Repositioning and increasing the weight of borders around dendrogram produced with R 'plot' function

I am trying to cut a dendrogram into three classes using the rect.hclust function, but when I export the graph, it cuts off the borders at the bottom of the graph. In addition, I would like to increase the weight of the borders, but I am not sure how to do this, as the lwd argument doesn't seem to exist for this function. What can I do to fix these parameters?

Data:

cluster <- data.frame(plot=c(1,
                             2,
                             3,
                             4,
                             5,
                             6,
                             7,
                             8,
                             9,
                             10,
                             11,
                             12,
                             13,
                             14,
                             15,
                             16,
                             17,
                             18,
                             19,
                             20),
                      meas1 = c(443,
                                836,
                                903,
                                684,
                                94,
                                125,
                                733,
                                846,
                                625,
                                234,
                                437,
                                775,
                                269,
                                774,
                                17,
                                502,
                                80,
                                51,
                                523,
                                229),
                      meas2 = c(735,
                                574,
                                793,
                                261,
                                961,
                                136,
                                404,
                                138,
                                45,
                                935,
                                698,
                                675,
                                594,
                                497,
                                152,
                                153,
                                30,
                                667,
                                547,
                                745),
                      meas3 = c(23,
                                526,
                                36,
                                93,
                                708,
                                970,
                                399,
                                111,
                                456,
                                439,
                                569,
                                503,
                                337,
                                213,
                                399,
                                850,
                                614,
                                491,
                                28,
                                452))

Code for hierarchical clustering and dendrogram generation:

#Generate distance matrix
dist_mat <- dist(cluster, method = 'euclidean')

#Hierarchical clustering 
hclust_avg <- hclust(dist_mat, method = 'average')

#Cut into 3 classes
cut_avg <- cutree(hclust_avg, k = 3)

#Plot dendrogram
plot(hclust_avg, xlab = "", ylab = "Euclidean Distance", sub = "", 
     main = "Cluster Dendrogram", lwd = 2)

#Generate borders around each group
rect.hclust(hclust_avg, k=3, border = 2:6)

Exporting the graph looks as follows:

I need the red-green-blue borders to be thicker (higher border weights), and I also need them to not be cut off at the bottom of the graph.