MM3(96) conformational analysis of D-glucaramide and x-ray crystal structures of three D-glucaric acid derivatives-models for synthetic poly(alkylene D-glucaramides)

Journal of Carbohydrate Chemistry

ISSN: 0732-8303 (Print) 1532-2327 (Online) Journal homepage: http://www.tandfonline.com/loi/lcar20

MM3(96) CONFORMATIONAL ANALYSIS

OF d-GLUCARAMIDE AND X-RAY CRYSTAL

STRUCTURES OF THREE d-GLUCARIC ACID

DERIVATIVES—MODELS FOR SYNTHETIC

POLY(ALKYLENE d-GLUCARAMIDES)

Susan D. Styron , Alfred D. French , Joyce D. Friedrich , Charles H. Lake &

Donald E. Kiely

To cite this article: Susan D. Styron , Alfred D. French , Joyce D. Friedrich , Charles H. Lake &

Donald E. Kiely (2002) MM3(96) CONFORMATIONAL ANALYSIS OF d-GLUCARAMIDE AND X-RAY

CRYSTAL STRUCTURES OF THREE d-GLUCARIC ACID DERIVATIVES—MODELS FOR SYNTHETIC

POLY(ALKYLENE d-GLUCARAMIDES), Journal of Carbohydrate Chemistry, 21:1-2, 27-51, DOI:

10.1081/CAR-120003736

To link to this article: http://dx.doi.org/10.1081/CAR-120003736

Published online: 28 Oct 2011.

Submit your article to this journal

Article views: 62

View related articles

Citing articles: 20 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=lcar20

Download by: [McMaster University]

Date: 26 August 2016, At: 01:03

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

J. CARBOHYDRATE CHEMISTRY, 21(1&2), 27–51 (2002)

MM3(96) CONFORMATIONAL ANALYSIS OF

D-GLUCARAMIDE AND X-RAY CRYSTAL

STRUCTURES OF THREE D-GLUCARIC ACID

DERIVATIVES—MODELS FOR SYNTHETIC

POLY(ALKYLENE D-GLUCARAMIDES)

Susan D. Styron,¹Alfred D. French,²Joyce D. Friedrich,¹

Charles H. Lake,³and Donald E. Kiely ^4,

*

¹Department of Chemistry, The University of Alabama at

Birmingham, Birmingham, Alabama 35294, USA

²Southern Regional Research Center, U.S. Department of

Agriculture, P.O. Box 19687, New Orleans, Los Angeles 70179, USA

³Department of Chemistry, Indiana University of Pennsylvania,

Indiana, Pennsylvania 15705-1087, USA

⁴Shafizadeh Rocky Mountain Center for Wood and Carbohydrate

Chemistry, The University of Montana, Missoula, Montana 59812, USA

ABSTRACT

An exhaustive conformational analysis of ^D-glucaramide was carried out

using MM3(96) [MM3(96). Molecular Mechanics Software used with per-

mission from N.L. Allinger; University of Georgia]. Nine torsion angles were

each driven in increments of 120ꢀ, generating 19,683 starting conformations.

Each conformation was then fully energy-minimized using MM3’s block

diagonal/full matrix optimization option at dielectric constants of both 3.5

and 6.5. Conformer populations were calculated based on the modeling re-

1

sults and calculated theoretical average H vicinal coupling constants were

compared to experimental values obtained in D₂O solution. Crystal structures

* Corresponding author. Fax: 406-243-6166; E-mail: jccumt@selway.umt.edu

27

Copyright D 2002 by Marcel Dekker, Inc.

www.dekker.com

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

28

STYRON ET AL.

of three acyclic D-glucaric acid derivatives (N,N⁰-dimethyl-D-glucaramide,

dipotassium ^D-glucarate, and sodium potassium D-glucarate) are reported.

These structures and that of previously reported monopotassium glucarate

correspond closely with model conformations that were within one kcal/mol

of the global minimum.

INTRODUCTION

Previously we used ^D-glucaric acid as an available chiral diacid monomer in

polycondensation reactions with primary alkylenediamines to make so-called ‘‘hydroxy-

lated nylons’’, or poly(alkylene ^D-glucaramides).^[2–5]D-Glucaric acid [(2R,3S,4S,5S)-

tetrahydroxyhexanedioic acid] introduces four chiral centers into each diacid residue

incorporated into a polyamide chain. We have taken advantage of the asymmetry of this

monomer to synthesize examples of three classes of poly(alkylene ^D-glucaramides):

‘‘randomly aligned’’ polyamides^[2,3]and two classes of stereoregular polymers,

repeating head, tail–polyamides^[4]and alternating head, tail–tail, head–polyamides^[6]

(Figure 1). Our interest in studying the conformational characteristics of the glucaric acid

component of these polymers stems from an effort to better understand the properties and

potential applications of the resultant polyamides. To this end, we report here an

exhaustive molecular mechanics investigation of ^D-glucaramide and x-ray crystal struc-

tures of various ^D-glucaric acid derivatives as model compounds for the repeating

glucaryl acid unit of these polymers.

D-Glucaramide is a good model compound because it incorporates both the chiral

glucaryl moiety of the polymer as well as the characteristic amide bonds, and its rela-

tively small size allows for computation of the important conformational attributes of the

polymers within a reasonable amount of computer time. The alkylene portion of the

diamine component was not included in the calculations, as these aliphatic diamine units

O

H

N

O

OH

H

N

)

(

)

(

x

(CHOH)₄

N

H

N

H

n

OH

O

"Random" poly(alkylene

D

-glucaramides)

Stereoregular repeating Head, tail -

poly(alkylene -glucaramides)

D

O

OH

H

O

OH

H

N

)

(

x

N

H

x N

n

OH

O

OH

O

H

Stereoregular alternating Head, tail - tail, head - poly(alkylene

D-glucaramides)

x= 1,2,3,4,5, or 6

Figure 1. Three classes of poly(alkylene D-glucaramides).

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

29

should adopt a conformation similar to polyethylene, with an extended or ‘‘zig-zag’’

conformation being dominant.^[7]

Literature data suggest that an unfavorable steric interaction results from hydroxyl

groups that are in a 1,3-parallel arrangement in acyclic carbohydrates, and that such

compounds typically undergo a 120ꢀ rotation about an appropriate C-–C bond to alleviate

this interaction, producing bent or ‘‘sickle’’ conformations.^[8]These interactions are

similar to the destabilizing syn-diaxial interactions (1.9 kcal/mol) between 1,3-diaxial

hydroxyl groups in chair conformations of substituted cyclohexanes.^[9]This phenomenon

accounted for results obtained previously in this laboratory from the molecular

modeling of N,N ⁰-dimethylxylaramide and N,N ⁰-dihexylxylaramide using MacroModel

V2.0.^[10–12]For both compounds, it was found that two sickle conformations were

lower in steric energy than the extended conformation. Presumably, the reduction in

steric energy observed for the two sickle conformations with respect to the extended

conformation is in part due to the alleviation of the 1,3-interaction between the hydroxyls

at C2 and C4 in the sickle conformations. Other studies on acetylated carbohydrate

derivatives also point to the importance of 1,3-interactions in determining preferred con-

1

formations. Sweeting et al.^[13]utilized H NMR vicinal coupling constant data as a

means of studying the conformational characteristics of six peracetylated hexononitriles.

They showed that an extended conformation predominated for their hexononitriles unless

there was a 1,3-parallel interaction present. For penta-O-acetyl-^D-glucononitrile, in which

the acetyl groups on C2 and C4 are eclipsed in the extended conformation, it was found

that a sickle conformation was adopted preferentially. Angyal and co-workers^[8,14]also

conducted ¹H NMR conformational studies on various derivatives of glucose. In

particular, they found that hexa-O-acetyl-^D-glucitol prefers a sickle conformation over

an extended conformation. Several other reports^[15,16]also highlighted the role of

1,3-interactions in determining acyclic carbohydrate conformations, and it was expected

that this phenomenon might contribute to the distribution of conformations observed in

the present modeling study.

These molecules are fairly complex subjects for conformational analysis. Further-

more, the particular atomic sequences are not widely studied. It seemed reasonable to use

an empirical force field (molecular mechanics) method because of the sheer number of

conformations that must be considered. Because of the functional groups present and the

absence of ring forms, we chose to use MM3(96) which has been seen to work well for

mono- and disaccharide molecules.^[17]This choice has some consequences. For

example, hydrogen bonding in general purpose force fields has some limitations. Even

MM3, which uses a complex, angle-dependent Buckingham potential plus dipole–dipole

interactions, does not incorporate a dependence of the energy on the orientation of the

acceptor hydroxyl group. Nor are the geometric consequences of donor-acceptor-donor-

acceptor chains included.

Although MM3 does not provide methods for dealing with solvent, either explicitly

in a molecular dynamics sense or implicitly with a continuum method, it does allow

changing the dielectric constant and thus altering the strength of hydrogen bonds. For

comparison of observed and calculated results for condensed phase systems, it is

necessary to use an increased dielectric constant compared to the value deemed

appropriate for an isolated molecule. This necessity can be explained in several ways,

including the bulk dielectric of the phase. Importantly, the fact that hydrogen bonds can

(and often do) form with solvent molecules reduces the influence of intramolecular

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

30

STYRON ET AL.

hydrogen bonds on the experimental conformations. This effect can be partly mimicked

by reducing the strength of intramolecular hydrogen bonds. Another reason is that the

formation of a hydrogen bond involves a loss of entropy. The experimental structures

result from their free energies, including entropy, but energy minimization studies based

on steric energy do not. MM3 was able to calculate total zero point vibrational energies

that were similar to HF/6-31G* values for glucose, but it is unknown how comparable

these values are to solution values.

Recently^[18]it was shown that the observed crystal structures are well predicted at

dielectric constants of either 3.5 or 7.5. A value of 1.5 was clearly inferior, while the

population vs. energy lines for the 3.5 and 7.5 calculations approximated a Boltzmann

distribution. The similarity of these two results suggested that a higher dielectric constant

would not help to explain the distribution of crystal structures. Values of 3.5 and 6.5 were

chosen in the present work.

Other limitations include the lack of provision for specific solvent effects during

conformational analysis in the MM3 code. These limitations of MM3 were accepted

because of the positive experiences with carbohydrates and because of the large number

of structures that required examination.

There are several ways that the conformations of these compounds might be

explored. There are nine torsional angles that allow rotation including five in the back-

bone and four for the hydroxyl groups. In isolation, these torsion angles would be

expected to each take one of the three staggered conformations, resulting in 3⁹distinct

conformations. Energy minimization of this number (19,683) of structures is within the

capabilities of even a (now obsolete) personal computer. We preferred the exhaustive

examination of conformations, since it was reasonably inexpensive, to the use of more

sophisticated routines that randomly search for stable conformers until no new low-

energy forms are found.^[19]Such routines, or a molecular dynamics simulation, would

become a practical necessity for molecules that are just a little larger.

An obstacle that was encountered at the beginning of the study was the absence of

MM3(96) torsional parameters for the N-–C-–C-–O sequence in the a-hydroxyamide

À

functionality (connectivity of N-–C( O)-–C-–OH). We undertook ab initio calculations

À

on a simpler a-hydroxyamide, glycolic amide, in order to obtain a reasonable estimate of

the needed torsional constants, and applied ab initio results from HF/6-31G* level

calculations to generate the needed torsional parameters^[20]that were subsequently used

throughout this study.

RESULTS AND DISCUSSION

Ab Initio Calculations on Glycolic Amide

Energies for glycolic amide were calculated at 20ꢀ increments of rotation about the

C-–C bond at the HF6-31G* level of ab initio theory (GAUSSIAN94^[21]), along with

comparable MM3 calculations with values of V1 =V2 = V3 =0 for the torsion angle in

question. These data were then sent to N. L. Allinger’s group at the Computational Center

for Molecular Structure and Design at the University of Georgia, where the following

parameters were estimated using their TorsFind program: V₁=À2.157, V₂= À0.592, and

V₃= 0.466.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

31

23

25

H

14

21

17

O

15

H

O

H

O

10

12

19

H

C

6

C

N

8

5

C

3

C

1

C

7

N

4

2

H

22

O

H

O

18 16

11

13

H

O

9

20

H

24

26

Figure 2. Numbering scheme for model compound.

Conformational Analysis of D-Glucaramide

Nine torsion angles of ^D-glucaramide were varied in the conformational analysis.

Five of the angles correspond to rotation about the glucaramide backbone, while the

other four account for hydroxyl rotamers. The numbering scheme for the model

compound is shown in Figure 2, and the varied torsion angles are listed in Table 1.

Torsion angles were varied in increments of 120ꢀ, starting with À180ꢀ, to give the three

staggered orientations for each angle. Varying these nine angles between three values

resulted in 3⁹(19,683) starting conformations. Eighty-one separate input files with 243

conformations each were generated using the torsional driver routine of MM3(96)^[1]

(À3 option, no minimization). Using the separate input files kept the output files from

becoming too large to process the resulting data under MS-DOS. Each of the 81 input

files was next submitted to the block diagonal minimization/full matrix optimization

routine of MM3(96) using the newly derived torsional parameters and a dielectric

constant (e) of 3.5. This value of e is higher than MM3’s default value of 1.5 (which

estimates conformational behavior in a vacuum),^[22]and was used in order to probe the

condensed-phase conformational characteristics of the model compound.^[17]The full

matrix minimization method calculates both a steric energy term and a value for the free

energy which takes molecular vibration into account.

The energies of the resulting 19,683 minimized conformations were extracted from

the output files using a local QBASIC program. They were sorted in order of lowest final

Table 1. Torsion Angles Varied in MM3(96) Studies

on D-glucaramide

Backbone

Hydroxyl Rotamers

O14-–C6-–C5-–H18

H15-–C2-–C1-–O9

H15-–C2-–C3-–H16

H16-–C3-–C4-–H17

H17-–C4-–C5-–H18

H23-–O10-–C2-–H15

H24-–O11-–C3-–H16

H25-–O12-–C4-–H17

H26-–O13-–C5-–H18

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

32

STYRON ET AL.

steric energy to highest. A histogram was generated showing the number of starting

conformations (Nc) that minimized to each particular final energy. Conformations with

energies varying by 0.0001 kcal/mol were considered to be equivalent.

The lowest-energy structures (ranging from the minimum observed energy to the

minimum +1 kcal) were then used to generate conformer populations for each, using the

following equations (10):

Na=No ¼ e^ÀÁE=RT

ð1Þ

ð2Þ

Pa ¼ ½ðNa=NoÞ=ÆðNi=NoÞ Â 100

Na/No is the molar ratio of some rotamer a to the most stable rotamer o, and the

energy difference between the two rotamers is DE. Pa is the population (in percent) of

any rotamer a among a total of i rotamers. Two sets of populations were calculated

and presented herein; one set is based on the relative final steric energies of the low-

energy conformations, and the other is based on the relative free energies of the low-

energy conformations.

The vicinal coupling constants for H15–H16, H16–H17, and H17–H18 were cal-

culated for each low-energy conformation using the program ‘‘Vicinal Coupling Con-

stants for Microcomputers’’ (Quantum Chemistry Program Exchange), which employs an

empirical generalization of the Karplus equation.^[23,24]It takes into account the

electronegativities of substituents on each carbon, as well as the configuration of those

3

substituents. Theoretical average coupling constants for J_15,16, J_16,17, and J_17,18were

calculated based on the equation:^[10]

Jcalcd ¼ ÆXi Â Ji

where Xi is the percent population (Pa) of each rotamer and Ji is the corresponding

calculated coupling constant for that particular rotamer.

The entire conformational analysis experiment was repeated at a dielectric constant

of 6.5 in order to emulate a more polar environment, and the data were worked up in the

same manner as that employed for the calculations done at e= 3.5. Results were

compared between data sets at the two different dielectric constants as well as with

1

data from H NMR in D₂O and a number of crystal structures of acyclic D-glucaric

acid derivatives.

Results from Conformational Analysis at Dielectric Constant of 3.5

Figure 3 summarizes the computational results. The number of initial structures that

minimized to each observed final steric energy is depicted. A total of 2085 distinct

energies resulted from minimization of the initial 19,683 starting conformations. The

energies ranged from 3.89 kcal/mol for the global minimum to the highest minimum of

28.33 kcal/mol. Both the potential energy (final steric) and the free energy were retained

for each structure. The free energy is based on both the potential energy and the

vibrational entropy at room temperature.

Only ten different conformations were found within the range of the minimum

energy and minimum energy +1 kcal/mol. These ten conformations are shown in Figure 4

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

33

Figure 3. Histogram showing the number of initial structures (Nc) that minimized to each

observed final steric energy at e = 3.5.

with the appropriate rotational labels,^[25]and their corresponding energies and percent

of population for each are given in Table 2. While several of the low energy

conformations are those in which one or more sets of hydroxyl groups are eclipsed

in a 1,3-parallel interaction, 45–50% of the population as a whole (depending on

whether DE or DG was used to calculate conformer populations) accounts for

conformations in which this interaction is alleviated. The global minimum (labeled

conformation 1a) and the related conformation 3a (which differs from 1a only in the

orientation of the hydroxyl group at C3) are indeed bent structures that have no 1,3-

parallel interactions. However, conformation 2a and related conformation 4a (again, the

2a

S (₂G^-,₄G^-)

1a

4.05 kcal/mol

DS (₃G⁺,₄G^-)

3.89 kcal/mol

3a

DS (₃G⁺,₄G^-)

4.12 kcal/mol

5a

4a

S (₄G^-)

S (₂G^-,₄G^-)

4.56 kcal/mol

4.54 kcal/mol

Figure 4. Ten lowest energy structures from molecular modeling at e = 3.5.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

34

STYRON ET AL.

7a

6a

S(4G-)

4.72 kcal/mol

S(4G-)

4.72 kcal/mol

8a

DS(₃G⁺,₄G+)

4.84 kcal/mol

10a

9a

S(4G-)

P

4.87 kcal/mol

4.86 kcal/mol

Figure 4. Continued.

two differ only in the orientation of the C-3 hydroxyl) have a 1,3-parallel interaction

between the hydroxyl groups at C3 and C5. Even more interesting is the presence of

two sets of hydroxyl groups in a 1,3-relationship in related conformations 5a and 10a

(differing in the orientation of the hydroxyl groups at C4), and conformations 6a and

7a. Conformation 8a is unique, with no 1,3-interactions. Conformation 9a is fully

extended and has the hydroxyl groups at C2 and C4 eclipsed. These calculations

suggest that conformations having 1,3-interactions are stabilized to some extent by

intramolecular hydrogen bonding between alternating hydroxyl groups.

Table 2. Calculated Energies and Percent Populations for Ten Lowest Energy Structures (e = 3.5)

Steric Energy

(kcal/mol)

Free Energy

(kcal/mol)

a

b

Conformation

%

1a

2a

3a

4a

5a

6a

7a

8a

9a

10a

3.8865

4.0488

4.1214

4.5372

4.5579

4.7216

4.7223

4.8383

4.8583

4.8743

104.545

104.292

104.919

105.783

106.347

105.877

105.878

104.914

106.141

106.060

24.0

18.3

16.2

8.0

24.2

37.1

12.9

3.0

7.7

5.9

1.2

2.6

5.9

4.8

2.6

13.0

1.6

4.7

4.5

1.8

^aCalculations based on relative steric energies.

^bCalculations based on relative free energies.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

35

Figure 5. Histogram showing the number of initial structures (Nc) that minimized to each

observed final steric energy at e = 6.5.

Results from Conformational Analysis at Dielectric Constant of 6.5

Figure 5 shows the number of initial structures at e = 6.5 that minimized to each

observed final steric energy. A total of 3335 distinct energies were generated from the

initial 19,683 starting conformations. The energies ranged from 8.91 kcal/mol for the

global minimum to the highest minimum of 41.80 kcal/mol.

Thirty-five different conformations were found within the range of the minimum

energy and minimum energy +1 kcal/mol. Many of these conformations were very

1b

DS(₃G⁺,₄G⁺)

2b

DS(₂G^-,₄G^-)

4b

3b

S(₄G^-)

DS(₃G^-,₄G⁺)

Figure 6. Four groupings of low-energy conformations from modeling at e = 6.5.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

36

STYRON ET AL.

Table 3. Percent of Population Occupied by Each Grouping of Conformations (e = 6.5)

Grouping

Percent of Population^a

Percent of Population^b

1b

2b

3b

4b

44.4

27.9

14.4

13.3

16.2

43.9

36.0

3.9

^aCalculations based on relative final steric energies.

^bCalculations based on relative free energies.

similar, so for simplicity they were grouped together into four general conformational

families. Variations within the families were primarily due to hydroxyl rotamers.

General structures with appropriate rotational labels are shown for the four groupings in

Figure 6, and the corresponding percent of population occupied by each family is given

in Table 3. The global minimum fell into Group 1b. Both groups 1b and 3b have no

1,3-parallel interactions, whereas group 2b has one set of hydroxyl groups in a 1,3-

relationship and group 4b has both sets of hydroxyl groups eclipsed. Together groups

1b and 3b account for 52–59% of the population. It appears that at the higher di-

electric constant, the 1,3-parallel interaction is slightly more destabilizing and therefore

less favored. Thus, the intramolecular hydrogen bonding contribution between alter-

nating hydroxyl groups is somewhat reduced but not eliminated in a simulated more

polar environment.

Factors That Influence the Calculated Conformational Distribution of

D-Glucaramide-Alleviation of 1,3-Parallel Dihydroxy Interactions

and Stabilization by Intramolecular Hydrogen Bonding

As previously indicated, in the fully extended conformation of ^D-glucaramide, the

hydroxyl groups at C2 and C4 are in a 1,3-parallel relationship. Five of the ten low-

energy structures, including the global minimum from the modeling at e =3.5 (1a–4a

and 8a) correspond to conformations in which this parallel interaction has been

avoided. In addition, groups 2b and 3b from the modeling at e =6.5 are also composed

of conformations without 1,3-parallel interactions between the C2 and C4 hydroxyl

groups. Interestingly, conformations 2a and 4a, as well as group 2b, incorporate instead

a new 1,3-parallel interaction between the hydroxyl groups at C3 and C5. Intramo-

: : :

lecular hydrogen bonding between H-–O H of the two hydroxyl groups appears to

: : :

˚

stabilize these conformations. The approximate calculated O H distance = 2.1 A in all

cases (calculated with Alchemy 2000^[26]software).

Four other conformations resulting from calculations at e= 3.5 (5a, 6a, 7a, and

10a), as well as group 4b from calculations at e= 6.5, also benefit from hydrogen

: : :

bonding between 1,3-parallel hydroxyl groups (again, calculated O H distance=

˚

2.1 A).

An additional source of hydrogen bonding, that between a vicinal amide N-–H and

hydroxyl group, is encountered at both terminal amide bonds in all of the calculated

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

37

: : :

low-energy conformations. The calculated N-–H O-–H distance is approximately 2.1

˚

A. This preference for hydrogen bonding accounts for the relatively small dihedral

angles exhibited for N8-–C1-–C2-–O10 and N7-–C6-–C5-–O13 (generally less than

30ꢀ in all low-energy conformations). This phenomena can be seen in the extended

backbone crystal structures of 11 discussed below. This structure possessed torsion

angles N9-–C1-–C2-–O10 =20.2⁰and N7-–C6-–C5-–O13 = À23.7⁰.

We have also observed some special cases of hydrogen bonding between the

oxygen of a carbonyl group and the hydrogen of a hydroxyl group. One such case

presents itself in conformations 1a and 3a, where an eight-membered ring is

completed by hydrogen bonding between the carbonyl oxygen O14 and the hydrogen

(H23) of the hydroxyl group at C2. This hydrogen bond may be a special stabilizing

factor for these two conformations, compensating somewhat for the 1,3-parallel

interaction introduced between C2 and C6 as a result of the alleviation of the original

1,3-parallel interaction between the C2 and C4 hydroxyl groups. A second case is

encountered in the extended conformation 9a, where a five-membered hydrogen

bonding system is formed between the carbonyl oxygen O14 and the hydrogen

: : :

(H25) of the hydroxyl group at C4. In both of these cases, the H-–O C-–O distance

˚

was approximately 2.0 A.

In effect, the two dominant considerations when evaluating the calculated low-

energy conformations from the molecular modeling of ^D-glucaramide appear to be

1,3-parallel interactions and intramolecular hydrogen bonding. Many literature

examples^{[8,10,13–16]}provide evidence for the destabilizing effects of 1,3-parallel inter-

actions, and we have indeed encountered several low-energy conformations that are

stabilized by the alleviation of such interactions. However, many other calculated low-

energy conformations for ^D-glucaramide have either the C3 and C5 hydroxyl groups in

1,3-parallel orientations, or have both sets of hydroxyl groups (C3 and C5 as well as

C2 and C4) in a 1,3-relationship. In addition, acyclic glucaric acid derivatives 11 and

12 from this study have crystal structures with extended conformations with 1,3-pa-

rallel hydroxyl groups.

Comparison of Calculated and Experimentally Determined

Coupling Constants

The theoretical average vicinal coupling constants based on the two data sets from

the modeling studies are given in Table 4, along with experimentally determined

coupling constants for ^D-glucaramide as well as a number of other amides derived from

D-glucaric acid.

3

Calculated values for J_15,16and J_16,17at both dielectric constants (Table 4) agree

fairly well with observed values for all of the listed glucaramides. However, both

3

calculated values for J_17,18are underestimated. The modeling approach therefore

effectively simulates the average conformational characteristics about the C2-–C3 and

C3-–C4 bonds, but is not as accurate at the C4-–C5 junction.

In general, both theory and experiment confirm that the preferred conformational

relationship between H15 and H16 is gauche with some contribution from anti con-

formations. When populations are calculated based on the final steric energies, the per-

centage of gauche conformations is approximately 75%, while the percentage of anti is

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

38

STYRON ET AL.

Table 4. Comparison of Calculated and Experimental Coupling Constants (Hz)

Compound

³J_15,16

³J_16,17

³J_17,18

D-glucaramide (observed^a)

3.08

2.98

2.87

2.73

3.01

3.94

4.37

4.88

4.83

4.53

4.47

5.56

5.00

5.08

5.16

5.23

2.58

2.16

2.36

2.07

N,N ⁰-dimethyl-D-glucaramide (observed^a)

Head, tail–poly(tetramethylene D-glucaramide) (observed^a)

D-glucaramide (theoretical;^be = 3.5)

D-glucaramide (theoretical;^ce = 6.5)

D-glucaramide (theoretical;^de = 3.5)

D-glucaramide (theoretical;^ee = 6.5)

^{a 1}H NMR, D₂O.

^bCalculated using the set of 10 lowest energy conformations and their relative steric energies.

^cCalculated using the set of 35 lowest energy conformations and their relative steric energies.

^dCalculated using the set of 10 lowest energy conformations and their relative free energies.

^eCalculated using the set of 35 lowest energy conformations and their relative free energies.

25%. Using populations calculated based on the free energies, the percentage of gauche

conformations decreases to approximately 60%, with the anti conformation accounting

for the remaining 40%. About the C3-–C4 bond, however, approximately equal quan-

tities of gauche and anti arrangements between H16 and H17 are predicted by theory

(regardless of whether final steric or free energy is used to calculate the populations).

3

Experimental data are well matched to the theoretical data. The J_16,17value of ap-

proximately 5 Hz is intermediate between usual values observed for either all-gauche or

all-anti conformations, indicating conformational averaging.^[27]

Theoretical J values of approximately 2.0 for vicinal coupling between H17 and H18

support a largely gauche arrangement about the C4-–C5 bond (ꢀ 95% gauche, with ꢁ5%

anti), a prediction which is not borne out by the experimentally observed coupling constants

which are greater than 5.0 for all of the glucaramides listed. Thus, the experimental results

show much more conformational averaging for the C4-–C5 bond than predicted from the

modeling study. The reasons for the discrepancy between theoretical and experimental

coupling constants at the C4-–C5 junction are not clear at this time. It is possible that

Figure 7. The labeling of the atoms in the crystal structure of N,N⁰-dimethyl-D-glucaramide (11).

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

39

Figure 8. Labeling of the atoms of the crystal structure of dipotassium D-glucarate monohydrate

(12). The potassium ions and water molecule are shown in the unit cell for 12, Figure 12.

specific solvent (D₂O) effects account for some discrepancies between experiment and

prediction, and since the relative energies of the various conformers are very close, small

errors in the force field could account for some of the discrepancy.

In addition, it appears that the calculations carried out here employing MM3(96) at

dielectric constants of 3.5 and 6.5 generate conformations that are overly favored by

Figure 9. The labeling of the atoms of the crystal structure of sodium potassium D-glucarate

dihydrate (13). The sodium and potassium ions are shown in the unit cell for 13, Figure 13.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

40

STYRON ET AL.

intramolecular hydrogen bonding. This is manifested by the relationship between

dielectric constant used and calculated values for the dipole–dipole component of the

final steric energies obtained in the conformational analysis. The dipole–dipole con-

tribution to final steric energy varies greatly with dielectric constant. At a low dielectric

constant of 1.0, the dipole–dipole contribution to final steric energy is very high, on the

order of À31 kcal/mol. At higher dielectric constants (3.5 and 6.5), those used in the

calculations described here, the stabilizing dipole–dipole contributions drop to about À8

kcal/mol and À4 kcal/mol, respectively. However, these contributions still contribute

significantly to biasing structures with various stabilizing intermolecular hydrogen bonds

at the expense of destabilizing steric interactions such as those associated with 1,3-

parallel interactions. Thus in conformations 1a–10a, only 9a (the extended structure) has

a large (anti) dihedral angle between H–17, H–18 about the C4-–C5 bond, a result that is

not consistent with the observed H17–H18 coupling of ca. 5.1 Hz for ^D-glucaramide.

Consequently, based on the results presented here, it is suggested that care be exercised in

0

˚

Table 5. Selected Bond Distances (A) and Angles (ꢀ) for N,N -Dimethyl-D-glucaramide (11),

Dipotassium D-Glucarate Monohydrate (12) and Sodium Potassium D-Glucarate Dihydrate (13)

11

12

13

Bond

C1-–C2 = 1.517 (9)

C1-–C2 = 1.533 (3)

C2-–C3 = 1.533 (3)

C3-–C4 = 1.520 (3)

C4-–C5 = 1.548 (3)

C5-–C6 = 1.524 (3)

C1-–O7 = 1.240 (3)

C1-–O8 = 1.264 (3)

C2-–O10 = 1.418 (3)

C3-–O11 = 1.421 (3)

C4-–O12 = 1.426 (3)

C5-–O13 = 1.432 (3)

C6-–O14 = 1.243 (3)

C6-–O15 = 1.260 (3)

O10-–H23 = 1.10 (5)

O11-–H24 = 0.86 (3)

O12-–H25 = 0.89 (3)

O13-–H26 = 1.02 (4)

C2-–H15 = 0.90 (4)

C3-–H16 = 0.96 (4)

C4-–H17 = 0.98 (4)

C5-–H18 = 1.01 (5)

C1-–C2 = 1.534 (3)

C2-–C3 = 1.526 (3)

C3-–C4 = 1.528 (3)

C4-–C5 = 1.535 (4)

C5-–C6 = 1.536 (4)

C1-–O7 = 1.252 (3)

C1-–O8 = 1.249 (3)

C2-–O10 = 1.427 (3)

C3-–O11 = 1.426 (3)

C4-–O12 = 1.440 (3)

C5-–O13 = 1.417 (3)

C6-–O14 = 1.255 (3)

C6-–O15 = 1.251 (3)

O10-–H23 = 0.87 (3)

O11-–H24 = 0.93 (3)

O12-–H25 = 0.76 (3)

O13-–H26 = 0.83 (3)

C2-–H15 = 0.92 (2)

C3-–H16 = 0.91 (2)

C4-–H17 = 0.93 (2)

C5-–H18 = 1.02 (2)

C1-–C2-–C3 = 109.9 (2)

C2-–C3-–C4 = 112.7 (2)

C3-–C4-–C5 = 113.6 (2)

C4-–C5-–C6 = 115.3 (2)

Distances C2-–C3 = 1.541 (8)

C3-–C4 = 1.517 (9)

C4-–C5 = 1.526 (8)

C5-–C6 = 1.544 (9)

C1-–O9 = 1.255 (8)

C1-–N9 = 1.313 (9)

C2-–O10 = 1.422 (7)

C3-–O11 = 1.425 (10)

C4-–O12 = 1.415 (7)

C5-–O13 = 1.406 (7)

C6-–O14 = 1.224 (8)

C6-–N7 = 1.334 (9)

N7-–H20 = 0.98 (7)

N9-–H21 = 0.78 (8)

Bond

Angles

C1-–C2-–C3 = 108.8 (5) C1-–C2-–C3 = 109.9 (2)

C2-–C3-–C4 = 112.9 (5) C2-–C3-–C4 = 113.4 (2)

C3-–C4-–C5 = 111.4 (5) C3-–C4-–C5 = 111.6 (2)

C4-–C5-–C6 = 108.9 (5) C4-–C5-–C6 = 112.0 (2)

H20-–N7-–C6 = 111 (4)

H21-–N9-–C1 = 116 (6)

H23-–O10-–C2 = 113 (2) H23-–O10-–C2 = 105 (2)

H24-–O11-–C3 = 103 (2) H24-–O11-–C3 = 105 (2)

H25-–O12-–C4 = 108 (2) H25-–O12-–C4 = 102 (2)

H26-–O13-–C5 = 105 (5) H26-–O13-–C5 = 107 (2)

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

41

Figure 10. The superimposed structures of N,N⁰-dimethyl-D-glucaramide (line drawing, 11) and

dipotassium D-glucarate monohydrate (dashed drawing, 12).

assigning relative importance to various conformations of acyclic carbohydrates, even at

dielectric constants of 3.5 and higher, due to overemphasis of hydrogen bonding in

structure stabilization.

Discussion of the X-ray Crystal Structures of Three Acyclic

D-Glucaric Acid Derivatives 11, 12 and 13

X-ray diffraction studies were performed on one ^D-glucaric acid diamide, N,N⁰-

dimethyl-^D-glucaramide (11), and two disalts, dipotassium D-glucarate monohydrate (12),

and sodium potassium ^D-glucarate dihydrate (13). The labeling of the atoms of (11), (12)

and (13) is reported in Figures 7, 8, and 9, respectively. Selected bond distances and

angles are reported in Table 5.

Table 6. Comparison of Selected Torsion Angles (ꢀ) for Compounds 11, 12, and Theoretical

Conformation 9a

Torsion Angle

11

12

Conformation 9a

C1-–C2-–C3-–C4

O10-–C2-–C3-–C4

C1-–C2-–C3-–O11

O10-–C2-–C3-–O11

C2-–C3-–C4-–C5

C2-–C3-–C4-–O12

O11-–C3-–C4-–C5

O11-–C3-–C4-–O12

C3-–C4-–C5-–C6

O12-–C4-–C5-–C6

C3-–C4-–C5-–O13

O12-–C4-–C5-–O13

H15-–C2-–C3-–H16

H16-–C3-–C4-–H17

H17-–C4-–C5-–H18

À177.9

59.7

60.4

À170.8

66.5

65.0

À170.8

67.1

66.3

À61.9

179.3

À61.6

À58.1

61.0

À57.7

165.7

À70.4

À72.8

51.1

À55.9

À177.7

À57.9

À54.9

64.9

À179.4

57.8

174.2

50.9

À175.8

64.5

À57.0

À179.8

À60.4

À60.9

À177.4

À63.4

173.3

70.8

À55.3

À175.0

67.5

À79.1

À167.7

À59.2

À173.8

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

42

STYRON ET AL.

Table 7. Comparison of Selected Torsion Angles (ꢀ) for Compound 13

Theoretical Conformation 1a

Torsion angle

13

Conformation 1a

C1-–C2-–C3-–C4

À161.4

72.5

76.1

175.2

53.0

56.8

O10-–C2-–C3-–C4

C1-–C2-–C3-–O11

O10-–C2-–C3-–O11

C2-–C3-–C4-–C5

À50.0

67.9

À65.3

63.8

C2-–C3-–C4-–O12

O11-–C3-–C4-–C5

O11-–C3-–C4-–O12

C3-–C4-–C5-–C6

À173.0

À168.9

À49.8

À73.0

167.8

53.4

À179.4

À177.0

À60.2

À72.1

170.0

51.4

O12-–C4-–C5-–C6

C3-–C4-–C5-–O13

O12-–C4-–C5-–O13

H15-–C2-–C3-–H16

H16-–C3-–C4-–H17

H17-–C4-–C5-–H18

H15-–C2-–O10-–H23

H16-–C3-–O11-–H24

H17-–C4-–O12-–H25

H18-–C5-–O13-–H26

À65.8

67.8

À66.4

55.8

À168.4

À60.1

À21.7

26.7

175.7

À64.7

32.7

45.2

30.8

57.6

À61.6

À59.0

N,N⁰-Dimethyl-D-glucaramide (11) and dipotassium D-glucarate monohydrate (12)

both adopt extended conformations with similar skeletal carbon backbones in the crys-

talline state. Both of these molecules possess intramolecular 1,3-parallel interactions.

Figure 10 provides a superimposed image of both structures which shows the high

degree of overlap. The structure of sodium potassium ^D-glucarate (13) is distinctly

different. It adopts the double-sickle conformation in which there are no 1,3-parallel

interactions. This bent structure is similar to that observed for crystalline monopotas-

sium ^D-glucarate.^[28]

Figure 11. The unit cell of N,N⁰-dimethyl-D-glucaramide (11) as viewed perpendicular to the ac-

plane showing the hydrogen bonding system associated with the crystal structure.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

43

All three of the crystal structures correspond to populated conformations in the

modeling study. Structures 11 and 12 are similar to the extended conformation 9a from

the modeling study at e= 3.5. A comparison of selected torsion angles for 11 and 12 and

the theoretical conformation 9a are given in Table 6. Structure 13 with two sickle

arrangements (₃G⁺, G^À) is very similar to the global minimum (conformation 1a) from

4

the modeling done at e =3.5 as well as conformations in Group 3b from the modeling

done at e =6.5. Selected torsion angles for 13 are compared to the theoretical con-

formation 1a in Table 7.

All three crystal structures have extensive hydrogen bonding systems, and all

four hydroxyl hydrogen atoms are included in the hydrogen bonding system for each

crystal structure. The hydroxyl hydrogen atoms were fixed in the crystal structure of

N,N⁰-dimethyl-D-glucaramide (11) to maximize the intermolecular hydrogen bonding

(Figure 11). This resulted in the strong intermolecular hydrogen bonding interactions

: : :

˚

d(H26 O9) = 1.898 A, d(H25 O12) = 2.094 A, d(H24 O13) = 2.144 A and

: : :

d(H23 O14) =1.895 A. The amide hydrogen atoms were refined and possessed strong

˚

: : : : : :

˚

intramolecular hydrogen bonds d(H21 O10) =2.151 A and d(H20 O13) =2.233 A

˚

: : :

˚

with H21 also forming an intermolecular hydrogen bond d(H21 O9) =2.298 A.

In the extended dipotassium ^D-glucarate monohydrate stucture (12) (Figure 12),

the atoms H23, H24 and H26 form strong intermolecular hydrogen bonds with adja-

Figure 12. The unit cell associated with the crystal structure of dipotassium D-glucarate mono-

hydrate (12) perpendicular to the bc axes showing the hydrogen bonding system associated with the

crystal structure.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

44

STYRON ET AL.

Figure 13. The unit cell associated with the crystal structure of sodium potassium D-glucarate

dihydrate (13) perpendicular to the bc axes showing the hydrogen bonding system associated with

the crystal structure.

: : :

cent molecules [d(H23 O15) =1.790 A, d(H24 O14) = 1.908 A and d(H26 O7) =

: : :

˚

: : :

2.111 A], while H25 forms an intramolecular hydrogen bond (d(H25 O14) =1.872

˚

: : :

A). Hydrogen atom H23 forms a weaker intramolecular hydrogen bond [d(H23 O7)=

˚

2.231 A]. It was also observed that compound 12 crystallized in sheets perpendicular to

the ab plane. These sheets are held together by hydrogen bonding to water molecules.

In the hydrogen bonding system associated with the crystal structure of sodium

potassium ^D-glucarate dihydrate (13) (Figure 13), all four hydroxyl hydrogen atoms ex-

: : :

˚

hibit intermolecular hydrogen bonding [d(H23 O15) =1.821 A, d(H24 O8)= 1.788

: : :

˚

A, d(H25 O10) = 2.266 A and d(H26 O14) = 1.912] but there is no evidence for

stabilizing intramolecular hydrogen bonds for ‘‘bent’’ 13. It is noted that potassium

sodium salt 13 forms ring-like structures, with the water molecules (Figure 13) and

cations located in the cavities.

CONCLUSIONS

1

Theoretical calculations and H NMR solution data show D-glucaramide to have

some degree of conformational flexibility, primarily about the C3-–C4 and C4-–C5

bonds. However, for the most part, the conformational relationship between H15 and

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

45

H16 is gauche. The conformations that are populated at dielectric constants of 3.5 and

6.5 include several sickle and double-sickle conformations, and to a much lesser extent,

extended conformations. In the crystalline state, N,N⁰-D-glucaramide and dipotassium D-

glucarate are fully extended conformations, while crystalline sodium/potassium ^D-glu-

carate has a double-sickle conformation.

Consequently, based on these model experiments, the glucaric acid portion of

poly(alkylene ^D-glucaramides) can in solution be expected to take on a variety of

conformations, mostly those with sickle arrangements which give rise to a bend in the

glucaric acid monomer unit. The inherent stereoregularity of particular polyamides

(head, tail or head, tail–tail, head alignment of the glucaric acid moiety) will also be a

large factor in determining the overall polyamide shape. The computational and x-ray

structure studies described have provided us with some basic information on the

conformational complexity of glucaric acid that should prove useful as we pursue

conformational studies on the poly(^D-glucamides) themselves.

EXPERIMENTAL

1

General Methods. All H NMR spectra were recorded in D₂O at 400.13 MHz.

Chemical shifts are reported as ppm (d) downfield from 3-(trimethylsilyl)propionic-

2,2,3,3-d₄acid, sodium salt (TSP).

D-Glucaramide (1). To a solution of 7N methanolic ammonia (10 mL, 70

mmol) cooled in an ice bath with stirring was added dropwise over a period of 15 min a

solution of methyl ^D-glucarate-6,3-lactone (0.749 g, 3.63 mmol) in distilled methanol

(15 mL). The reaction mixture was stirred in an ice bath for 1.5 h, resulting in the

formation of a white precipitate. The reaction mixture was then removed from the ice

bath, covered with a Kimwipe, and left in a fume hood overnight to allow evaporation of

the excess ammonia. The solid was removed by vacuum filtration, washed with

methanol (37Â10 mL), and dried at reduced pressure for 6 h to give 1 (0.663 g, 3.18

mmol, 87.7%): mp 168–170 ꢀC (lit.^[29]168 ꢀC); ¹H NMR d 4.36 (d, 1H, H–15,

J_15,16= 3.08 Hz); d 4.28 (d, 1H, H–18); d 4.13 (dd, 1H, H–16, J_16,17=4.88 Hz); d 4.00

(t, 1H, H–17, J_17,18=5.16 Hz). X-ray grade crystals could not be obtained despite

multiple recrystallization attempts.

N,N⁰-Dimethyl-D-Glucaramide (11). Methyl D-glucarate-1,4-lactone (0.5034 g,

2.442 mmol) was dissolved in methanol (10 mL). A solution of methylamine (0.5 mL of

40 wt% in water, 180 mg, 5.8 mmol) was added slowly, and the flask was covered with

a septum. A shiny white precipitate was observed almost immediately after the addition

of the methylamine solution. The reaction mixture was stirred at room temperature for 2

h, and the suspended solid was filtered and washed with methanol (3Â5 mL) then

acetone (3Â5 mL) to give 11 (0.442 g, 1.87 mmol, 76.6%): mp 188–191 ꢀC (lit. 188–

190 ꢀC);^{[30] 1}H NMR d 4.31 (d, 1H, H–15, J_15,16=2.98 Hz); d 4.25 (d, 1H, H–18); d

4.10 (dd, 1H, H–16, J_16,17= 4.83 Hz); d 3.96 (t, 1H, H–18, J_17,18=5.16 Hz). X-ray grade

crystals were obtained by redissolving a small amount of the white solid in methanol

with the aid of a heat gun, then cooling the solution to room temperature and storing at

approximately 5 ꢀC for several days.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

46

STYRON ET AL.

Table 8. Details of the Data Collection, Solution and Refinement of the Crystal Structures of

N,N⁰-Dimethyl-D-Glucaramide (11), Dipotassium D-Glucarate Monohydrate (12) and Sodium

Potassium D-Glucarate Dihydrate (13)

11

12

13

Empirical formula

Crystal size (mm)

Crystal system

Space group

C₈H₁₆N₂O₆

0.25Â0.20Â0.20

Monoclinic

P2₁

C₆H₁₀K₂O₉

0.45Â0.40Â0.20

Orthorhombic

P2₁2₁2₁

C₆H₁₂KNaO₁₀

0.35Â0.25Â0.25

Orthorhombic

P2₁2₁2₁

Unit cell

dimensions

˚

a (A)

˚

8.5371 (18)

5.2125 (11)

12.2551 (36)

90

7.8798(7)

8.6524(16)

15.9714(17)

90

7.5348(16)

7.8276(10)

18.9413(29)

90

b (A)

˚

c (A)

a (ꢀ)

b (ꢀ)

c (ꢀ)

106.88 (2)

90

3

˚

V (A )

521.9 (3)

2

236.2

1087.5(5)

4

304.3

1117.2

4

306.2

Z

Formula weight

r_calc.(g/cm³)

1.503

0.129

1.859

0.908

1.821

0.561

abs. coeff (mm^À1

F(000)

)

252

2.0–45.0

w-2q

À9 = h = 9

À5 = k = 5

À13 = l = 13

2720

624

2.0–45.0

w-2q

0 = h = 8

À9 = k = 9

À17 = l = 17

3064

632

2.0–45.0

w-2q

0 = h = 8

À8 = k = 8

À20 = l = 20

3169

2q range (ꢀ)

Scan type

Index ranges

Reflections collected

Independent reflections

1362

1393

1465

(R_merge= 3.88%)

944 (F > 6.0s)

N/A

(R_merge= 1.49%)

1393 (F > 6.0s)

0.0197 (11)

0.5517/0.6019

197

(R_merge= 2.05%)

1358 (F > 6.0s)

0.0064 (4)

0.7965/0.8930

213

Observed reflections

Extinction correction (c)

Min/max transmission

Number of parameters

Final R (observed data)

0.7347/0.9894

153

R = 6.55%

wR = 6.04%

R = 8.66%

wR = 10.97%

0.80

R = 2.11%

wR = 2.48%

R = 2.15%

wR = 3.43%

0.69

R = 2.12%

wR = 2.20%

R = 2.48%

wR = 2.59%

1.11

Final R (all data)

Goodness of fit

Largest D/s

Mean D/s

Data to param ratio

0.009

0.001

0.000

6.5:1

0.71

7.2:1

0.31

6.9:1

0.25

À3

˚

Largest diff. peak (eA_À3

)

˚

Largest diff. hole (eA

)

À0.59

À0.28

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

47

Dipotassium D-GlucarateÁH₂O (12). KOH was added to an aqueous (30 mL)

slurry of monopotassium ^D-glucarate (10.0 g, 40 mmol) until all of the solid had dissolved

(pH > 11). The pale yellow mixture was applied to a Bio-Rad AG¹11 AG (50–100

mesh) ion-retardation column. The product (detected by H₂SO₄char) was contained in

fractions 6–10. The fractions were combined and the colorless solution was concentrated

at reduced pressure. The residue (11.1 g, 39 mmol, 96%) was crystallized from water to

give x-ray grade crystals of 12.

Anal. Calcd for C₆H₈K₂O₈ÁH₂O (386.32): C, 23.68; H, 3.31; K, 25.69. Found: C,

23.86; H, 3.29; K, 25.76.

Sodium Potassium D-GlucarateÁ2H₂O (13). NaOH pellets (ꢂ 0.8 g, 20 mmol)

were added to monopotassium ^D-glucarate (5.0 g, 20 mmol) slurried in deionized water.

Insolubles were removed by filtration and the filtrate was applied to a Bio-Rad AG¹11

AG (50–100 mesh) ion-retardation column (previously washed with 0.5 M 1:1 KCl–

NaCl). Fractions 7–15 (detected by H₂SO₄char) were combined and concentrated to a

clear, colorless syrup. When crystallization did not occur overnight, the syrup was seeded.

Crystallization began immediately. The resulting crystals were washed with 1:1 meth-

anol–water. The latter (4.25 g, 16 mmol, 78%) were then recrystallized from 1:1

methanol–water to give x-ray grade crystals of 13.

Anal. Calcd for C₆H₈KNaO₈Á2H₂O (270.21): C, 23.53; H, 3.95. Found: C, 23.70;

H, 4.02.

Collection of the X-ray Diffraction Data and Solution of the Crystal Structures

for 11, 12, and 13. The crystals were sealed into thin-walled glass capillaries, which

were then mounted and aligned on an Enraf Nonius CAD4 single crystal diffractometer

with k-geometry. The data were collected with graphite-monochromatized Mo Ka

radiation which were subsequently corrected for absorption as well as for Lorentz and

polarization effects. Structures 12 and 13 were corrected for extinction by refining the c

parameter.^[31]There was no evidence of decay in any of the data collections. Details of

the data collection are found in Table 8.

All crystallographic calculations were carried out with the aid of an IBM–PC and

the SHELXTL–PC program package.^[32]The analytical form of the scattering factors

for neutral atoms was used with both of real (Df⁰) and imaginary (Df ⁰⁰) components

of anomalous dispersion included in the calculations.^[33]The positional and aniso-

tropic thermal parameters were refined for all non-hydrogen atoms. Refinement con-

tinued until convergence was reached. Each structure was checked by means of a

final difference —Fourier synthesis. Specific details are collected in Table 7 or are out-

lined in the text.

N, N ⁰-Dimethyl-D-Glucaramide (11). A single crystal (dimensions 0.25 mmÂ

0.20 mmÂ0.20 mm) was selected for the study. Cell parameters and Laue symmetry (2/

m) revealed that the crystal belongs to the monoclinic crystal system. The systematic

absence 0k0 for k =2n +1 define the space group as being either the centrosymmetry

space group P2₁/m (No. 11) or the noncentrosymmetric space group P2₁(No. 4).

Intensity statistics clearly favored the latter case. This was later confirmed by correct

solution and refinement of the crystal structure. 2720 reflections were collected which

were merged into a unique set of 1362 reflections (R_int=3.88%).

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

48

STYRON ET AL.

The structure was solved by direct methods. Hydrogen atoms could not be located

by direct examination of the difference — Fourier maps. Attempts to locate the

hydrogen atoms by using low angle data (q = 1ꢀ to 10ꢀ) failed (the scattering factor

of hydrogen atoms drops off quickly with increasing q). By using only low angle data

the contribution from the hydrogen atoms is maximized). The positional and isotropic

thermal parameters were refined for both amine hydrogen atoms. All hydrogen atoms

bound to carbon atoms were placed at calculated positions in the appropriate staggered

geometry.^[34]The hydroxyl hydrogen atoms were placed in calculated positions which

maximized the intermolecular hydrogen bonding interactions. The positions determined

resulted in torsion angles that closely resembled those for the crystal structure of 12

and the theoretically determined conformation 9a as shown in Table 5. The isotropic

thermal parameter of each hydrogen atom was set equal to the U_eqvalue of the atom to

which it is bound. Absolute conformation was determined due to the crystal belonging

to a noncentrosymmetric space group. This was attempted by the h-refinement pro-

cedure. This was later verified by the overlap of the structure with that of dipotassium

D-glucarate monohydrate (12) (Figure 8).

The refinement converged (largest D/s =0.009) with R =6.55% and wR =6.04%

for those 944 reflections with F₀=6s(jF_oj). The final difference — Fourier map showed

no anoma_Àlo₃us features after refinement, with the residual electron density in the range

À3

˚

À0.59 eA to 0.71 eA

.

Dipotassium D-Glucarate Monohydrate (12). A single crystal (dimensions 0.45

mmÂ0.40 mmÂ0.20 mm) was selected for the study. Cell parameters and Laue

symmetry (2/m 2/m 2/m) revealed that the crystal belongs to the orthorhombic crystal

system. The systematic absences h00 for h =2n+ 1, 0k0 for k =2n +1 and 00l for l =2n+ 1

uniquely define the space group the noncentrosymmetric space group P2₁2₁2₁(No. 19).

3064 reflections were collected which were merged into a unique set of 1420 reflections

(R_int= 1.49%).

The structure was solved by direct methods. All hydrogen atoms were located from

difference—Fourier maps and had their positional and isotropic thermal parameters

refined. Because the compound crystallized in a noncentrosymmetric space group,

absolute conformation needed to be determined. This was accomplished by the h-

refinement procedure with h =0.98(8).

The refinement converged (largest D/s =0.001) with R =2.11% and wR =2.48%

for those 1393 reflections with F₀=6s (jF_oj). The final difference—Fourier map showed

no anoma_Àlo₃us features after refinement, with the residual electron density in the range

À3

˚

À0.28 eA to 0.31 eA

.

Sodium Potassium D-Glucarate Dihydrate (13). A single crystal (dimensions

0.35 mmÂ0.25 mmÂ0.25 mm) was selected for the study. Cell parameters and Laue

symmetry (2/m 2/m 2/m) revealed that the crystal belongs to the orthorhombic crystal

system. The systematic absences h00 for h=2n+1, 0k0 for k =2n+ 1 and 00l for l =2n+ 1

uniquely define the space group the noncentrosymmetric space group P2₁2₁2₁(No. 19).

3169 reflections were collected which were merged into a unique set of 1359 reflections

(R_int= 2.05%).

The structure was solved by direct methods. All hydrogen atoms were located from

difference—Fourier maps and had their positional and isotropic thermal parameters

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

49

refined. Because the compound crystallized in a noncentrosymmetric space group,

absolute conformation needed to be determined. This was accomplished by the h-

refinement procedure with h = 1.11(10).

The refinement converged (largest D/s= 0.000) with R = 2.12% and wR = 2.20%

for those 1358 reflections with F₀=6s (F_o). The final difference—Fourier map showed

no anoma_Àlo₃us features after refinement, with the residual electron density in the range

À3

˚

À0.28 eA to 0.25 eA

.

SUPPLEMENTARY MATERIAL

Details of x-ray crystallographic data for compounds 11, 12, and 13 are included as

supplementary material. Tables of atomic coordinates, bond lengths, and bond angles

have been deposited with the Cambridge Crystallographic Data Centre. These tables may

be obtained on request from the Director, Cambridge Crystallographic Data Centre, 12

Union Road, Cambridge CB2 IEZ, UK.

ACKNOWLEDGMENT

Financial support for this work was provided in part by a Graduate Assistance in

Areas of National Need (GAANN) fellowship to S. Styron.

REFERENCES

1. MM3(96). Molecular Mechanics Software used with permission from N.L. Allin-

ger; University of Georgia.

2. Kiely, D.E.; Chen, L.; Lin, T.-H. Hydroxylated nylons based on unprotected esteri-

fied D-glucaric acid by simple condensation reactions. J. Am. Chem. Soc. 1994,

116, 571–578.

3. Kiely, D.E.; Chen, L.; Lin, T.-H. Synthetic polyhydroxypolamides from galactaric,

xylaric, D-glucaric, and mannaric acids and alkylenediamine monomers—some

comparisons J. Polym. Sci.: Part A: Polym. Chem. 2000, 38, 594–603.

4. Chen, L.; Kiely, D.E. Synthesis of stereoregular head, tail hydroxylated nylons

derived from D-glucose. J. Org. Chem. 1996, 61, 5847–5851.

5. Carter, A.; Morton, D.; Kiely, D.E. Synthesis of some poly(4-alkyl-4-azaheptame-

thylene-D-glucaramides) J. Polym. Sci.: Part A: Polym. Chem. 2000, 38, 3892–3899.

6. Styron, S.D. PhD Dissertation; The University of Alabama: Birmingham, 1999.

7. Carey, F.A.; Sundberg, R.J. Conformational, Steric, and Electronic Effects. In Ad-

vanced Organic Chemistry, Part A: Structure and Mechanisms, 3rd Ed.; Plenum

Press: New York, 1990; 121.

8. Angyal, S.; LeFur, R.; Gagnaire, D. Conformations of acyclic sugar derivatives part

II: determination of the conformations of alditol acetates in solution by the use of

250-MHz NMR spectra. Carbohydr. Res. 1972, 23, 121–134.

9. Eliel, E.L.; Wilen, S.H. Configuration and Conformation of Cyclic Molecules. In

Stereochemistry of Organic Compounds; Wiley: New York, 1994; 707.

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

50

STYRON ET AL.

10. Chen, L.; Haraden, B.; Kane, R.W.; Kiely, D.E.; Rowland, R.S. In Computer Mod-

eling of Carbohydrate Molecules; French, A.D., Brady, J.W., Eds.; ACS Symposium

Series 430; American Chemical Society: Washington, DC, 1991; 141–151.

11. Goldsmith, D.J.; Bowen, J.P.; Qamhiyeh, E.; Still, W.C. Stability relationships of

decalindiones—modified MM2 force field calculations. J. Org. Chem. 1987, 52,

951–953.

12. Chang, G.; Guida, W.C.; Still, W.C. An internal coordinate Monte Carlo method for

searching conformational space. J. Am. Chem. Soc. 1989, 111, 4379–4386.

13. Sweeting, L.; Coxon, B.; Varma, R. Conformational analysis of peracetylated hex-

ononitriles. Carbohydr. Res. 1979, 72, 43–55.

14. Angyal, S.; LeFur, R.; Gagnaire, D. Conformations of acyclic sugar derivatives part

III: 3,4,5,6,7-penta-acetoxy-trans-1-nitro-1-heptenes. Carbohydr. Res. 1972, 23,

135–138.

15. Jeffrey, G.A.; Kim, H.S. Conformations of the alditols. Carbohydr. Res. 1970, 14,

207–216.

16. Blanc-Muesser, M.; Defaye, J.; Horton, D. Conformations in solution of the stereo-

isomeric, peracetylated aldohexose dimethyl acetals and diethyl dithioacetals. Car-

bohydr. Res. 1980, 87, 71–86.

17. French, A.D.; Dowd, M.K. Exploration of disaccharide conformations by molecular

mechanics. J. Mol. Struct. (Theochem.) 1993, 286, 183–201.

18. Barrow, S.E.; Dulles, F.J.; Cramer, C.J.; Truhlar, D.G.; French, A.D. Relative

stability of alternative chair forms and hydroxymethyl conformations of D-gluco-

pyranose. Carbohydr. Res. 1995, 276, 219–251.

19. Goto, H.; Osawa, E. An efficient algorithm for searching low-energy conformers of

cyclic and acyclic molecules. J. Chem. Soc., Perkin Trans. 2 1993, (2), 187–198.

20. MM3 1994 Force Field Manual. Tripos, 1995; 144.

21. Gaussian94 Software. Gaussian: Pittsburgh, Pennsylvania.

22. MM3 1994 Force Field Manual. Tripos, 1995; 137–138 and 177.

23. 3JHHPC: Vicinal Coupling Constants for Microcomputers Software (program num-

ber QCMP025) purchased from Quantum Chemistry Program Exchange: Bloo-

mington, Indiana.

24. Haasnoot, C.A.G.; de Leeuw, F.A.A.M.; Altona, C. The relationship between pro-

ton–proton NMR coupling constants and substituent electronegativities—I. An

empirical generalization of the Karplus equation. Tetrahedron 1980, 36, 2783–

2792.

25. Horton, D.; Wander, J.D. Conformations of acyclic derivatives of sugars XI: con-

formations of the D-aldopentose diethyl and diphenyl dithioacetals in solution. J.

Org. Chem. 1974, 39 (13), 1859–1863.

26. Alchemy 2000 Molecular Modeling Software purchased from Tripos: St. Louis,

Missouri.

27. Friebolin, H.H.H. Coupling Constants and Chemical Structure. In Basic One- and

Two-Dimensional NMR Spectroscopy; VCH Publishers: New York, 1991; 78–79.

28. Taga, T.; Kuroda, Y.; Osaki, K. Interactions of calcium ions with carbohydrates:

x-ray diffraction and NMR spectroscopic studies on the potassium salt and the

calcium salt of D-glucaric acid. Bull. Chem. Soc. Jpn. 1977, 50 (12), 3079–

3083.

29. Kiely, D.E.; Navia, J.L.; Miller, L.A.; Lin, T.-H. A general synthesis of diterminal

MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016

MM3 ANALYSIS OF D-GLUCARAMIDE

51

diaminodideoxyalditols and 1-amino-1-deoxyalditols using trimethylsilyl protecting

groups. J. Carbohydr. Chem. 1986, 5 (2), 183–197.

30. Chen, L.; Kiely, D.E. Computer aided structural studies of poly(alkylene D-Glucar-

amide)s. Polym. Prepr. 1993, 34 (2), 550–551.

31. Siemens SHELXTL–PC Manual; Siemens Analytical Instruments: Madison, Wis-

consin, 1992; Vol. 4.2.

32. Siemens SHELXTL–PC Manual; Siemens Analytical Instruments: Madison, Wis-

consin, 1992; Vol. 4.2, 154–155.

33. International Tables for X-Ray Crystallography; Kynoch Press: Birmingham, Eng-

land, 1974; 148–150.

34. Churchill, M.R. Some comments on carbon–hydrogen and nitrogen–hydrogen

distances assumed in, and determined from, recent x-ray diffraction studies on

inorganic complexes. Inorg. Chem. 1973.

Received May 1, 2001

Accepted December 17, 2001

Article Doi

MM3(96) conformational analysis of D-glucaramide and x-ray crystal structures of three D-glucaric acid derivatives-models for synthetic poly(alkylene D-glucaramides)

DOI: 10.1081/CAR-120003736

Source and publish data:

Authors:

Article abstract of DOI:10.1081/CAR-120003736

Full text of DOI:10.1081/CAR-120003736

Products guided by the article

R&D Labs maybe for 6614-50-2

Relevant to this article

Hot Product