mass numbers m/z = 1-50 with the mass spectrometer. Species concentrations were calculated by flowing calibration gases of known composition and comparing mass intensities for each gas sample.
10 Propylene    glycol    yields    as    percent    of
theoretical yield for the catalysts are shown in Figure 3 as a function of time for a reaction temperature of 150°C, 1.5 MPa, and 1% catalyst in solution. The results show that the 5% Ru/C  (CG6M-F)  catalyst was active for
15 propylene glycol formation.    The CG-5P-G and the CG5P-H
catalysts produced one-half the yield of propylene glycol as the CG6M-F catalyst. However, the conversion rates for lactic acid to propylene glycol over the prepared ruthenium catalysts were more rapid than those
20 of the prior art, and the reactions were performed under
milder reaction conditions and at hydrogen pressures as low as 3.3    MPa.

COMPARATIVE EXAMPLE

This   example  was  performed to  evaluate  the
efficacy in a batch reactor of commercially available catalysts that consisted of ruthenium catalysts on several supports and a Raney nickel catalyst.
The commercially available catalysts that were
3 0 evaluated   were    a    5%    Ruthenium   on   aluminum   oxide
(Ru/Al203;   CC4-230   in  Figure  3;   catalog No.   H214  XR/W from   Degussa   Corp.,    Calvert   City,    Kentucky);    a   5% Ruthenium on carbon (Ru/carbon) with a dispersion of 34% (12046   in   Figure   3;   catalog  No.   1940C   from  precious
35 Metals Corp.,  Sevierville, Tennessee); a 3% Ruthenium on
titanium oxide (Ru/Ti02) with a dispersion of 40% (CC4-
226 in Figure 3; catalog No. H7701 X/D from Degussa
Corp.); and, a Raney Nickel (59968 in Figure 3; catalog
No. A-7063 from Activated Metals & Chemicals, Inc.,
5 Sevierville,  Tennessee).
The results are shown in Figure 3 and show that the 5% Ru/carbon catalyst (12 046) gave results that were comparable to the prepared catalyst 5% Ru/C (CG6M-F)  catalyst whereas the 5% Ru/A1203  (CC4-230)  3% Ru/Ti02
10 (CC4-226), and Raney Ni  (59968)  catalysts showed limited
propylene glycol yield. Other catalysts such as palladium on a carbon support, nickel on alumina, and bulk copper chromite showed essentially no activity for propylene glycol formation from lactic acid (results not
15 shown).
Figure 4 shows the propylene glycol yield after 5 hours in the reactor for commercial Ruthenium catalyst 1940C (12046 in Figure 3) as a function of temperature from 13 0 to 170°C and hydrogen pressure from
20 3.3  to 13.2 MPa.     The results  show that high hydrogen
pressure and moderate temperature was most favorable for propylene glycol formation. By comparison, using the prepared CG6M-F catalyst, a maximum propylene glycol yield of 91% was obtained after 5 hours at a temperature
25 of 150°C and 13.5 MPa.    The primary reaction by-product
was methane. In a separate set of experiments, the rate of reactor stirring had no influence on the conversion of lactic acid to propylene glycol.
Figure 5 shows the effect of initial lactic
30 acid concentration in the feedstock over the range of 5%
by weight in solution to 3 0% by weight in solution. The reactions shown in Figure 5 were conducted with 5% Ruthenium catalyst (1940C) at 150°C and 14.5 MPa hydrogen pressure at  a  loading of  1 gram catalyst per
35 100 grams solution.    The fractional conversion rate was
lower as the feedstock concentration was increased; however, the absolute rate of conversion increased as the lactic acid concentration was increased in the feedstock.
5 This example demonstrates that the process can
effectively convert lactic acid to propylene glycol at different lactic acid concentrations.

EXAMPLE 3
10 This example shows an important advantage of
the present invention which is the conversion of "unrefined" lactic acid to propylene glycol. Lactic acid made by a microorganism in a fermentor exits the fermentor as a salt,  usually calcium lactate,  and must
15 be  acidified prior  to  further use   (usually by  adding
excess sulfuric acid.
Impure lactic acid is either a fermentate or lactic acid containing lactate salts. Impure lactic acid   preferably   contains   50   to   100%   lactic   acid.
20 In fermentation,   lactic acid is produced as
lactate salt, which may be partially acidified to produce a lactic acid and lactate salt mixture. This mixture can be treated by the process of the present invention which provides  an economic  source of  lactic
25 acid.    Residual lactate can be further acidified and/or
can be recycled into the crude broth. Lactate salt can be in the alkali metal, alkali earth, or ammonium lactate form as a fermentate. The lactic fermentate can be partially or fully acidulated.
3 0 Hydrogenation    reactions    performed    as    in
Example 2 showed that the conversion of calcium lactate or potassium lactate to propylene glycol was unsuccessful, and, in fact, the reaction was thermodynamically  unfavorable.      However,   as   shown   in
35 Figure   6,    when   sulfuric   acid   was   added   to   calcium
lactate in a 1 to 1 molar ratio before performing the
hydrogenation reaction, high yields of propylene glycol
were obtained which were comparable to the conversion of
neat lactic acid to propylene glycol. The reaction
5 conditions were as  in Example 2  at  150°C and 14.5 MPa
hydrogen pressure.
This result shows that the present invention
enables  the  conversion of   "unrefined"   lactic  acid  to
propylene glycol which is in contrast to the prior art
10 wherein high purity lactic acid is to be the starting
material.

EXAMPLE 4
This  example  shows  the preferred method for
15 preparing the catalysts of the present invention.    These
catalysts have the properties of the catalysts shown in Table 1.
The   catalyst   preparation   method   shown   in Example 1 was used for the preparation of the additional
2 0 catalysts  except  that  the drying  times  and  reduction
conditions were slightly different. The newer, improved
catalysts were dried at 25°C for 5 hours and then under
vacuum at 100°C for 12 hours. The improved catalysts
were then placed in a quartz tube reactor and heated
25 from 25°C to 400°C at 0.5°C per minute under a flow of 3 0
ml per minute of a gas consisting of 10 volume percent hydrogen and 90 volume percent helium. Once 400°C was reached, the gas was switched to pure hydrogen and reduction was continued for 16 hours.    The catalyst was
3 0 then cooled under a helium flow to room temperature and
passivated in a stream of 2% oxygen in argon for 1 hour. The support comprising the preferred catalysts are microporous and have a BET surface area between about 400 to 1,000 m2 per gram.    The preferred loading is 4 to
A number of catalysts were prepared which had
the properties of the catalysts shown in Table 1.    These
catalysts are preferred.   These catalysts were evaluated
for dispersion using hydrogen chemisorption  in a  flow
5 cell system.

EXAMPLE 5
This example demonstrates the utility of the process    of    the   present    invention    in   a    continuous
10 trickle-bed reactor.    This example shows the reaction
mediated by catalysts CG5P-A and CG6M-F which were made according to the improved method in Example 4.
The reactor (inner diameter 0.62 inches and 18 inches  in  length)   which was  jacketed and  temperature
15 controlled   with   a   temperature   controller   from   Omega
Engineering, Stanford, Connecticut was charged with 50 to 75 ml of catalyst. The feed gas rate to the reactor was controlled by mass flow controllers and the feed hydrogen was passed through water saturators prior to
20 entering the reaction to avoid flashing of the liquid
feed in the reactor. The liquid feedstock solution was fed via an HPLC pump (available from Bio-Rad, Inc.) . The reactor effluent was passed through a cold-water condenser and into a phase separator wherein the liquid
25 product was allowed to settle out.    The gas effluent was
passed through a back pressure regulator and to the exhaust. A portion of the gas effluent was drawn into a mass spectrometer wherein the effluent gas composition was determined.
30 The     reactions    were     carried     out     at     a
temperature from about 80° to 156 С and at pressures ranging from about 3.4 to 8.3 MPa (500 to 1,200 PSIG). Lactic acid was fed to the reactor in aqueous solutions of   5  to  20%  by weight.     The  liquid  feed  rate  to  the
3 5 reactor ranged from 0.5 to 3.0 ml per minute, which gave
a weight hourly space velocity (kg lactic acid/kg catalyst/hour) of 0.3 to 2.0. The hydrogen to lactic acid feed molar ratio ranged from 2:1 to 10:1.
To   perform   the   trickle-bed   reaction,    the
5 catalyst   was   first   loaded   into   the   reactor  and  then
reduced at the reaction temperature for 1 hour to remove the passivating oxide layer. Once the catalyst had been reduced, water was fed to the reactor while the temperature   was   brought   to   the   desired   steady-state
10 condition.    The liquid was then switched to the lactic
acid containing feedstock solution and the hydrogen pressure was increased to the desired value. It generally took about 90 to 12 0 minutes for the steady-state   product    compositions   to   be   achieved   at   any
15 particular set of conditions.    Thus,   several different
reaction conditions could be evaluated over the course of a day.
A summary of several reactions conducted in a
trickle-bed reactor over two different ruthenium over
20 carbon   (Ru/C)   catalysts   is   shown  in Table  2.     These
catalysts were CG6M-F and CG5P-A with the properties for these catalysts shown in Table 1. In these reactions, the feedstock was 10 wt% lactic acid in water at a flow rate of 1.0 ml per minute.    The quantity of
2 5 catalyst  in the bed was  27-3 0 grams  and the hydrogen
pressure was 8.3 MPa (1,200 PSIG). The temperature ranged from 80°C to 120'C. The results in Table 2 clearly demonstrate that the process of the present invention is    efficient when performed in a continuous
3 0 mode of operation.    In particular, the present invention
shows that the CG6M catalyst is preferred over the CG5P
catalyst.    Reactions n9 and nlO which were both run at
120°C,  showed that the CG6M catalyst had a selectivity of
86% with a conversion of over 95%.
3 5 In contrast,  the CG5P catalyst under the same

conditions had a selectivity of about 82% with a
conversion of about 95%. Table 2 further shows that the
process of the present invention can be performed over
a temperature range between 10 0°C and Г4 0 С and
5 stillproduce yields of propylene glycol greater than 70%
when using the CG6M catalyst and between 12 0°C and 14 0°C to achieve similar with the CG5P catalyst.

Temp (°C) H2:lactic Ratio Lactic Conv. % PG Yield % PG
Selectivity % Yield СЩ (% of LA) Yield СгНв (% of LA) Carbon Balance % recov. exp.
80 3.4 35% 27% 77% 0.1% 9.1% 101.3% n1
8 100 3.4 67% 58% 87% 0.0% 10.6% 101.5% n7
о 120 3.4 82% 72% 88% 0.6% 4.5% 95.1% n8
ysl 80 5.7 31% 24% 77% 0.9% 7.3% 101.3% n2
Calal 100 5.7 66% 58% 88% 0.7% 6.2% 98.9% пб

120 5.7 95% 82% 86% 3.0% 5.5% 95.5% n9
80 9.5 36% 29% 81% 0.2% 6.6% 99.9% rv4
100 9.5 62% 56% 90% 0.9% 7.1% 102.0% n5
120 9.5 98% 84% 86% ~ 4.0% 11.3% 101.3% n10
80 2.5 25% 20% 80% 1.1% 6.3% 102.4% 13-n2
100 2.5 50% 43% 86% 4.3% 3.3% 100.6% 14-П1
CG5P 120 2.5 67% 57% 85% 5.4% 3.9% 99.2% 14-n2

80 5 23% 18% 78% 0.4% 2.8% 98.3% 13-пЗ

100 5 54% 46% 85% 1.7% 7.1% 100.8% 13-л7
'St 120 5 90% 74% 82% 10.1% 8.2% 102.2% 14-n3
PI
09 130 5 93% 72% 77% 14.4% 7.7% 101.1% 14-n6
CO
О 140 5 97% 72% 74% 10.8% 15.6% 101.4% 14-n5

80 7.5 22% 17% 77% 0.6% 3.6% 99.2% 13-n4

MPLE 6
This example shows the effect of changing the
flow rate of the  feedstock solution in a trickle-bed
reactor.     The  reactions were performed using catalyst
5 CG5P  as  in Example  5  except  that  the  temperature was
100°C, the hydrogen pressure was 8.3 MPa (1,200 PSIG), and the hydrogen to lactic acid molar ratio was 4:1.
Figure   7   shows   the   effect   of   changing   the liquid flow rate in the continuous trickle-bed reactor
10 containing the CG5P catalyst.    The reaction conditions
were 100°C, 8.3 MPa (1,200 PSIG) hydrogen, and a 4 to 1 molar ratio of hydrogen to lactic acid. The feed concentration of lactic acid was 10 wt% in water. Figure 7 shows that as the flow rate is increased there
15 is   a   concomitant   decrease   in   both   conversion   and
selectivity at the above chosen conditions. Reactions conducted at 5.5 and 6.9 MPA (800 and 1,000 PSIG) hydrogen pressure were also performed which gave slightly lower but acceptable   conversion of lactic acid
2 0 to propylene glycol and lower selectivity to propylene
glycol.      Reactions   were   not   performed   at   pressures
greater than 8.3 MPa  (1,200 PSIG)  hydrogen; however,   it
is expected that higher pressures would further improve
selectivity to propylene glycol yield.
25 Thus,  when the reactions were performed at a
hydrogen pressure of 8.3 MPa (1,200 PSIG), increasing the flow rate resulted in a decrease in both conversion and selectivity. Additional reactions were conducted at 5.5  to 6.9 MPa   (800  to 1, 000  PSIG)   hydrogen pressure.
3 0 In these reactions, there was a slightly lowerconversion
of lactic acid and lower selectivity to propylene glycol (data not shown).
This  example  shows  the  efficacy of  the CG5P
catalyst in the conversion of lactic acid to propylene
glycol at a feed concentration of 17.2 wt% compared to
5 the catalyst's  efficacy at a  feed concentration of  10
wt%.
The reactions were performed in a trickle-bed reactor using the conditions similar to those shown in Example   5   with   the   following   noted.       The   reaction
10 temperature was 100°C, the hydrogen pressure was 8.3 MPa
(1,200 PSIG), and the H2 to lactic acid ratio was 5 to 1. The results are shown in Table 3. It is noteworthy that these two reactions in Table 3, while conducted at the same overall flow rate,  have different space velocities
15 because of  the higher  lactic  acid concentration.     The
drop off in selectivity in going to the higher concentration corresponds to the decrease in selectivity in going to a higher flow rate, as seen in Figure 7. Thus,   at  the  same WHSV,   the higher feed concentration
2 0 gives the same performance.

EXAMPLE 8
This example compares the efficacy of the
catalysts having the properties of the catalysts shown
25 in Table 1 in the conversion of lactic acid to propylene
glycol in a batch reactor.
The reactions were performed as in Example 2 but with the following conditions. The temperature was 150°C and the pressure was  14.2  MPa   (2,100  PSIG).   The
3 0 initial feedstock solution was 100 grams which contained
5  grams  of  lactic  acid,   and  1  gram of  catalyst.     The
catalysts were those shown in Table 1. The results shown
in   Figure   8   were   from   measurements   taken   after   a
reaction time of 5 hours.
35 Figure   8   shows   that   the   CG6M-F,   SG6-D,   and
TiP25-C19-J  catalysts   (runs   3,   5,   and  8   of   Table   1,
respectively)    gave   propylene    glycol    yields    87%    or
greater  with   selectivities   96%   or  greater.      Clearly
these catalysts are preferred.     Catalysts  CG5P-N01-I,
5 CG5P-A,    and   ALg-E    (runs   6,    7,    and   9   of   Table   1,
respectively) were effective as well, but not as effective as the preferred catalysts. Thus, catalysts 3 through 9 were very effective in the conversion of lactic acid to propylene glycol with high selectivity.
10 As    expected,     the    efficacy    of     catalyst     10    was
substantially less. In all the reactions, some lactic acid was converted to the gaseous products methane and ethane. However, the overall carbon material balance in the reactions closed to within ±4%, which indicated that
15 the reaction results were reliable.

EXAMPLE 9
This example shows the preparation of preferred catalyst CG6M-F.