Single-Event Spectroscopy and Unravelling Kinetics of Covalent Domains Based on Cyclobutane Mechanophores

Domains Based on Cyclobutane Mechanophores

Article

Single-Event Spectroscopy and Unravelling Kinetics of Covalent

Brandon H. Bowser, Shu Wang, Tatiana B. Kouznetsova, Haley K. Beech, Bradley D. Olsen,*

Michael Rubinstein,* and Stephen L. Craig*

Cite This: J. Am. Chem. Soc. 2021, 143, 5269 −5276

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: Mechanochemical reactions that lead to an increase in

polymer contour length have the potential to serve as covalent

synthetic mimics of the mechanical unfolding of noncovalent “stored

length” domains in structural proteins. Here we report the force-

dependent kinetics of stored length release in a family of covalent

domain polymers based on cis-1,2-substituted cyclobutane mechano-

phores. The stored length is determined by the size (n) of a fused ring

in an [n.2.0] bicyclic architecture, and it can be made suﬃciently large

(

>3 nm per event) that individual unravelling events are resolved in

both constant-velocity and constant-force single-molecule force

spectroscopy (SMFS) experiments. Replacing a methylene in the

pulling attachment with a phenyl group drops the force necessary to

achieve rate constants of 1 s⁻from ca. 1970 pN (dialkyl handles) to 630 pN (diaryl handles), and the substituent eﬀect is attributed

to a combination of electronic stabilization and mechanical leverage eﬀects. In contrast, the kinetics are negligibly perturbed by

changes in the amount of stored length. The independent control of unravelling force and extension holds promise as a probe of

molecular behavior in polymer networks and for optimizing the behaviors of materials made from covalent domain polymers.

INTRODUCTION

changes in other bulk material properties, such as the fracture

■

toughness of network elastomers. The design, synthesis, and

characterization of CDPs is therefore of growing interest, and

recent years have seen the emergence of covalent stress-

responsive monomers (mechanophores) that unveil stored

length through force-coupled covalent reactions in highly

stretched polymer chains. A promising class of mechanophores

in this regard involves cyclobutanes (Figure 1b) that are fused

Biological systems have a variety of mechanisms for

productively responding to an applied mechanical load. The

muscle protein titin, for example, releases elastic energy via the

−8

unfolding of structural domains,

and the resulting

mechanics couple to the activity of actin and myosin to

build muscle strength in response to repeated load bearing.

The folded domains of titin and similar proteins (Figure 1a)

act as sources of “stored length” that are unveiled when a force

is applied, leading to an extension in the overall length of the

protein. Similar strategies have recently been brought into

12−14,16,17,19−26

to other cyclic, covalent domains.

Stress is

typically applied to one side of the cyclobutane via the polymer

backbone, causing the cyclobutane to ring-open through a [2 +

2] cycloreversion, releasing the length stored in the fused ring

on the other side of the cyclobutane “gate.” Whereas many

mechanophores, such as those based on the ring-opening

,10

synthetic polymers.

Chung et al. designed a biomimetic

polymer in which individual repeats comprise large macro-

cycles that are contracted by hydrogen-bonding interactions

27,28

reactions of gem-dihalocyclopropanes and cyclobutenes,

that span the loop interior. Using single-molecule force

spectroscopy (SMFS), they were able to observe the kinetics of

individual unfolding and refolding events that are responsible

for mechanical property enhancements in bulk materials.

The energy dissipation achieved by domain unfolding can be

greatly enhanced by stored length that is released through the

scission of covalent, rather than noncovalent, interactions

involve moderate (<1 nm) changes in polymer contour length,

the architecture of fused cyclobutanes provides access to

greater length changes that can, in rare examples, be

individually resolved in single molecule force spectroscopy

Received: February 24, 2021

Published: March 30, 2021

2−17

(

Figure 1a).

Because covalent bonds require higher forces

for rupture, these “covalent domain polymers” (CDPs) are

capable of single-strand toughness that far exceeds that of

conventional synthetic polymers. Within a polymer network,

increased single-chain toughness could also contribute to

2021 American Chemical Society

Journal of the American Chemical Society

J. Am. Chem. Soc. 2021, 143, 5269−5276

269

Article

transition state in the cycloreversion of CB4 as a key

determinant of the relative ease with which its network is

torn. Together, these growing uses of cyclobutane mechano-

phores in polymer networks motivate a more detailed

characterization of their force-coupled reactivity.

Herein we explore structure−activity relationships in the

mechanochemical reactivity of a subset of fused ring

cyclobutanes that are built via a modular approach, reminiscent

13,14

of the work of Weng and Boulatov (Figure 2).

The

modularity allows the force-coupled kinetics and the amount of

stored length that is released upon activation to be tuned

independently. When the stored length exceeds 3 nm per

mechanophore, we are able to resolve individual ring-opening

events via SMFS under both constant-velocity and constant-

force conditions. The resolution of individual events allows the

constant-velocity data to be converted into the equivalent

force-clamping data through a binning strategy. This analysis

unites the two data sets and expands the dynamic range of the

force-coupled kinetics experiments, allowing a fuller picture of

the mechanochemical cycloreversion reaction to be achieved.

We ﬁnd that the size of the fused macrocycle does not

substantially inﬂuence the force-coupled reactivity, whereas the

chemical structure of the pulling attachments has a profound

eﬀect. In particular, the use of 1,2-cis-diphenyl substituents as

pulling attachments provides a mechanical advantage relative

to dialkyl analogs. That mechanical advantage works in concert

with electronic stabilization of the force-coupled transition

state for cycloreversion, which combine to substantially reduce

the force required for activation for the same time scale regime

(

ms to min). The characterized force dependencies and

structural modularity of the mechanophores provide a

quantitative foundation for the use of these cyclobutane

mechanophores, in general, and their CDPs, in particular, in

force-responsive polymers and polymer networks.

EXPERIMENTAL SECTION

■

Monomer and Polymer Synthesis. The mechanophores and

polymers studied are shown in Figure 2. Cyclobutane monomers

CB1−CB3 have the same mechanophore “gate” structure, only

diﬀering in the size of the fused ring. Monomer CB4 has the same

fused ring as CB3, but diﬀers from CB1−CB3 in that the pulling

attachments are phenyl rings as opposed to methylenes. Similar

phenyl-substituted cyclobutane mechanophores have been reported

Figure 1. (a) Cartoon representation of SMFS of noncovalent (left)

and covalent (right) domain polymers. (b) Reported applications of

cyclobutane-based mechanophores.

experiments. Pill et al., for example, were able to observe the

ring-opening of a single cyclobutane mechanophore embedded

in an end-tethered polymer, and more recently Weng and

Boulatov have resolved individual ring-opening events within

previously, but we note that the synthetic strategy employed here

provides greater stereochemical ﬁdelity than the previous method,

which in turn provides greater control over single chain

mechanochemical properties. All monomers were prepared using a

modular approach, wherein the size of the fused ring is ultimately

determined by the length of the alcohol used during the ﬁrst

esteriﬁcation reaction (Figure 2). Once these linkers with terminal

alkenes are attached to the cyclobutane core, ring-closing metathesis

13,14

polymers that contain multiple cyclobutane repeats.

In one

of the latter examples, the polymer contour length more than

doubles as a result of the complete opening of all

mechanophores.

The impact of force-coupled cyclobutane cycloreversion

kinetics have also recently come to the fore in the context of

polymer network fracture. Poly(ethylene glycol) (PEG) gels

were synthesized through the end-linking of azide-terminated

(

RCM) closes the macrocycle that serves as the source of stored

length. The alkene resulting from RCM is then reduced via a

palladium-catalyzed hydrogenation so that the “handle side” of the

cyclobutane core can be functionalized with another cyclic alkene.

Rather than being used for stored length, this cyclic alkene is

copolymerized with epoxy-COD via entropy-driven ring-opening

metathesis polymerization (ED-ROMP), generating high molecular

tetra-arm PEG (M = 5 kDa) with bis-alkyne linkers. When the

bis-alkyne includes either a cis-diaryl or cis-dialkyl linked

cyclobutane “weak link” mechanophore, the tearing energy of

the gels is substantially reduced relative to that of a control

weight (MW) polymers that contain multiple cyclobutane repeats

network. In addition, the cis-diaryl cyclobutane gel is much

more easily torn than the cis-dialkyl analog. The diﬀerence in

fracture energies is correlated with qualitative characterization

of the force-coupled scission kinetics of the mechanophores

observed in single-molecule force spectroscopy experiments,

implicating local resonance stabilization of a diradical

(

P1−P4). The epoxide comonomer is mechanically inert and has

been shown to increase the adhesion force between the polymer and

the tip of the atomic force microscope (AFM) cantilever. Percent

monomer incorporation was determined by H NMR. Molecular

weights were determined by gel permeation chromatography

equipped with a multiangle light scattering detector (GPC-MALS).

270

Journal of the American Chemical Society

J. Am. Chem. Soc. 2021, 143, 5269−5276

Article

Figure 2. (a) Cartoon of monomer synthesis, highlighting the modular approach for incorporating stored length. (b) Chemical structures of

monomers and ROMP-derived copolymers.

Single Molecule Force Spectroscopy. Pulling experiments were

ﬁnal contour length after all covalent domains have unraveled (L₂).

For P3 and P4, the same approach was used to ﬁt individual domain

unfolding events within a single force curve, in order to determine the

contour length before (l ) and after (l ) a single activation. For the

conducted using a homemade AFM at ambient temperature in

30,32−34

toluene, using a similar procedure to that described previously.

Force curves used for analysis were obtained with rectangular-shaped

cantilevers (205 μm × 15 μm, nominal tip radius ∼2 nm, nominal

spring constant k ≈ 0.02 N/m, frequency ∼15 kHz). Multiple probes

of the same type were used throughout the course of the experiments.

The spring constant of each cantilever was calibrated in air, using the

thermal noise method, based on the energy equipartition theorem as

purposes of this analysis, we only measured the change in contour

length for a given event if we could resolve and successfully ﬁt two

adjacent peaks (Figures S5 and S6 ). For constant-force data, the

force-coupled length before (l ) and after (l ) each rupture event is

determined directly from steps in the distance vs time curves.

Modeling. The expected change in contour length was calculated

by employing a relaxed potential energy scan across a range of ﬁxed

end-to-end distances, similar to the well-established CoGEF method-

described previously. Cantilever tips were prepared by soaking in

piranha solution for ∼15 min at room temperature. Silicon surfaces

were prepared by soaking ∼30 min in hot piranha solution, followed

by washing with DI-water and drying under a stream of nitrogen. The

surface and cantilever were then placed in a UVO cleaner for 15 min.

After ozonolysis, the cantilever was mounted into the AFM, and ∼20

μL of a ∼0.1 mg/mL polymer solution were added to the silicon

surface and allowed to dry.

ology (Constrained Geometry simulates External Force) but without

31,32,36−39

attempting to simulate the actual reactivity.

Here, the

approach is used only to calculate monomer contour lengths as a

function of an applied force. We use the contour lengths found at

relatively high forces (e.g., forces relevant in SMFS experiments) to

extrapolate to a force-free contour length, in a computational process

that is analogous to the extrapolation of a force-free contour length

from the high-force extensional behavior of chains in the SMFS

experiments. Because we are only interested in the relative contour

lengths before and after unravelling, these calculations can be done at

a low level of theory.

Brieﬂy, the equilibrium conformers of monomers CB3 and CB4, in

both their ring-closed and ring-opened states, were minimized at

either the molecular mechanics or semiempirical levels of theory. The

end-to-end distance of a monomer was constrained until the bonding

geometries were noticeably distorted. Energy as a function of

displacement was then obtained by shortening the constraint in 0.1

Measurements were carried out in a ﬂuid cell with scanning set for

a series of constant velocity approaching/retracting cycles. To collect

constant-force” data, cantilever deﬂection was monitored during each

“

retraction cycle. Once the cantilever deﬂection reached a threshold

value (set here to 200 pN), the system was switched to the force-

control mode to achieve a desired set point force value for the

constant force measurement. This active control of force was

maintained for a set period of time (10−30 s), after which the

force-control mode was switched oﬀ and constant velocity retraction

resumed to ﬁnish the pulling cycle. During acquisition, data were

ﬁltered at 500 Hz and collected in dSPACE and Matlab.

Data Analysis. Matlab was used to analyze changes in polymer

contour length associated with mechanophore activation for both

constant-velocity and constant-force experiments. For constant-

velocity experiments, an extended freely jointed chain (EFJC)

model was used to ﬁ t the force curves before and after the transition

region (Figures S1 −S4 ), allowing for the determination of the initial

Å increments. The incremental change in energy (E

− En−1) vs

change in distance (d − dn−1) was taken as the force at the midpoint

of the increment, and the resulting force vs displacement curve was

extrapolated to zero force to give a force-free contour length (Figures

S15 −S18 ).

polymer contour length before any mechanophores react (L ) and the

271

J. Am. Chem. Soc. 2021, 143, 5269−5276

Journal of the American Chemical Society

Figure 3 a and b. As a polymer is pulled, the force along its

Article

RESULTS AND DISCUSSION

Representative distance vs time curves obtained from

constant-force SMFS, also known as force-clamp (FC)

spectroscopy, are shown in Figure 3c and d for P3 and P4.

These polymers were each clamped for ∼10−30 s at several

diﬀerent forces (∼1800 pN for P3 and ∼600 pN for P4) that

are just below the plateau forces observed in the respective

constant-velocity experiments (Figure 3a and b). If held at a

suﬃciently high force for enough time, the CB monomers react

and release slack, leading to an increase in polymer contour

length. Thus, the “plateaus” in the FC data represent how

much time the polymer spends at a given degree of unravelling

before another unravelling event occurs (which is indicated by

a sudden jump from one plateau to the next).

Domain Size. We ﬁrst analyze the amount of slack that is

released (measured by SMFS) for a typical unravelling event

and correlate slack release to the size of the fused covalent

domain. As seen in a magniﬁed section of the plateau region

for P3 and P4 (Figure 3a and b), the plateau comprises a series

of sawtooth features that are ascribed to the unravelling of

individual covalent domains. These features are detectable, but

not as easily resolved, in P1 and P2 (Figures S1 −S3), which

contain smaller fused domains than P3 and P4. The

observation that the ability to resolve the “teeth” of the

plateau depends on domain size is evidence that each tooth

likely corresponds to individual events. We therefore limit the

domain size analyses and single event-based extraction of

force-dependent kinetics to polymers P3 and P4.

■

The ED-ROMP methodology yielded high MW copolymers

that contain multiple fused-ring cyclobutanes. Polymer

molecular weight (GPC-MALS) and monomer content ( H

NMR) are summarized in Table 1. For polymers P3 and P4,

multiple polymers with varying composition and molecular

weight were synthesized and characterized.

Table 1. Summary Characterization Data of Polymers

Employed in This Study

Polymer

M (kDa)

M_w(kDa)

% CB

% epoxide

245

226

370

266

219

289

171

139

329

356

474

425

268

449

313

160

1.3

1.6

1.3

1.6

1.2

1.6

1.8

1.2

P3-a

P3-b

P3-c

P3-d

P4-a

P4-b

Molecular weights were determined by GPC-MALS and monomer

content by H NMR.

Representative force vs separation curves obtained from

constant-velocity (CV) SMFS for each polymer are shown in

For P3 and P4, each “tooth” of the sawtooth pattern is ﬁt

with an EFJC model using the Kuhn length and segment

elasticity parameters obtained on nascent polymer, as

described above, in order to determine the change in contour

length for individual ring-opening events. The frequency

histogram of the aggregate data (0.2 nm bin size) was ﬁt to

a Gaussian distribution (Figure 4a and b), and the average

change in contour length per event was found to be 3.16 ±

.21 nm and 3.13 ± 0.35 nm for P3 and P4, respectively.

Uncertainties correspond to the standard deviation of the

Gaussian ﬁt. These experimental values are consistent with

values of 3.13 and 3.40 nm obtained from calculations on P3

and P4, respectively (Figure 4e and f).

A similar analysis is applied to the staircase pattern of the

corresponding FC curves (Figure 3c and d), where the

transition between steps corresponds to the unravelling of

individual covalent domains. The distribution of step sizes

(

Figure 4c and d) is again characterized and used to determine

the average change in length per single event. Because the force

range explored in FC experiments is narrow (less than 100 pN

per polymer), we aggregate data across all pulls. Fitting the

aggregate data (0.2 nm bin size) with a Gaussian distribution

yields average values of 3.30 ± 0.39 nm for P3 and 3.42 ± 0.51

nm for P4, which are in agreement with those extracted from

constant-velocity experiments and those predicted by the

CoGEF calculations for a single activation event. The

uncertainty in the extrapolations precludes a deep comparison

of the extensions between the two experiments. We note,

however, that it makes sense for the changes in contour length

extracted from the constant velocity and FC measurements to

diﬀer, because the FC analysis obtains the change in contour

length under an applied force, whereas the constant velocity

curves are extrapolated back to zero force.

Figure 3. (a and b) Representative normalized force vs separation

curves for polymers P1−P3 (a) and P4 (b), obtained by constant-

velocity SMFS. Insets highlight the sawtooth pattern observed in the

plateau regions for P3 and P4. (c and d) Representative distance vs

time curves for P3 at 1860 pN (c) and P4 at 600 pN (d), obtained by

constant-force SMFS. Insets highlight the staircase pattern and

extraction of stored length release per activation.

backbone increases, and the probability of mechanophore

unravelling increases with it. The force continues to increase

until the rate at which stored length is released becomes

competitive with the rate at which the polymer is being

stretched (300 nm/s). The competition between these two

processes results in a plateau in the force−extension curve,

where the overall force remains relatively constant as the

contour length of the polymer increases.

Force-Coupled Kinetics. Before discussing the extraction

of force-dependent kinetics from the data, we make an initial,

qualitative observation about the kinetics by comparing the

272

Journal of the American Chemical Society

J. Am. Chem. Soc. 2021, 143, 5269−5276

from CV force −extension curves for P3 and P4 using a bin-

Article

methylene, CH(CH )(C H ), is ∼13 kcal/mol lower than the

CH bond of the analogous methyl-substituted methylene,

CH(CH ) . In addition to this electronic eﬀect, we also

considered that the phenyl substituents might provide better

mechanical leverage compared to the alkyl attachments, as

31,38

demonstrated in other mechanophores.

Such mechanical

lever-arm eﬀects would also inﬂuence the force-coupled

kinetics.

Diﬀerences in mechanical leverage correspond to diﬀerences

in force sensitivity−how relative rate changes as a function of

force. The force-dependent kinetics were extracted via two

approaches: (i) survival statistics that take into account the

changing force of the constant velocity experiments and (ii)

survival statistics in the constant force experiments. The latter

approach, based on FC spectroscopy, is often used to study

,41,42

protein unfolding,

but it has been employed to study

34,39

covalent mechanophores in only rare instances.

We ﬁt the

distance vs time curves with a single exponential decay

function to provide k(f) at several diﬀerent forces for both P3

and P4. The various rate constants are plotted as a function of

the force in Figure 5 (“×” symbols).

Additional kinetic data (Figure 5, circles) were extracted

Figure 4. (a−d) Histograms of changes in polymer contour length

per detected activation, obtained from constant-velocity (a and b) and

constant-force (c and d) experiments for polymers P3 (a,c) and P4

(b,d). All bin sizes are 0.2 nm, and the histograms were ﬁt with a

Gaussian distribution (black curve). (e and f) Representative

structures of mechanophore repeats and their associated end-to-end

distances in both their nascent and fully unraveled states, obtained

from CoGEF calculations.

transition forces (the force at which the plateau occurs) for

P1−P4 that are observed in the constant-velocity experiments.

For P1−P3, where the only diﬀerence in the monomer

structures is the size of the fused ring, we see that the

transitions in the force curves overlay (Figure 3a), indicating

that, within the range of ring sizes studied here, a change in the

size of the fused macrocycle does not lead to substantial

changes in the force-coupled reactivity of the cyclobutane, and

that the amount of stored length can be treated as a parameter

that can be tuned independently within this class of

mechanophores.

As seen in comparing Figure 3a to Figure 3b, however, the

attachments (phenyl vs alkyl) through which the polymer is

coupled to the cyclobutene have a major inﬂuence on the

transition force. The transition force of P4 is ∼1300 pN lower

than that of P1−P3. Boulatov and co-workers have recently

shown experimentally that the stereochemistry of phenyl

pulling attachments can greatly alter the force-coupled

reactivity of cyclobutane, but we are not aware of previous

experimental comparisons of diﬀerent handles with the same

stereochemistry (here, “cis-pulling”) in cyclobutanes. The

observed diﬀerence in reactivity between phenyl and alkyl

pulling attachments can be partially explained by diﬀerences in

the ability of the substituent to stabilize the radicals that are

Figure 5. Rate vs force plots for polymers P3 (a) and P4 (b). “×”

symbols correspond to data from constant-force experiments, and

circles correspond to data from constant-velocity experiments. Slopes

of linear ﬁts of the combined data are reported, along with the

16,23

presumably generated from the ﬁrst bond-breaking event;

the CH bond dissociation energy of a phenyl-substituted

‡

mechanochemical coupling parameter, Δx , obtained from eq 1.

273

J. Am. Chem. Soc. 2021, 143, 5269−5276

Journal of the American Chemical Society

independently by Osterhelt.

Article

based approach previously developed by our group and then

Lastly, the reaction lengths determined here (∼1 Å) are far

smaller than the change in contour length expected for the

formation of a fully open 1,4-diradicaloid intermediate or its

subsequent disproportionation into alkenes. The empirical

activation lengths are therefore consistent with a mechanistic

picture in which the homolytic scission of the ﬁrst bond is rate-

determining at the forces examined here, as proposed by

43,44

Full details are provided in the

Supporting Information, but, brieﬂy, Matlab is used to count

the total time spent by cyclobutane mechanophores within a

given force range t(f), where f here refers to all forces that fall

within a range (typically 20−50 pN). The residence time t(f)

is determined by multiplying the number of digital data points

found within a given force range by the acquisition time and

the number of cyclobutanes in the loaded polymer strand that

are still intact at that time each data point is recorded. The

number of bond ruptures in a given force range, N(f), is

counted using Matlab’s built-in findpeaks function. Adjust-

ments are made to account for the fact that some single events

are unable to be resolved, and double or triple events (<25%)

are determined based on the measured change in strand length

and accounted for in the analysis (Figures S7 −S9). Note that

t(f) and N(f) are usually taken from multiple curves with

diﬀerent numbers of rupture events, since a single curve might

not have enough rupture events for analysis. We then total t(f)

and N(f) across several pulls to obtain a group of t (f ) and

N (f) data with more than 100 rupture events (Tables S1 and

S2) and calculate the force-dependent event frequency, k(f) =

Boulatov. Furthermore, the enhanced mechanical leverage

provided by the phenyl group suggests that this scission event

likely involves an increase in the torsional angle between the

C−C bond that bridges the fused rings and the C−C bond

connecting the cyclobutane to the pulling attachment. A given

increase in bond length of the scissile carbon−carbon displaces

the two pulling attachments to the same extent, but

contributions from outwardly rotating attachments would be

sensitive to the increased length of the phenyl “lever” relative

to the methylene, as observed in electrocyclic reactions of

alkene-substituted cyclopropanes.

CONCLUSION

■

Cycloreversion releases an amount of stored length that is

determined by the size of a fused macrocycle, and large

unravelling events can be observed as single events in each of

two diﬀerent single molecule measurements. This allows us to

characterize contour length changes associated with domain

unravelling as well as details of the relative force dependencies.

Here, the statistics of ring-opening are well-described by

models in which the mechanophores act as independent

reactants with identical force-coupled reaction probabilities;

reaction at one mechanophore does not detectably inﬂuence

reactivity at another. The apparent independence of

mechanophore reactivity need not be general, however, and

approaches that permit stochastic events to be observed might

prove useful in probing such systems. In addition, the

∑

N(f )/∑ t(f ). We calculate the mean and standard error of

k(f) by averaging values obtained from multiple groups.

The combined constant-velocity and constant-force data are

ﬁt with a log−linear relationship (Figure 5), in accordance with

the simplest model of mechanochemical coupling (eq 1).

FΔx^‡

ln(k(f )) = C +

(1)

As seen in Figure 5, the consensus data obtained by FC and

CV measurements are well ﬁt by eq 1 across force-coupled rate

−

constants from approximately 0.05 to 20 s , with correspond-

‡

ing empirical mechanochemical coupling parameters, Δx , of

14,25,50

dynamics of mechanically coupled tandem processes

‡

.81 ± 0.006 Å for CB3 and 1.46 ± 0.09 Å for CB4. Both Δx

are proving to be rich territory for mechanistic investigation,

and with suﬃcient resolution single-event measurements might

eventually be used to capture the force-coupled lifetimes and

subsequent fate of reactive intermediates.

The diﬀering reactivities of the two cyclobutane mechano-

phores arises from a combination of electronic and mechanical

values obtained here are within experimental uncertainty of the

approximated values derived from calculations previously

reported by Boulatov and co-workers on similar cyclobutanes

(

∼0.8 Å for diester “handles” and ∼1.4 Å for diphenyl

“handles”). The SMFS results therefore provide experimental

support for the mechanism proposed therein, in which initial

“

lever arm” eﬀects of the phenyl substituent relative to alkyl

bond breaking to a diradical intermediate is the rate-limiting

substituents. Similar eﬀects have been observed in electrocyclic

reactions, but we believe this to be the ﬁrst experimental

characterization of such an eﬀect for a homolytic reaction. It

suggests a torsional component to the mechanical scission

reaction that need not be present in the force-free reaction.

Regardless of the origin, the diﬀerences in reactivity make

these cyclobutane promising candidates to help address

fundamental questions of molecular extension, including the

release of stored length, that accompanies the macroscopic

stress−strain behavior of macroscopic polymer networks. We

also note that, when the stored length is removed, these

substituted cyclobutanes might continue to be used as weak

bonds with characterized scission behavior to probe the

contributions of chain scission to polymers and polymer

networks.

step of the net cycloreversion. In addition, the consistency

within the FC and CV measurements indicates the success of

the “binning” methodology in converting CV data into its FC

equivalent.

A number of contributions go into the force dependency of

reaction rate, most notably distortions of the reactant and

transition state and the force-coupled change in length

45−47

between the reactant and transition state.

The depend-

ence changes with force, often in complicated ways that

48,49

include contributions from multiple conformations,

as well

as competing reaction paths and changes in mechanism that

have been shown to be relevant to the mechanical dissociation

3,24

of cyclobutanes.

Nonetheless, given the similarity in the

core reactant of both systems, the most likely contribution to

the diﬀerence in the two systems is the better geometric

coupling of the reaction path to the stretching force. In other

ASSOCIATED CONTENT

sı Supporting Information

■

words, the phenyl groups of P4 act as metaphorical crowbars,

relative to their alkyl counterparts in P3, allowing the applied

force to be transmitted across a greater distance as the bond-

breaking reaction proceeds.

The Supporting Information is available free of charge at

https://pubs.acs.org/doi/10.1021/jacs.1c02149.

274

J. Am. Chem. Soc. 2021, 143, 5269−5276

Journal of the American Chemical Society

(1) Tskhovrebova, L.; Trinick, J.; Sleep, J. A.; Simmons, R. M.

Article

REFERENCES

Synthetic procedures, SMFS data collection and analysis

■

(PDF )

Elasticity and Unfolding of Single Molecules of the Giant Muscle

Protein Titin. Nature 1997, 387, 308−312.

2) Rief, M.; Gautel, M.; Oesterhelt, F.; Fernandez, J. M.; Gaub, H.

AUTHOR INFORMATION

■

E. Reversible Unfolding of Individual Titin Immunoglobulin Domains

by Afm. Science 1997, 276, 1109.

Corresponding Authors

Bradley D. Olsen − NSF Center for the Chemistry of

Molecularly Optimized Networks, Duke University, Durham,

North Carolina 27708, United States; Department of

Chemical Engineering, Massachusetts Institute of Technology,

Cambridge, Massachusetts 02139, United States;

orcid.org/0000-0002-7272-7140; Email: bdolsen@

mit.edu

(3) Oberhauser, A. F.; Hansma, P. K.; Carrion-Vazquez, M.;

Fernandez, J. M. Stepwise Unfolding of Titin under Force-Clamp

Atomic Force Microscopy. Proc. Natl. Acad. Sci. U. S. A. 2001, 98,

4) Linke, W. A.; Grutzner, A. Pulling Single Molecules of Titin by

68−472.

Afm Recent Advances and Physiological Implications. Pfluegers

Arch. 2008, 456, 101−115.

(5) Chen, H.; Yuan, G.; Winardhi, R. S.; Yao, M.; Popa, I.;

Michael Rubinstein − NSF Center for the Chemistry of

Molecularly Optimized Networks, Department of Chemistry,

and Departments of Physics, Mechanical Engineering and

Materials Science, and Biomedical Engineering, Duke

University, Durham, North Carolina 27708, United States;

World Premier Institute for Chemical Reaction Design and

Discovery, Hokkaido University, Sapporo, Japan;

orcid.org/0000-0002-5455-0951 ;

Fernandez, J. M.; Yan, J. Dynamics of Equilibrium Folding and

Unfolding Transitions of Titin Immunoglobulin Domain under

Constant Forces. J. Am. Chem. Soc. 2015, 137, 3540−3546.

(6) Rivas-Pardo, J. A.; Eckels, E. C.; Popa, I.; Kosuri, P.; Linke, W.

A.; Fernández, J. M. Work Done by Titin Protein Folding Assists

Muscle Contraction. Cell Rep. 2016, 14, 1339−1347.

7) Eckels, E. C.; Tapia-Rojo, R.; Rivas-Pardo, J. A.; Fernández, J. M.

The Work of Titin Protein Folding as a Major Driver in Muscle

Contraction. Annu. Rev. Physiol. 2018 , 80 , 327 −351.

Email: michael.rubinstein@duke.edu

(8) Freundt, J. K.; Linke, W. A. Titin as a Force-Generating Muscle

Stephen L. Craig − NSF Center for the Chemistry of

Molecularly Optimized Networks and Department of

Chemistry, Duke University, Durham, North Carolina

Protein under Regulatory Control. J. Appl. Physiol. 2019, 126, 1474−

482.

(9) Guan, Z.; Roland, J. T.; Bai, J. Z.; Ma, S. X.; McIntire, T. M.;

Nguyen, M. Modular Domain Structure: A Biomimetic Strategy for

Advanced Polymeric Materials. J. Am. Chem. Soc. 2004, 126, 2058−

7708, United States; orcid.org/0000-0002-8810-0369;

Email: stephen.craig@duke.edu

2065.

Authors

(10) Fang, J.; Mehlich, A.; Koga, N.; Huang, J.; Koga, R.; Gao, X.;

Hu, C.; Jin, C.; Rief, M.; Kast, J.; Baker, D.; Li, H. Forced Protein

Unfolding Leads to Highly Elastic and Tough Protein Hydrogels. Nat.

Commun. 2013, 4, 2974.

Brandon H. Bowser − NSF Center for the Chemistry of

Molecularly Optimized Networks and Department of

Chemistry, Duke University, Durham, North Carolina

(11) Chung, J.; Kushner, A. M.; Weisman, A. C.; Guan, Z. Direct

7708, United States; orcid.org/0000-0002-8313-9301

Correlation of Single-Molecule Properties with Bulk Mechanical

Shu Wang − NSF Center for the Chemistry of Molecularly

Optimized Networks and Department of Chemistry, Duke

University, Durham, North Carolina 27708, United States

Tatiana B. Kouznetsova − NSF Center for the Chemistry of

Molecularly Optimized Networks and Department of

Chemistry, Duke University, Durham, North Carolina

Performance for the Biomimetic Design of Polymers. Nat. Mater.

014, 13, 1055−1062.

(12) Bowser, B. H.; Ho, C. H.; Craig, S. L. High Mechanophore

Content, Stress-Relieving Copolymers Synthesized Via Raft Polymer-

ization. Macromolecules 2019, 52, 9032−9038.

(

13) Zhang, H.; Li, X.; Lin, Y.; Gao, F.; Tang, Z.; Su, P.; Zhang, W.;

Xu, Y.; Weng, W.; Boulatov, R. Multi-Modal Mechanophores Based

on Cinnamate Dimers. Nat. Commun. 2017, 8, 1147.

7708, United States

Haley K. Beech − NSF Center for the Chemistry of

Molecularly Optimized Networks, Duke University, Durham,

North Carolina 27708, United States; Department of

Chemical Engineering, Massachusetts Institute of Technology,

Cambridge, Massachusetts 02139, United States;

orcid.org/0000-0003-3276-8578

(14) Tian, Y.; Cao, X.; Li, X.; Zhang, H.; Sun, C.-L.; Xu, Y.; Weng,

W.; Zhang, W.; Boulatov, R. A Polymer with Mechanochemically

Active Hidden Length. J. Am. Chem. Soc. 2020, 142, 18687−18697.

(15) Levy, A.; Goldstein, H.; Brenman, D.; Diesendruck, C. E. Effect

of Intramolecular Crosslinker Properties on the Mechanochemical

Fragmentation of Covalently Folded Polymers. J. Polym. Sci., Part A:

Polym. Chem. 2020, 58, 692−703.

Complete contact information is available at:

https://pubs.acs.org/10.1021/jacs.1c02149

(16) Chen, Z.; Mercer, J. A. M.; Zhu, X.; Romaniuk, J. A. H.;

Pfattner, R.; Cegelski, L.; Martinez, T. J.; Burns, N. Z.; Xia, Y.

Mechanochemical Unzipping of Insulating Polyladderene to Semi-

conducting Polyacetylene. Science 2017, 357, 475.

Author Contributions

^#B.H.B. and S.W. contributed equally.

(

17) Craig, S. L. Up Another Rung. Nat. Chem. 2017, 9, 1154.

18) Wang, S.; Panyukov, S.; Rubinstein, M.; Craig, S. L.

Notes

Quantitative Adjustment to the Molecular Energy Parameter in the

Lake-Thomas Theory of Polymer Fracture Energy. Macromolecules

The authors declare no competing ﬁnancial interest.

(

019, 52, 2772−2777.

19) Pill, M. F.; Holz, K.; Preußke, N.; Berger, F.; Clausen-

Schaumann, H.; Luning, U.; Beyer, M. K. Mechanochemical

ACKNOWLEDGMENTS

■

Cycloreversion of Cyclobutane Observed at the Single Molecule

This work was funded by Duke University and the Center for

the Chemistry of Molecularly Optimized Networks

Level. Chem. - Eur. J. 2016, 22, 12034−12039.

(20) Yang, J.; Horst, M.; Romaniuk, J. A. H.; Jin, Z.; Cegelski, L.;

Xia, Y. Benzoladderene Mechanophores: Synthesis, Polymerization,

and Mechanochemical Transformation. J. Am. Chem. Soc. 2019, 141,

6479−6483.

(MONET), a National Science Foundation (NSF) Center

for Chemical Innovation (CHE-1832256).

275

J. Am. Chem. Soc. 2021, 143, 5269−5276

Journal of the American Chemical Society

Distal Conformational Locks on Ferrocene Mechanophores Guide

Article

(

21) Li, M.; Zhang, H.; Gao, F.; Tang, Z.; Zeng, D.; Pan, Y.; Su, P.;

Reaction Pathways for Increased Mechanochemical Reactivity. Nat.

Chem. 2021, 13, 56−62.

(40) Hioe, J.; Zipse, H. Radical Stability and Its Role in Synthesis

and Catalysis. Org. Biomol. Chem. 2010, 8, 3609 −3617.

(41) Schlierf, M.; Li, H.; Fernandez, J. M. The Unfolding Kinetics of

Ubiquitin Captured with Single-Molecule Force-Clamp Techniques.

Proc. Natl. Acad. Sci. U. S. A. 2004 , 101 , 7299.

Ruan, Y.; Xu, Y.; Weng, W. A Cyclic Cinnamate Dimer

Mechanophore for Multimodal Stress Responsive and Mechanically

Adaptable Polymeric Materials. Polym. Chem. 2019, 10, 905−910.

22) Kean, Z. S.; Black Ramirez, A. L.; Yan, Y.; Craig, S. L.

Bicyclo[3.2.0]Heptane Mechanophores for the Non-Scissile and

Photochemically Reversible Generation of Reactive Bis-Enones. J.

Am. Chem. Soc. 2012, 134, 12939−12942.

23) Kean, Z. S.; Niu, Z.; Hewage, G. B.; Rheingold, A. L.; Craig, S.

(42) Hoffmann, T.; Dougan, L. Single Molecule Force Spectroscopy

Using Polyproteins. Chem. Soc. Rev. 2012, 41, 4781−4796.

(43) Serpe, M. J.; Kersey, F. R.; Whitehead, J. R.; Wilson, S. M.;

Clark, R. L.; Craig, S. L. A Simple and Practical Spreadsheet-Based

Method to Extract Single-Molecule Dissociation Kinetics from

Variable Loading-Rate Force Spectroscopy Data. J. Phys. Chem. C

2008, 112, 19163−19167.

L. Stress-Responsive Polymers Containing Cyclobutane Core

Mechanophores: Reactivity and Mechanistic Insights. J. Am. Chem.

Soc. 2013, 135, 13598−13604.

(44) Oberbarnscheidt, L.; Janissen, R.; Oesterhelt, F. Direct and

24) Wang, J.; Kouznetsova, T. B.; Boulatov, R.; Craig, S. L.

Mechanical Gating of a Mechanochemical Reaction Cascade. Nat.

Commun. 2016, 7, 13433.

(

25) Chen, Z.; Zhu, X.; Yang, J.; Mercer, J. A. M.; Burns, N. Z.;

Model Free Calculation of Force-Dependent Dissociation Rates from

Force Spectroscopic Data. Biophys. J. 2009 , 97 , L19 −L21.

Martinez, T. J.; Xia, Y. The Cascade Unzipping of Ladderane Reveals

Dynamic Effects in Mechanochemistry. Nat. Chem. 2020, 12, 302−

(45) Huang, Z.; Boulatov, R. Chemomechanics: Chemical Kinetics

309.

for Multiscale Phenomena. Chem. Soc. Rev. 2011 , 40 , 2359 −2384.

46) Akbulatov, S.; Boulatov, R. Experimental Polymer Mechano-

chemistry and Its Interpretational Frameworks. ChemPhysChem 2017,

8, 1422−1450.

47) Tian, Y.; Boulatov, R. Comparison of the Predictive

Performance of the Bell −Evans, Taylor-Expansion and Statistical-

Mechanics Models of Mechanochemistry. Chem. Commun. 2013, 49,

(

26) Yang, J.; Horst, M.; Werby, S. H.; Cegelski, L.; Burns, N. Z.;

Xia, Y. Bicyclohexene-Peri-Naphthalenes: Scalable Synthesis, Diverse

Functionalization, Efficient Polymerization, and Facile Mechanoacti-

vation of Their Polymers. J. Am. Chem. Soc. 2020, 142, 14619−14626.

Craig, S. L. Molecular Stress Relief through a Force-Induced

27) Wu, D.; Lenhardt, J. M.; Black, A. L.; Akhremitchev, B. B.;

Irreversible Extension in Polymer Contour Length. J. Am. Chem.

Soc. 2010, 132, 15936−15938.

48) Boulatov, R. The Challenges and Opportunities of Contem-

187−4189.

porary Polymer Mechanochemistry. ChemPhysChem 2017, 18, 1419−

421.

49) Akbulatov, S.; Tian, Y.; Huang, Z.; Kucharski, T. J.; Yang, Q.-

(

28) Wang, J.; Kouznetsova, T. B.; Niu, Z.; Ong, M. T.; Klukovich,

H. M.; Rheingold, A. L.; Martinez, T. J.; Craig, S. L. Inducing and

Quantifying Forbidden Reactivity with Single-Molecule Polymer

Mechanochemistry. Nat. Chem. 2015, 7, 323−327.

(

Z.; Boulatov, R. Experimentally Realized Mechanochemistry Distinct

from Force-Accelerated Scission of Loaded Bonds. Science 2017, 357,

(

29) Wang, S.; Beech, H. K.; Bowser, B. H.; Kouznetsova, T. B.;

Olsen, B. D.; Rubinstein, M.; Craig, S. L. Mechanism Dictates

Mechanics: A Molecular Substituent Effect in the Macroscopic

Fracture of a Covalent Polymer Network. J. Am. Chem. Soc. 2021, 143,

50) Nixon, R.; De Bo, G. Three Concomitant C −C Dissociation

99.

Pathways During the Mechanical Activation of an N-Heterocyclic

31) Klukovich, H. M.; Kouznetsova, T. B.; Kean, Z. S.; Lenhardt, J.

714−3718.

30) Lin, Y.; Kouznetsova, T. B.; Craig, S. L. Mechanically Gated

Degradable Polymers. J. Am. Chem. Soc. 2020, 142, 2105−2109.

Carbene Precursor. Nat. Chem. 2020, 12, 826−831.

(

M.; Craig, S. L. A Backbone Lever-Arm Effect Enhances Polymer

Mechanochemistry. Nat. Chem. 2013, 5, 110−114.

(32) Lin, Y.; Kouznetsova, T. B.; Craig, S. L. A Latent Mechanoacid

for Time-Stamped Mechanochromism and Chemical Signaling in

Polymeric Materials. J. Am. Chem. Soc. 2020, 142, 99−103.

(33) Barbee, M. H.; Kouznetsova, T.; Barrett, S. L.; Gossweiler, G.

R.; Lin, Y.; Rastogi, S. K.; Brittain, W. J.; Craig, S. L. Substituent

Effects and Mechanism in a Mechanochemical Reaction. J. Am. Chem.

Soc. 2018, 140, 12746−12750.

(34) Kouznetsova, T. B.; Wang, J.; Craig, S. L. Combined Constant-

Force and Constant-Velocity Single-Molecule Force Spectroscopy of

the Conrotatory Ring Opening Reaction of Benzocyclobutene.

ChemPhysChem 2017, 18, 1486−1489.

(35) Florin, E. L.; Rief, M.; Lehmann, H.; Ludwig, M.; Dornmair, C.;

Moy, V. T.; Gaub, H. E. Sensing Specific Molecular Interactions with

the Atomic Force Microscope. Biosens. Bioelectron. 1995, 10, 895−

(

01.

36) Wang, J.; Kouznetsova, T. B.; Kean, Z. S.; Fan, L.; Mar, B. D.;

Martínez, T. J.; Craig, S. L. A Remote Stereochemical Lever Arm

Effect in Polymer Mechanochemistry. J. Am. Chem. Soc. 2014, 136,

37) Gossweiler, G. R.; Kouznetsova, T. B.; Craig, S. L. Force-Rate

5162−15165.

Characterization of Two Spiropyran-Based Molecular Force Probes. J.

Am. Chem. Soc. 2015, 137, 6148−6151.

(38) Wang, J.; Kouznetsova, T. B.; Niu, Z.; Rheingold, A. L.; Craig,

S. L. Accelerating a Mechanically Driven Anti-Woodward-Hoffmann

Ring Opening with a Polymer Lever Arm Effect. J. Org. Chem. 2015,

0, 11895−11898.

39) Zhang, Y.; Wang, Z.; Kouznetsova, T. B.; Sha, Y.; Xu, E.;

Shannahan, L.; Fermen-Coker, M.; Lin, Y.; Tang, C.; Craig, S. L.

(

276